HKSAR v. CHAN KIT YI AND ANOTHER

DCCC 684/2017

[2020] HKDC 628

IN THE DISTRICT COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

CRIMINAL CASE NO 684 OF 2017

---------------------------

----------------------------

-----------------------------------------

REASONS FOR VERDICT

-----------------------------------------

1.  D1 and D2 are jointly charged with two counts of dealing in arms without a licence contrary to section 14 of the Firearms and Ammunition Ordinance, Cap 238. It is alleged that they, on or about 5 June 2015 (1st charge) and 3 August 2015 (2nd charge), respectively, at MH (Trading) Company in Hong Kong, by way of trade or business, dealt in arms without a licence, namely selling a stunning device. Both defendants pleaded not guilty.

Prosecution case

2.  Prosecution has called seven witnesses, the first two being DPC48019 (“PW1”) and DPC8607 (“PW2”).  Their evidence is largely not in dispute and is embodied also in the Admitted Facts. 

3.  The prosecution relied on the expert evidence of Professor Carmen C Y Poon (“PW3”).  They also rely on the evidence of Mr Tang Kwok Ming (“PW6”) and Chan Tak Kwong James(“PW7”).

4.  As for Dr Ng Chung Ki (“PW4”) and Dr Chiao Wing Fu (“PW5”), both senior forensic pathologists and forensic pathologists, the prosecution does not rely on their evidence.  They were merely tendered for cross-examination at the request of the defence.

Admitted Facts

5.  Under section 65C of the Criminal Procedure Ordinance, Cap 221, parties admitted, inter alia, that on 5 June 2015, PW1 posed as a customer, went to MH Trading Company (the shop) and asked to buy a lighter device on display.  D1, then a salesperson of the shop, took out a lighter device of the same model.  PW1 then bought the same, Exhibit P1, from D1, and left the shop.

6.  Likewise, on 3 August 2015, PW2 again posed as a customer again, went to the shop.  Again, D1 was the salesperson there.  PW2 pointed at a lighter device on display in the glass showcase and inquired D1 of the same.  D1 then took out another similar lighter device from the cashier countertop and explained to him the said device and demonstrated to him how to use the same.

7.  During this time, D1 pressed down the operating switch of the device, Exhibit P2.  It showed electric current generated between the metal tips of the device in form of a blue electric arc.  Later, PW2 paid D1 $398 and bought another lighter device of the same model placed in the cabinet in the shop, Exhibit P12.

8.  Then PW2 revealed his identity and the shop was searched.  Police seized Exhibit P2 and 24 other lighter devices from within the shop.

9.  D2 was the owner of the shop at the time of the two offences and he was also the employer of D1.  He was the one responsible for buying goods for the shop, including Exhibit P1, P2, P12, and the other seized exhibits.  D1 was employed at the shop since year 2014.  Apart from the lighter devices, the shop also displayed and sold ornaments like necklaces and earrings.

10.  After D1 was arrested, under caution she says that she only worked there.  The lighter arc devices were bought by her boss, D2, and she did not know anything.  D1 had later voluntarily taken part in four cautioned interviews.  Accurate records of such interviews are Exhibits P7 to P10. 

11.  D2 was arrested on 4 August 2015.  He had voluntarily taken part in a video-recorded interview.  The same is accurately recorded in the redacted Exhibit P11 and its transcript Exhibit P11A.

12.  Parties also agree that a PowerPoint headed, “Probe mode mechanism.” was issued by Medical and Health Department in June 2017, (Exhibit D2) and one of the references referred to in the PowerPoint is, “Conducted Electrical Weapon Drive - Stun Mode, Skin Rub vs Injection,” by Mark W Kroll (Exhibit D3).

13.  D1 has one previous conviction not similar and D2 has a clear record.

Reports of Dr LAM Wai-kwok

14.  The reports of Dr Lam Wai Kwok in relation to another case involving electric arc lighters with probe separation of 5 millimetres were also produced under section 65B of Criminal Procedure Ordinance (Exhibit P22 and P23).

15.  Dr Lam was not called during trial. 

16.  In his two reports, he said based on the tests performed by the COMMS, where the peak-to-peak pulsating voltage generated from the EUT when connected to resistor load of resistance close to human body was measured at 46,799 volts, the same was able to generate electric arcs from the electrodes on the head and the same was able to produce 44,643 high voltage pulses continuously in three seconds duration, based on the electrical properties and the electrical characteristics of the EUT, the devices are similar to stunning devices. 

17.  It was said also that although there was a protective cover on top of the EUT, when the protective cover was open, the electrodes on the head of the EUT were exposed and although the space and the current between the 5-millimetre distance electrodes was relatively small, the electrodes were relatively less exposed, it is still possible to apply the electrodes of EUT on human and cause shock.

18.  It was said that the effect of such device when applied on humans, depend on the number of factors such as the site and duration of the application, tolerance to pain and psychological preparedness.  It was said that generally, if applied on an average healthy person on brief application, subject could be startled and would experience intense local pain and muscle spasm.  If the application lasts 3 to 5 seconds, the subject may fall onto the ground due to muscle paralysis and remain dazed, weak or even immobilised for a while, possibly up to several minutes.

19.  He mentioned that while a Taser will last for 5 seconds, some models of stun devices lack features which guarantee a substantive discharge and so result in brief electric shock resulting in local pain and muscle spasm.  Dr Lam said, referring to the 5-millimetre electrode separation distance, that it is difficult to sustain the electric discharge as the human subject may move away from the initial contact with the electrode easily and, therefore, not result in neuromuscular incapacitation.

Records of interview

20.  D1 said in these interviews that the electric arc lighters were all bought by D2.  She had no idea that the selling of these lighters would be an offence.  She said D2 demonstrated to her how to operate these electric arc lighters.  She had no idea of the energy output of these lighters.  She was also asked to sell them at $398.

21.  D2 admitted in the interview that he had bought the electric arc lighters from mainland.  They were later delivered to Hong Kong.  D2 said he knew electric current would be generated when the operating switch was pressed, but he did not know the electricity output of these electric arc lighters.  He said he had read the instructions manual.

22.  He said he was told by the seller of these lighters that if one continuously pressed the switch of the lighters for 15 seconds the circuit would be cut.  He was also told that the electric current required to charge the lighter is small and the same is akin to that of a mobile phone.  He was also told that it was absolutely safe.

Exhibit evaluation reports

23.  Mr Tang Kwok Ming (“PW6”) had been the assistant police telecommunication inspector since 1 May 1989.  Since 2 March 2015, he was posted to the Communication Branch Operations and Support Service Division, Electric System Support Section.  The same was engaged by police to perform technical tests and to prepare technical reports on these suspected stunning devices.  These devices under tests were called “EUT”, exhibit under test.

24.  As a result of the tests on Exhibit P1 (EUT1), two exhibit evaluation reports were prepared, dated 18 June 2015 and 24 January 2017 by PW6 and endorsed by Mr Chan Tak Kwong (“PW7”), Exhibits P24 and P25.  As a result of the test on Exhibit P2 (EUT8), exhibit evaluation reports were similarly produced by PW6 and endorsed by PW7, Exhibit P26.  As for the rest 24 electric arc lighters, similar tests were performed and two exhibit evaluation reports were prepared by PW6 and endorsed by PW7, Exhibit P27 and P28.

25.  Exhibit P24 and P25 are both exhibit evaluation reports on Exhibit P1 (EUT1) between 10 June 2015 and 16 June 2015.  Except for the date of these reports, the contents are the same. 

26.  In summary, it says that Exhibit P1 was capable of generating electric arc across the electrodes.  The electrodes of the EUT were connected to resistor load or resistance close to that of a human being.  The peak-to-peak pulsating voltage generated from EUT across resistor load was measured at 114,486 volts.  The EUT was capable of generating 59,055 high voltage pulses continuously in 3 seconds duration.  With the above measurement results, it was said in the report that the electrical characteristic of the EUT was similar to that of a stunning device.

27.  The separation between the two electrodes in Exhibit P1 are 5 millimetres apart.  There is a protective covering of Exhibit P1.  When one opened the protective covering of Exhibit P1 and a button was pressed, Exhibit P1 was capable of generating an electric arc across the electrodes and a sparking sound was heard.  The report says that the evaluation test aims to investigate the electrical characteristic of the EUT output voltage.  There are four measurements in the evaluation test, namely measurement of peak-to-peak pulsating output voltage, pulse duration, pulse repetition interval and the number of pulses generated in 3 seconds.

28.  To observe the output voltage of the EUT, a digital oscilloscope with a high voltage probe was used.  The measurement was done by using the Potential Divider Resistor Network, which is a tailor-made resistor network consisting of ten 200Ω resistors and two 1Ω resistors connected in series.  The total resistance of the Resistor Network is around 2kΩ which is similar to the total body resistance of human being.  To ensure accurate calculations of measurements, the resistance of the Resistor Network is calibrated regularly and the value would be recorded in an evaluation report and the exhibit evaluation report shows the detailed equipment set up for the measurement.

29.  The evaluation test measured the peak-to-peak pulsating output voltage by discharging the EUT through the Potential Divider Resistor Network.  The peak-to-peak voltage across the resistor R_2Ω using the oscilloscope was measured and recorded 10 times in total.  The actual pulsating output voltage of the EUT was obtained by multiplying the measurement results with the resistor ratio of the Potential Divider Resistor Network for each measurement. The average pulsating output voltage of the EUT was then obtained by calculation from the results.

30.  The one pulse waveform diagram from the oscilloscope was captured and the diagram is attached as Annex B of the evaluation report.  The pulse duration and pulse repetition interval was measured by discharging the EUT through the Potential Divider Resistor Network.  The pulse waveform was then monitored using the oscilloscope.  The pulse duration and pulse repetition interval from the waveform diagram was then measured and recorded.  The captured output waveform diagram showed the pulse duration and the pulse repetition interval.  The diagrams are attached as Annex C and Annex D in the evaluation report.

31.  The number of pulses generated in 3 seconds, discharging the EUT through the Potential Divider Resistor Network for 3 seconds was also measured.  The waveform of the output voltage across the resistor R_2Ω was monitored using the oscilloscope.  The number of pulses generated was observed directly from the waveform if possible or obtained the result by calculation.  The output waveform diagram was captured and the same was shown in form of a diagram as Annex E in the evaluation report.

32.  The equipment set up is shown at Annex F in the form of a drawing.  It is said that the resistive characteristic of the Potential Divider Resistor Network is measured regularly and in the report the resistor ratio is 899.342 and the peak-to-peak voltage Vpp (R_2Ω) across the resistor R_2Ω was measured for 10 times until the separation, with the separation between each time at least 30 seconds, giving the average value of Vpp (R_2Ω) as 127.3 volts.  The peak-to-peak pulsating output voltage of Exhibit P1, Vpp, was obtained by multiplying the average value of Vpp (R_2Ω) with the resistor ratio 899.342 giving the answer as 114,486 volts.

33.  In the report it said that since the number of pulses generated is too large to be observed directly on the waveform shown, the result was calculated.  The number of pulses generated in three seconds was 59,055 pulses. 

34.  The procedure and methodology adopted for exhibit evaluation reports are all the same. 

35.  Regarding the exhibit evaluation report for Exhibit P2 (Exhibit P26), the report says that the peak-to-peak pulsating voltage generated from Exhibit P2 across the resistor load was measured at 87,377 volts and that it was capable of generating 122,950 high voltage pulses continuously in 3 seconds interval duration.  Similarly, it said that the electrical characteristic of Exhibit P2 was similar to a stunning device. 

36.  The pulse duration of Exhibit P2 is measured at 6.8 nanoseconds and the pulse repetition interval was 24.4 microseconds. 

37.  As for the exhibit evaluation report for Exhibit P12 (Exhibit P27), although the same was capable of generating electric arc across the electrodes, nevertheless when the electrodes were connected to the resistor load of resistance close to human body.  As in the previous equipment set up for testing Exhibits P1 and P2, the peak-to-peak pulsating voltage generated from Exhibit P12 across the resistor load was measured at

0 volt. It would not generate high voltage pulse continuously in 3 seconds interval duration.

38. Likewise, the exhibit evaluation report for the other 24 EUTs

(Exhibit P28) there were similar test results as Exhibit P12.

PW3 - Professor Carmen Poon

39.  Professor Poon is an assistant professor in the Department of Surgery, Chinese University of Hong Kong.  She graduated from the Engineering Science Biomedical Programme and obtained her Master’s degree from the collaborative programme at University of Toronto.  She completed her PhD in electronic engineering in the Chinese University of Hong Kong and she cofounded the Division of Biomedical Engineering Research at the Department of Surgery.  She is a senior member of the IEEE and has served as Administrative Committee Member and Chair of the Technical Committee of the Wearable Sensors and Systems for the IEEE, Engineering in Medicine and Biology Society.

40.  She is currently an editorial member of three international journals. She is also the recipient of Early Career Award from two of the world’s largest international professional societies of biomedical engineers, the International Federation of Medical and Biological Engineering/ the International Academy of Medical and Biological Engineering, as well as the IEEE Engineering in Medicine and Biology Society.

41.  She contributed to establish the Biomedical Engineering Program at Chinese University of Hong Kong and developed and taught courses on medical devices, undergraduate and graduate levels.  The course syllabus covers the kind of medical electronic devices that measure or apply electrical current on human body, for example, electric knives that dissect tissues in a surgical procedure and devices that provide neuromuscular electrical stimulations.  These devices share some common principles with Conducted Energy Devices (“CED”).

42.  She has also led research studies as principal investigator on blood pressure measurement devices.  Sudden drop in blood pressure (hypotension) is a known reason for syncope, a temporary loss of consciousness usually related to insufficient blood flow to the brain.  The underlying mechanism of syncope is related to the physiological responses produced by CED on human subjects.

43.  Professor Poon has made three statements prior to the trial, dated 25 June 2018, 28 November 2018, and 25 April 2019.  She further made another statement dated 29 July 2019 after the defence expert on electrical engineering raised new issues in court.  She said based on the exhibit evaluation report on Exhibit P1, the peak-to-peak pulsated voltage generated from Exhibit P1, measured at 114.5 kilowatts. 

44.  According to Ohm’s law, peak current on pulses when applied to resistance of 2,000Ω can be estimated to be 14.5 kilowatt divided by 2,000Ω equals 57.2 amp.

45.  Exhibit P1 was capable of generating 59,055 high voltage pulses continuously in 3 seconds, each pulse in the form of a sinusoidal wave and with an average duration of 7.6 nanoseconds.  Based on the above set out measurements, she approximated that Exhibit P1 is capable of delivering a root mean square current of 3 amp.  And as for Exhibit P2, the peak-to-peak pulsating voltage generated from Exhibit P2 was measured at 87.4 kilowatt.

46.  According to Ohm’s law, peak current on pulses when applied to resistance of 2,000 Ω can be estimated at 87.4 kilowatts, divided by 2,000 Ω equals 43.7 amp.  Exhibit P2 was capable of generating 122,950 high voltage pulses continuously in 3 seconds, each pulse in the form of a sinusoidal wave and with an average duration of 6.89 nanoseconds.  Based on the above set out measurements, she approximated Exhibit P2 as being capable of delivering a root mean square current of 4 amp.

47.  CEDs such as Tasers and stun guns, she said, are high voltage low current stimulators and can cause involuntary muscle contractions and pain. CEDs are designed to produce electrical stimuli in the form of short duration (small fraction of a millisecond) repetitive pulses (5 to 30 pulses per second) each of 50 kilowatt.

48.  In brief, the high voltage (analogous to pressure in a water hose) CED is required in order to create an electric arc that bridges poor conductors such as air gaps, clothing or dry skin (the corneum).  With the high voltage design, even when electrodes are embedded and do not touch the skin directly or when the skin is dry, the device can create an arc between the probe and the skin as well as break down the dryness of the skin to enable the electric current to enter the body.  On the other hand, she says, the electrical current, analogous to water flow in a hose, of CED determines the level of painfulness or injuries a device can bring to a person.  The higher the current the more potential injury it brings.

49.  There are several potential mechanisms leading to stunning and disabling effects of CED.  These include, she said, altering the cardiac rhythm by stimulating the cardiac cells causing imbalance of acid-base in the body by altering respiratory and metabolic patterns by stimulating nerves and muscles that control breathing causing transient paralysis and disability by stimulating the motor neurons that control major muscles inducing stress reactions in the human body and leading to excited delirium through pain.

50.  She said delivering a certain amount of charge above a threshold, either as a single pulse or a sequence of pules over a short period of time can produce one or more of these effects.  With external stimulus that accumulates charges over the threshold in less than one second the heart can be triggered to pump a different rhythm.  This can lead to disruptive regular blood flow patterns and dizziness.  During extreme heat, it can even result in reticular fibrillation and death.  Depending on the location where the CED is applied, one or more of those phenomenon can be triggered.  Some of these phenomena are inter-related and they can work together to lead to ultimate stunning and disabled effect.

51.  The current or delivered charge is critical for determining the effect of CED.  Human subject reported pain on average when brief high voltage pulses at 0.5 µH were delivered to the forearms.  Most subjects found that pain is intolerable at 1 µH.  When a commonly used stun gun by the armed forces, namely Taser X26, is applied to the human volunteer, a charge of 100 µH per pulse was recorded. 

52.  The amount of charge is comparable to a defibrillator which is to apply transthoracic electrical impulses for a cardiac capture for resuscitation. Since the safety of CED remains a topic of debate there is limited amount of human research studies conducted under controlled experiments.  She says most of the knowledge on the effects of CEDs is based on Taser X26 which delivers currents in the form of 15 to 19 pulses each and the duration about 150 microseconds.  The average current is approximated to be 2 amp.

53.  She says in a recent study that the probe spread is also a determining factor for the incapacitation effect of CED and severe incapacitation is achieved when probe distance is separated by at least 12 inches.  Nevertheless, the same study presents a scenario where the Taser X26 was applied in the lateral upper extremity capture during back exposure. 

54.  Level of incapacitation did not change much for the probe separation distance varying from 4 inches to 12 inches.  This suggests, therefore, that, she says, that the stunning or disabling effect of CED highly depends on whether the device is actually applied to a person, rather than just probe separation distance.

55.  Since the body, she says, is a heterogenous model, one cannot assume current travels only along the shortest physical distance between the two electrodes of a CED.  Once the voltage is high enough to allow current to enter the human body, there are chances that the current spreads and stimulates the nervous system inside the body.  Furthermore, this study was designed to test only the incapacitation effect of Taser X26 under probe operation mode and with probe separation greater than or equal to 4 inches.  It did not address the incapacitation effect of CED in the drive-stun mode or with electrode separation below 4 inches.

56.  She says both of these exhibits were estimated to operate with a current in the region of 3 to 4 amp range.  This operation range is 1.5 to 2 times more than that of Taser X26 and comparable to that of electronic devices for generating painful sensation including that of perception of shock on human subjects and pain studies.  Both of these exhibits, P1 and P8, require less than 1 millisecond to generate a total charge of 1 µH, a threshold that most subjects found pain is intolerable.

57.  The electrode separation in these two exhibits were both 5 millimetres which is similar to some of the neuromuscular implants, bearing in mind also that some of the electrical stimulators for vagal nerves and baroreceptors in the carotid region only required the two electrodes to be separated by a millimetre range. 

58.  Given the specification of both Exhibits P1 and P2, if either of them were applied to sensitive areas such as chest cavity and spine, both of them could alter heart rhythms, cause severe pain, disrupt blood flow to the brain and stimulate motor neurons resulting in a stunning or disabling effect on the subject.

59.  The high power of both devices can also damage tissues and cause burn marks around the local area where the electrodes were applied.  She, therefore, concluded that both Exhibits P1 and P2 are considered portable devices which are designed and adapted to stun and disable a person by administering an electric shock applied with or without direct contact with that person as stipulated in section 2 of the Firearms and Ammunition Ordinance.

60.  In reply to the report by the electrical engineer expert from the defence, Mr K W Cheng, dated 5 December 2018, Professor Poon disagrees with what Mr Cheng said under section 5.1 and 5.2 of his report on human impedance. She says, unlike what Mr Cheng says, electric excitation on cells have been studied for decades.  She quoted pioneer Nobel Prize work, the Hodgkin–Huxley Model, by doctors Alan Lloyd Hodgkin and Andrew Fielding Huxley.

61.  She says that, regarding figure 5.2 in Cheng’s report, she says human cells, including neurons and cardiomyocytes can be described by this model. Studies of high voltage electric excitation on human body are few because they are generally thought to be dangerous and would cause pain on human subjects. These studies are only ethically justified if the protocol is carefully designed and brings new knowledge to the field. 

62.  Figure 5.2 is the model commonly used in studying electrical shock a person may potentially face at home or at work.  To complete this circuit in this situation, it is always assumed the current enters the body from one limb and exits from the other.

63.  In the case of EUTs, the circuit loop and the resulting impedance/current highly depends on the specific body location where the EUTs were applied.  It is highly likely that there are pathways inside the human body when the resistances are smaller than the skin impedance.  The current would travel along these pathways rather than just on the skin.  Therefore, she does not agree to the comment that where “…the electrode (separation) is around 0.5 cm, therefore the impedance is only skin impedance.”

64.  She says Figures 5.3 and 5.4 of the same report by Mr Cheng are selected from quoted reference, “Effects of current on human beings and livestock.” Both references clearly showed that body impedance drastically reduced with voltage.  Figure 5.4 further shows that at 250 volts, the skin breaks down. This is in agreement with common understanding that skin breakdown voltage is below 500 volts and voltage above this threshold will cause skin to turn from an insulator into a conductor.  Since the EUTs in this case have peak-to-peak voltage in the region of 80,000 to 110,000 volts range, no doubt the EUTs can cause skin to breakdown and the current entering the human body to excite the cells inside.

65.  She says the absolute values of “Total body impedance” in Figure 5.3 and 5.4 are not a good reference for modelling the human body in our case.  She says these figures discuss impedance from the hand-to- hand and hand-to-feet pathways, used mainly for electric safety at home and at work and this should not be confused with our present situation.  Since the electrode separation is small (in the “mm” and “cm” range), both EUTs can be conveniently applied to specific targeted areas of the subject.  The human body impedance for small electrode separation can be much lower than that shown in the Figures 5.3 and 5.4.

66.  For comparison, electrical surgical knives use a load of 200Ω and 800Ω to report their specification by bipolar and monopolar electrodes respectively. The Canadian report used peak voltage divided by resistor load of 600Ω to estimate peak current from over 200 conducted energy devices (Taser X26). A stimulator in the carotid region also uses 600Ω to estimate the peak current from peak voltage.  If 600Ω were used instead of 2,000Ω as the model in this case, the estimated current would be at least three times larger.  Therefore, to conclude, the 2,000Ω model used in her report is already considered a safe margin to account for possible measurement and modelling errors.

67.  In relation to the high frequency human impedance, she says that the Figure 5.4 shows that the total body impedance drops with frequency and it is below the 1,000Ω for frequencies above 5,000 hertz.  For nanosecond pulses delivered in the EUTs, frequencies are in the megahertz range, many orders of magnitude above 5,000 hertz.  If 1,000Ω is used as the model instead of 2,000Ω, the estimated current would be doubled. 

68.  She also says that when the defendant expert claimed that the only real current measurements that has been reported in the literature can be referred to in reference [2][1], she has a number of concerns in the quoted study [2] for reasons:

(a) the applied stimulation was a 10 V step voltage, ie a low DC voltage that has no common specifications compared to EUTs in this case;

(b) no discussion was given regarding where the stimulation was applied;

(c) no discussion was given regarding the repeatability of the measurement and the detailed study design;

(d) no discussion was given regarding the number of subjects recruited for the study; and

(e) no discussion was given regarding whether ethics approval or informed consent has been obtained before the study.

69.  The study [2] is, therefore, not applicable and irrelevant to the question whether the EUTs in this case should be considered as “portable device which is designed or adapted to stun or disable a person by means of electric shock applied with or without direct contact with that person.”

70.  Although, human impedance is not a pure resistance, it is a relevant and appropriate model to decide whether human cells can be excited at a specific current threshold.  The value of 2,000Ω is already a conservative estimate of the human body in evaluating the EUTs.

71.  She said the sine wave assumption gives a 30% reduction in the current estimated from the peak-to-peak voltage.  Therefore, it is found at the present moment the most suitable model for estimating current of the device was as stated in the conclusions she made in the previous statement dated 25 June 2018. 

72.  As for Mr Cheng’s comment that, “Taser X26 is a very high-powered device, whereas the EUT1 and EUT8 are very small power.” Professor Poon does not agree.  She says that Professor Cheng did not take into account the fact that both EUTs in this case could produce trains of pulses in a very short period of time.

73.  The frequencies of pulses of both EUTS are much higher than those of Taser X26 which is in the millisecond and second range.  As a result, both EUTs were estimated to have higher current.  Power is defined by voltage x current, or voltage² / resistance.  When both the voltage and current of the EUTs are comparable or greater than those of the Taser X26, the power of the EUTs are obviously comparable or greater than Taser X26.

74.  Turning to the comment to her report in Mr Cheng’s report, section 6.2, Professor Poon says that temporal summation of electrical pulse trains is commonly used model in excitation of neuron cells.  Although the time range varies from milliseconds to seconds for different types of nerve fibres and cells, these ranges have been considered in her reports dated 25 June and 28 November 2018.  Studies in microbiology use nanosecond and electrical pulses to increase the permeability of the cell membrane indicating that it is possible to open up the impermeable cell membrane to alternate ion influx or out flux using nanosecond pulses.

75.  She does not agree with the comment that one should use the net charge to determine pain sensation.  She says it is well known that human being is very sensitive to AC current and in some cases, a higher threshold is required for a human to perceive DC current compared to AC current.  Both DC and AC current can cause electrical shock. 

76.  Moreover, she says that the defendant expert calculated the net charge of a single pulse on both EUTs and concluded it is very small.  She does not agree with using this sort of calculation to benchmark the EUTs with Taser.

77.  For Taser X26, each pulse has around 100 to 150 microsecond duration but any two pulses are separated by around 50 to 60 milliseconds.  It is, therefore, reasonable to calculate charge per pulse for Taser X26.  On the other hand, to operate the EUTs in a real life operation, a single simple click of the button on both EUTs is able to generate a train of pulses (rather than a single pulse) because human operator would not be able to control the device in a nanosecond time frame.  Therefore, it is reasonable to consider the temporal summation effect of the train of pulses as in her report dated 28 November 2018.

78.  Again, she reiterated, given the heterogeneity of the human body, infinite number of possible designs of CED and the potential injury a CED can bring to a human subject, reasonable assumptions has to be made to approximate the current and to compare a EUT with other CEDs and other medical devices. Voltage and current can be divided into AC and DC components.

79.  In her report dated 25 June and 28 November 2018, she has made the following assumptions as to the estimated current of the EUTs:

(1) RMS_DC is negligible;

(2) waveform is close to a sine wave;

(3) resistance value is relatively high (2,000 Ω).

80.  These assumptions have reduced the possible current by a factor of at least two to three times.  Therefore, it is reasonable, she said, to conclude that both EUT1 and EUT8, ie Exhibit P1 and P2, should be considered as portable devices which are designed or adapted to stun or disable a person by means of electric shock applied either with or without direct contact with that person.

81.  In response to Mr Cheng’s argument that a skin effect on the electronics components can be a factor affecting the measurement accuracy of the EUTs, and subsequently affect the conclusion whether or not the EUTs are stunning devices. Professor Poon replies that skin effect is the tendency of an alternating electric current (AC) to become distributed within a conductor such that the current density is largest near the surface of the conductor and decreases with greater depths in the conductor.  The electric current flows mainly between the outer surface and a level called the skin depth. The skin effect causes the effective resistance of the conductor to increase and high frequencies where the skin depth is smaller.

82.  For the test set-up, skin effect, she says, is most prominent on the conductive wires.  At higher frequencies in the megahertz range, the resistance of the wires can no longer be considered negligible.  By examining the figures of the circuit box, wires are considerably longer in the following segments:

(1) between the external electrode 1 of the EUT and the first test point (TP1);

(2) between the external electrode 2 of the EUT and the second test point (TP2);

(3) between the TP1 and the probe of the oscilloscope; and

(4) between the TP2 and the ground of the probe.

83.  Skin effect on the wires in all these four segments, she says, would result in underestimating resistance total of the test circuit and/or the measured peak-to-peak voltage between TP1 and TP2.  On the other hand the two 1Ω resistors are tightly connected to each other as well as TP1 and TP2.  There is no evidence that R_2Ω has been underestimated due to skin effect. Therefore, skin effect on the electronics components of test circuits in this case will only reduce the Resistor Ratio and the measured peak-to-peak pulsating output voltage of the EUTs which is to the benefit of the defendant.

84.  As for the parasitic effects of the electronic components used in the test setup, Professor Cheng is saying that there are parasitic impedance in the resistors R1-12, as well as in the measurement probe and the oscilloscope. Professor Poon agreed with him in this regard.  She further points out that since the arrangement and the exact values of these parasitic impedences are unknown, to postulate the effects and draw meaningful conclusions are difficult.  For example, to consider parasitic impedences of a single resistor, one has to consider its resistance, inductance and capacitance.

85.  These components can either be connected in series, in parallel or partly in series and partly in parallel.  With different postulated arrangement, the resulting frequency responses would be different.  If these factors were to be considered, the measured value might not be directly comparable to other existing CEDs which normally use only a single resistance value to report their specifications.  Moreover, since the output voltage and frequency of an exhibit is unknown before the test, it is impossible and also unfair to tune the test circuit for each individual EUT.

86.  As for Mr Cheng’s argument that due to the parasitic capacitance and inductance of the R_2Ω resistor, the test circuit forms a harmonic resonance oscillator for current and that even though the two EUTs in this case have very unique frequency, 123 megahertz and 147 megahertz, this could be due to the test circuit having multiple resonance peaks that amplify the two EUTs and the respective frequencies.

87.  Professor Poon noted that the test circuit is a relatively simple circuit, where across the measurement points TP1 and TP2, there can only be one effective RLC circuit.  Therefore, there is no evidence, she says, that it can oscillate at multiple resonance frequencies and generate pulsating voltage of different frequencies during the separate testing of EUT1 and EUT8. 

88.  Mr Cheng argued that the measurements could be as a result of noise being picked up by the test set up, rather than from the EUTs.  Different noise sources have been proposed, including:

(a) power lines;

(b) existence of motors or vibrating objects operating in the vicinity of the test site; and

(c) wireless electromagnetic interference.

89.  Professor Poon responded that since the measured signals were not in the frequency range of the standard electrical voltage in Hong Kong, ie 50 hertz and the tests were conducted in a quiet and isolated room, the first two possible noise sources can be excluded.  Further, the measured frequency of the EUT1, which is 147 megahertz, was in the frequency band of 146 to 148 megahertz, which is the frequency band allocated for land mobile and reserved for government use only.  The measured frequency of the EUT8, which is 132 megahertz, falls in the frequency band of 117.975 to 137 megahertz, which is the frequency band allocated for aeronautical mobile and 93 per cent of its frequency channels are vacant.

90.  She said that the police have already reduced the possible effect of electromagnetic interference during testing.  She further opines that as measured in a report dated 25 April 2019, given the heterogeneity of the human body, infinite number of possible design of CED and the potential injury a CED can bring to a human, reasonable assumptions are necessary to approximate the current and to compare EUTs with other common CEDs and/or relevant medical devices.

91.  She mentioned after considering the exhibit evaluation reports of the two exhibits, she confirmed and reiterated her conclusion in the previous reports that these two exhibits, P1 and P2, should be considered as portable devices which are designed and adapted to stun or disable a person by means of electric shock applied either with or without direct contact with that person.

92.  Professor Poon further is of the view that the view taken by Dr Ng, PW4, and Dr Chiao, PW5, on the issue of electrode separation that the subject devices could not produce sufficient energy was unsubstantiated.  She criticised them for not mentioning whether these devices could cause fainting. She added that it is practically possible to apply the electrodes to vital organs of the victim.  It is practically possible for the assailant to apply the electrodes to vital organs of the victim such as the neck for a prolonged period of time.

93.  Further, Professor Poon noted that these doctors relied on the study of taser guns on human subjects.  These studies suggested a wider electrode separation for effective stunning.  Professor Poon addressed on the limitations of the study report such as that the studies have not covered a probe separation of less than 4 inches, that these studies concerned lean on the disability of the test subjects but not on the stunning effect which are the issues in this case.  The test subjects have been trained and have heavy body build.  The study avoids dangerous locations such as midline or cross-midline of the test subjects, ie, the probes remain on the one side of the test subjects.  The study also avoids testing on locations such as the chest and the neck.

94.  PW3, Professor Poon, says that both the parasitic inductance and capacitance is negligible in the present case as the circuit is a simple one and the wires are short.  PW3, noting that PW7 had considered parasitic inductance, added that PW7 had not considered capacitance and other arrangements, that is how different electronic components in the circuit was arranged and she concluded that PW7 had leaned particularly to one side in terms of liability of the circuit in the calculation.

95.  She said that the final result could go up or down depending on whether the circuit was set up in series or in parallel and she claimed in her report, Exhibit P33, that, “The test circuit is a relatively simple circuit where cross measurements points of TP1 and TP2, there can only be one effective RLC and there is no evidence that it can oscillate at multiple resonant frequencies and general pulsating voltages of different frequencies during the separate testings of EUT1 and EUT8.” 

96.  She further stated that had there been resonance, she would have seen it on the oscilloscope and that the wave form grew bigger and bigger.  She was not concerned about the noise challenge by Professor Cheng as said previously. Apart from the noise, she also stated that the test set-up was enclosed in an aluminium box of non-negligible thickness.  As indicated, the precautions have been taken by the police to reduce the possible effects of EMI during testing.

PW4 - Dr Ng Chung Ki

97.  Dr Ng is a Senior Medical and Health Officer and a Senior Forensic Pathologist. He had prepared a report dated 10 July 2015, Exhibit P20, regarding the effect of Exhibit P1 on humans.  He was tendered for cross-examination at the request of the defendant. 

98.  PW4 confirmed that there was a longstanding practice for the police to hand over exhibit evaluation reports in relation to stunning devices to them. PW4 said it is his first time to testify in court concerning a suspected stunning device.  He admitted he is not an expert in electrical engineering.

99.  He could not be sure whether Exhibit P1 is capable to stun or disable a person but he concluded from the materials that he had that there is a possibility that it could do so. 

100.  He explained that the peak-to-peak voltage is very high because the amplitude of the peak is very high, but as for current, because it is a long period of time between each of the pulses produced by Exhibit P1 and although Exhibit P1 voltage is high, the total current that is passed through is, in fact, low.  He says that as the electric arc lighter, Exhibit P1,  was not designed as a stun gun with a U-Guard and a protective cover to avoid the electrodes from contacting another person, he is of the view that the likelihood of the Exhibit P1 being able to stun or disable a person is very low.

101.  He also stated that there are many papers that stipulate the greater the probe spread, the greater the chance of causing incapacitation and confirmed that in the smaller probe spread, the possibility of causing incapacitation becomes less likely and agreed that a probe spread of 5 millimetre is a very short distance and that he has never read any paper on experiments on devices with a five millimetre probe spread.  He says he could not, therefore, conclude whether, in fact, definitely whether Exhibit P1 could stun a person.

PW5- Dr Chiao Wing Fu

102.  Likewise, Dr Chiao is a Senior Medical and Health Officer and a Forensic Pathologist in Hong Kong.  Again, he had prepared a report dated 22 May 2017 concerning the six electric arc lighters, not any of the seized exhibits in this case, and their effect on human, Exhibit 21.  Again, he was tendered for cross-examination by the defence. 

103.  He says that the area of electrical separation distance and current flow in the human body is an area between biomedical engineering and forensic pathology.  He explained that while current is an area in medical engineering, effect on human he believed would be in the area of forensic pathology, and as an expert, he has the expertise. 

104.  PW5 was the author of the PowerPoint issued by Medical and Health Department, Exhibit D2.

105.  PW5 maintained that stun guns which deliver electrical pulses by contact of the skin only are unlikely to incapacitate a person.  They can only induce severe pain and not even cause dizziness.  He is of the opinion to compare the drive-stun mode of Taser X26 with the electric arc lighters with a 5 millimetre electrode spread is most appropriate.  He agreed that the probe mode of Taser X26 is completely different from electric arc lighters and is not comparable. He is of the opinion that without the dart probe, it would be very difficult for current to penetrate the skin.  He said that the design of the subject electric arc lighter in this case has the protective cover and their electrodes are protected by the U-Guard.  In addition, the gap between the electrodes are so small that the chance that one could easily apply current on another is very low.  He says if one grabs a person by the neck, and attempts to apply the electric arc lighter on his aorta, PW5 opines that it would be only a contact with the skin and the current would not penetrate any tissue. 

106.  Regarding Exhibit D6, PW5 concurs with the author’s illustration observation that the skin at the site of the electrode immediately following the discharge of the drive-stun conducted electrical weapon (CEW) displays a widened U-shape indentation in the epidermis. 

107.  There is no full thickness breach of the epidermis, but there is a loss of stratum corneum at the electrode contact site.  The dermis lacks hemorrhage, thrombosis, inflammation, or other collagen change. 

108.  He had stated, comparing to the drive-stun mode of the Taser X26, it would be even harder for the electric arc lighters to breach the epidermis. PW5 definitely concluded that in his opinion that the electric arc lighters in his case could not stun or disable a person.

PW6 - Tang Kwok Ming

109.  Mr Tang, Assistant Police Telecommunication Inspector, was the one responsible for actually performing tests on and preparing the exhibit evaluation reports of a suspected stunning device in this case. 

110.  In cross-examination in court about possible problems with the resistors in the test circuit, PW6 testified that had there been any problem with the resistors, for instance, when the current was too high, the resistors would have burnt and he would have noticed a burning smell.  In these circumstances, oscilloscopes would not have been able to show any measurement results. 

111.  PW6 said that he had not noticed any burning smell and was able to carry on all the 10 measurements with each EUT.  PW6 stated that he did check the resistors before the tests commenced but he would not check the resistors each time after a measurement had been taken since the resistors are inside a resistors box with a cover.

112.  PW6 also said that there was a trimmer switch on the probe to calibrate and optimise the square wave.  He testified that he did not use the trimmer at all to tune or adjust the waveforms to optimise the square wave.  He assumed that the probe was calibrated wherever the oscilloscopes were and that the trimmer would have been adjusted then.  PW6 also testified that the oscilloscope had been set to a capture mode, also known as the trigger mode. Once he placed the electrodes of the EUT at the two contact points marked, “To external electrode 1 of the EUT,” and, “To external electrode 2 of the EUT,” as shown in the exhibit evaluation reports, then if the oscilloscope could capture any voltage, an image would be shown on the screen automatically.

113.  Under cross-examination, as pointed out to PW6, the highest and lowest peak-to-peak pulsating output voltage for Exhibit P2 was 113 and 64 respectively and the extremity was significant.  He explained that he was merely responsible for recording the facts.  He followed procedure and took 10 measurement pulse samples, then calculated the average voltage output and recorded it down.  It was put to PW6 that other than the trigger mode, there is also another mode that of the oscilloscope being the manual mode so as to avoid the oscilloscope only automatically capturing pulses with high voltage.  PW6 confirmed that there is a manual mode but he did not use it.

PW7 – Chan Tat Kwong James

114.  Mr Chan has a Bachelor of Engineering Degree in Electrical Engineering and Master of Science in Management.  In years 2002 to 2004, he was the laboratory manager of the COMMS laboratory, Communications Branch, Information System Wing of  Hong Kong Police, which was recognised by the Hong Kong Laboratory Accreditation Scheme.  He now handles huge electronic system projects in Hong Kong Police. 

115.  He says that in his witness statement date 29 April 2019 (Exhibit P19) that the aims of the evaluation tests are to investigate electrical characteristics of the EUT output voltages in the measurement readings of peak-to-peak pulsating output voltages, pulse duration, pulse repetition intervals and numbers of pulses generated at three seconds intervals were recorded.  In order to avoid unstable equipment readings and eliminate unreliable measurement results, the measurement data was averaged for precise results.

116.  He says that based on the summarised observation and obtained readings, they concluded that both Exhibit P1 and P2 were capable of generating electric arc electrodes.  The peak-to-peak pulsating voltage generated was respectively 114,486 volts and 87,377 volts and they were capable of generating 59,055 and 122,950 high voltage pulses continuously respectively in three seconds interval duration.  He said with the above measurements that these electrical characteristics of the EUTs were similar to a stunning device.  He said that the waveform analysis and the measurement was not included in this evaluation.

117.  In reply to the queries raised by Professor Cheng regarding the sampling rate of the equipment, he relied on Nyquist Theorem that a signal can be measured by two times sample frequency of it.  This measurement, the sample frequency of the oscilloscope, model TDS3032B, using for measuring the output voltage of EUT is 2.5 gigahertz.  That means that the test equipment, oscilloscope, is capable to measure the signal frequency up to 1.25 gigahertz in theory.  In this measurement the sampling rate is higher than the tested signal frequency about 17 times.  It is over 8 times of the minimum sampling requirement.  Therefore, the test equipment is capable to measure the frequency of the EUTs.

118.  Therefore, Professor Cheng’s query about the equipment accuracy, PW7 said in his report that based on the record the test equipment, oscilloscope model TDS3032B, was sent to Hong Kong Calibration Laboratory for calibration on 14 August 2014 which was about one year before the dates of the two EUT measurements.  Therefore, one should not be concerned about the equipment accuracy.

119.  Referring to the query about the resistor for testing in the measurement, he says that the 2,000Ω resistance (resistor load) for simulating the resistance of human skin were formed by 10 numbers of

200Ω carbon film resistor in series connection and 2 numbers of

10Ω carbon film resistor in series connection. These resistors were brought from an electronic components shop. There is no specification of this resistor. However, the measurement record of the resistor load from

10 July 2014 to 25 November 2015 was listed in the table enclosed in his report, which clearly showed the range of total resistor value from 1,978 to 2,268Ω. It was indicated that the resistance value of the resistance divider box (resistor load) is quite stable without any changes, even occasionally loading the high voltage pulsating signal. The resistor load was solely used for measuring suspected stunning device.

120.  He then went on to respond to the query about the time constant of resistor circuit and said that the resistors were soldered one by one in series connection and formed a about 2,000 Ω value resistor load by a qualified technician of Hong Kong Police.  Professor Cheng said the measured waveforms are not accurate due to the LR circuit which may introduce up to five nanoseconds time constant.  However, he says this testing model for EUT evaluation is not able to obtain the exact waveform of the high voltage signal without distortion.  He said the report of the forensic pathologist also did not highlight or emphasize the waveform of the EUT output signal.

121.  PW7, in further reply to the issues challenge put forward by Professor Cheng during trial, further prepared another report dated 29 July 2019 in reply.  He said the resistance box in question had tested frequencies of different stun devices covering a wide range of frequency from 7 megahertz to 200 megahertz.  And there seems to be no obvious resonant frequency observed.  He believes that the waveforms shown in the oscilloscopes were not caused by the resonance of the resistor network but the generated signal from the EUT.

122.  He further added that a physical characteristic of carbon film resistor is stable.  He also agreed that the heat dissipation is lower and the thermal effect is also very small.  Therefore, the resistance values of resistors, R1 to R12, should not change due to the high but narrow voltage pulse of the EUTs.  The inference is in line with the result of the laboratory regular resistor network measurements.  He says that the results of measurements showed the resistance characteristic of the resistors is stable.  Therefore, the skin effect in the resistor network he wrote in his further statement that according to the formula of the skin effect, for high conductivity conductor, the skin effect can be obvious.  However, for poor conductivity material, the skin effect shall be smaller.  In addition, the resistor carbon film is a thin graphite layer, which is already a thin skin.  Therefore, it is reasonable to assume the skin effect in carbon film resistor is quite small and he said the skin effect would likely appear in the copper wires of the circuit.

123.  Referring to the resistor box (Photo 1) in his statement and the schematic diagram of the resistance box, Figure 1 in his report, the four copper wires that might induce skin effect resistance R(f1), R(f2), R(f3) and R(f4).  It is noticed that both R(f3) and R(f4) resistors were not the components of the circuit of the voltage divider and it could be considered that they become impedance of the probe or the oscilloscope.  Comparing to the impedance of the probe and the oscilloscope, above total 51M ohm, the value in R(f3) and R(f4) would be very small would not cause significant change in the waveform.  In the measured voltage across resistors, R1 and R2 would be conservative and lower than that of the actual voltage since voltage drop in R(f3) and R(f4) before reaching the probe and the oscilloscope.

124.  Coming to the issue of parasitic inductance in the Circuit, Mr Chan says that in his submitted report, the measurement of EUT1 and EUT8 were based on pure resistance model of voltage divider circuit.  In response to Professor Cheng’s query about the parasitic inductance in the circuit, further calculation including reasonable assumed values of parasitic inductance into the circuit was conducted.  In making the assumption the parasitic impedance of any single resistors, R1 to R12, is 2.5 nH in 140 megahertz, the copper wire resistance for R(f1) plus R(f2) caused by the stun effect under 140 megahertz is 140Ω.

125.  Therefore, he says if considered the effect of parasitic inductance into the circuit and accepting the above assumption, the measured values of output voltage of EUT1 and EUT8 shall be revised for forensic to re-evaluate the EUTs.  At the end of this estimation, with considering parasitic impedance in R1 to R12, he worked out the output voltage of EUT1 and EUT8, that is Exhibits P1 and P2, would become 38.75 kilowatt and 56.46 kilowatt respectively.  The pulse number of both EUTs in the three seconds remained the same as submitted before.

126.  In court, PW7 says that he implemented the third generation communication and power system to the police and the communication system in the marine digital radar and video transfer.  He also served as lab manager responsible for equipment calibration for various departments and the lab was accredited until year 2004. 

127.  As far as the set-up of the test circuit was concerned, he said that this was first designed by his senior colleague in collaboration with the pathologist and later the tests included also the three seconds functionality test. The test procedure was used to test the suspected stun device and would be sent to Telecommunication Department.

128.  In relation to the parasitic inductance/capacitance, PW7 said this exists in all components and that the parasitic inductance and capacitance might compensate each other negating the overall effect.  As for the skin effect, he said with reference to other telecommunication system that he had worked with, he considered the effect was so minimal and insignificant to be considered.  It is also impractical and reasonable to assume that any such skin effect would convert unevenly to affect the result.  He further commented that in order to produce oscillation on a certain frequency, that would require careful design.  It would hardly occur by chance.

129.  In court, he stated that after checking with the certificates of calibration, the oscilloscope model TDS3032B used to measure Exhibit P2 and the other 25 suspected stunning devices were calibrated on 14 August 2014.  When the devices were tested on 28 August 2015, it was only two weeks after the 12 period from the last calibration.  The same was recalibrated again with certificate as Exhibit P29(4).  PW7 claimed that by comparing the error data on the certificate of calibration one could confirm the amount of deviation in the measurement on the oscilloscope.  As for the oscilloscope used to test the voltage output of Exhibit P1, model TDS 3034B, PW7 said that the same was calibrated once every 18 months.  It was calibrated 16 December 2013 and again on 5 to 6 August 2015.  Exhibit P1 was tested on 6 June 2015 when the same was supposed to be recalibrated.  He says he confirmed that he did not have specification from manufacturer as far as the resistors used in the test circuit is concerned or documentation to show that the resistors are able to perform under high voltage situations.

130.  He said, however, that the resistance value was measured by a resistance meter, almost known as multi-meter, with a voltage around 3 to 9 volts.  He agreed that by so measuring does not mean that the resistors were capable of functioning properly under high voltage situations of 8 kilowatt or above.  He also agreed that the frequency of the multi-meter is completely different from that of the suspected stunning devices and he agreed that the resistance value of the resistors would be affected by high voltage and high frequency, but the change would be insignificant.

131.  As for the accuracy of the probe, PW7 admitted the same had not been calibrated.  He said one could use the trimmer to tune the waveform and optimise the square wave to ensure that the probe was functioning properly when used on the oscilloscope.  He had mistakenly believed that PW6 had done the same.  He admitted that he had no knowledge of the age of the probe or from where the same was purchased and he never inspected the probe for ageing.  As for the other stunning devices which were able to produce electric arc but gave the final zero volt measurement, PW7 said that is a fact and he denied any inconsistency.  He did not find it necessary to find a suitable oscilloscope for not measuring any voltage.  PW7 admitted there were steps to analyse why the EUT were able to generate electric arc but without any measured voltage, but he did not perform such test.

132.  PW7 agreed that there are parasitic inductance effects and resistance at high frequency which would make the resistors not just resistive but capacitive and inductive.  He also admitted in cross-examination that parasitic inductance and capacitance would exist at the contact point between the probe and the resistor board, between the probe and the oscilloscope, in the solder used to solder the resistor to the resistor box, between the EUTs and the contact points when testing.  He admitted he did not know the range for parasitic inductance and parasitic capacitance in the circuit.

133.  Although, PW7 claimed the parasitic inductance are negligible, he put in a presumed value for parasitic inductance in his recalculation which resulted in a drop of 50 per cent roughly of the peak-to-peak pulsating voltage for Exhibit P1 and P2.  PW7 said he tried to make them as close to or equal to or possibly equal to the values of R1 and R2.  When cross-examined that the value should be a larger figure, PW7 said that he, in fact, did not know and he had no experience in measurement of this kind of resistance.

134.  Coming to the issue of oscillation and LC resonance, PW7 says in his statement, Exhibit P32, that the tested devices are of different types of stun devices and they cover a wide range of frequencies and there is no resonant frequency observed.  He believed that the waveform in the oscilloscope were not caused by the resonance of the resistor network but generated signal from the EUTs.

135.  In relation to the possible EMI noise affecting the test result, he said that the tests were performed in a room and, although, there was no strict protocol in place to avoid EMI when tests were conducted and there was no measurement to confirm that there was any induced voltage in EMI in the resistor box.

Defence case

136.  The defendant in this case elected not to give evidence.

DW1 - Professor Cheng

137.  Professor K W Cheng is an expert witness in electrical engineering.  On 5 December 2018, he prepared a report in this case including commenting to the exhibit evaluation report on Exhibit P1 and Exhibit P2, prepared by Mr Tang (PW6) and Mr Chan(PW7).  He is the professor of the Department of Electrical Engineering of the Hong Kong Polytechnic University.  His research areas are on power electronics, electronic circuit, converters, energy saving and renewable energy.  He has in electrical engineering for more than 30 years.

138.  He claims to be experienced in power circuit research and development including high voltage, high conversion, transformer and bioelectronics.  He has developed more than 100 projects in power electronics and electrical engineering with over 300 publications in those areas.  He has also received awards in relation to power conversion.  Recent to his writing of the said report, he had conducted research on human impedance and associated protection.  He has experience as an expert witness in electrical engineering in court.

139.  In his report he raises doubt on the accuracy of the measurements in the exhibit evaluation reports.  He criticised the use of oscilloscope in the exhibit evaluation report test, ie TDS3032B.  In the case of Exhibit P2 and TDS3034B in the case of Exhibit P1.  He says both equipment are of sampling rate of 2.5 GS/s.  In the former case, the measured waveform is of 6.8 nanoseconds and the latter 7.6 nanoseconds.  He says that oscilloscope uses interpolation to construct the waveform and the accuracy is not confirmed because the sample rate is slow to measure the waveform with small pulse width.  He also says the equipment should be calibrated once every 12 months, otherwise, the accuracy of the measurement is not confirmed.

140.  As for the set-up of the testing equipment, firstly, he said because of the high voltage measurement, the use of 10 times 200Ω resistors and 2 times 1Ω resistors are used to reduce the voltage to low enough for measurement.  There are no descriptions of the accuracy and the voltage ratings of the resistors used.  He estimated that the voltage impressed on the 200Ω resistors is 4.36 kV.  He said further that there are very few resistors in the market that have voltage ratings of over 4 kV.

141.  He says that if the resistor is not of sufficient voltage rating, the measured result is not accurate.  He says most of the resistors in the market is only about 50 to 400V rating. Also most resistors specified at voltage ratings at DC, and there is no information for most resistors in the market specified with high frequency AC, especially if the frequency of the measurement is around 147 megahertz, estimated using 6.8 nanosecond pulse width.  If the voltage rating deviates by many times, the result of the measurement, he says, is not reliable.

142.  He then goes to say that if in the electric current, there is a parasitic effect of the circuit coil inductance.  A short wire of a few centimetres will form inductance in the order of µH, the use of 10 resistors of 200Ω and 2 resistors of 1Ω connected in a series would create a certain amount of series inductance.  There is no information of the resistors types used and also the wiring between the wires.  There is no information of the measured inductance on the set-up.

143.  He says that if the 12 resistors connected in a series and also with the wires, the estimated parasitic inductance can vary from 1 µH to 10 µH.  The time constant T of the set-up is estimated to be between 0.5 nanoseconds to 5 nanoseconds and using this time constant to measure 6.8 nanosecond waveform is not accurate. 

144.  Professor Cheng also commented that as the output of the two EUTs are high voltage and high frequency, there is a chance that impedance drops at high frequency.

145.  Coming to the report by Professor Poon, Professor Cheng says it is uncertain to say human impedance is 2,000Ω and that they behave strangely at high frequency and high voltage.  He says the definition of 2,000Ω does not apply to around 100 megahertz frequency and 100 kilowatt condition and that the response of the current may not be a sinewave.  Further, the rise time of the current under such condition is unknown.

146.  He says further that as the voltage is AC waveform, the charge developed as positive and negative with positive cancelling the negative effect and net charge is very small or approaches zero.  Further, he says parasitic effect on human subject arise time current has not been considered.  He says that the actual charges are much smaller.  As the actual current passing through the human subject is unknown he says it is unfair that Exhibit P1 are comparable devices to other CEDs such as stunning gun or Taser.

Discussion

Issues

147.  There are three issues in this case as described by the defence;

(1) whether testing method and equipment on the seized devices are reliable and whether they yield an accurate result;

(2) whether Exhibits P1 and P2 are portable devices which is designed and adapted to stun or disable a person by means of electric shock applied either with or without direct contact with that person; and

(3) whether D1 and D2 had any knowledge of the nature of the seized devices; namely, whether P1 and P2 are portable devices which are designed and adapted to stun or disable a person by means of electric shock applied either with or without direct contact with that person.

148.  D2 is a man of clear record and due consideration regarding his propensity to commit crime and credibility has been given. 

149.  This court in dealing with the three issues has already carefully considered all submissions by both parties and would not repeat the evidence here because the same has already been set out previously.

Issue 1

150.  This court now deals with the first issue. 

151.  Both Professor Poon and Professor Cheng for the defence are experts in electrical engineering.  Both of them have not taken part in the actual evaluation tests performed on Exhibit P1 and Exhibit P2 and Mr James Chan and Mr Tang were responsible for these tests. 

152.  Mr Chan is an expert.  Mr Chan has served over 20 years and has performed and managed a lot of projects concerning high frequency electronics. He used to be the lab manager for equipment calibration of various departments and the lab was an accredited one by Hong Kong Calibration Lab, the most authoritative body up to the year 2004.

153.  It is noted that at this juncture that the laboratories used by Professor Cheng at all material times was not accredited. 

154.  It is not challenged that the testing procedure had been designed for some time in collaboration with the pathologists and revised to include the three seconds functionality test.  Throughout the years, the same has been employed to test these stun guns or devices.  The criticism on sampling rate has been dealt with by both Mr Chan and Professor Poon based on the Nyquist Theory which this court accepts. 

155.  For the calibration issue of the oscilloscope, firstly, although they were not punctually calibrated every 12 months, the delay is not substantial; secondly, from the calibration results of this oscilloscope following the tests of Exhibit P1 and Exhibit P2 in question, they did not show any irregularity in the error data.  It is also accepted by Professor Cheng that Tektronix is a high quality brand for oscilloscope.

156.  As for the probe, according to the prosecution witnesses, no apparent abnormality was observed in the waveform.  There is, therefore, no need, they say, to use the trimmer to fine tune the probe.  Further, both Mr Chan and Mr Tang did not observe any abnormality throughout the test and in the test results.  There is, therefore, no evidence to suggest that the probe had not been properly performing as it should.

157.  Coming to the resistors, it was submitted by the defence that there is no specification from the manufacturer regarding the resistors employed. Nevertheless, according to Mr Tang there is again no sign of abnormality noticed.  Whilst the tests were performed, there is no smell or signs of burning and sudden loss of readings.  In fact, there had always been 10 measurements during these tests and overall there is no abnormalities observed even in the measurement readings.  All resistors have been measured in the lab by a multi-meter.  Coming to the criticism posed by Professor Cheng, again, there is nothing from the test results and indeed any evidence before this court to suggest that the resistors have not been performing as measured by the multi-meter.

158.  As Professor Cheng admitted there is seldom any factory specification for any resistors based on AC conditions, it is, therefore, impracticable for this to be expected unless, of course, there is evidence leading to suspicion that these resistors have not been performing as they should under the present high voltage test situation.

159.  In the course of Professor Cheng giving evidence in court, he presented further new issues to challenge the accuracy of the test results.  When questioned why he did not present these new issues in his report earlier, he explained that he had only been given the photographs of the test circuit, Exhibit P19B, at the start of the trial.  This court noted that the set-up of the test circuit was in his report in the form of a diagram and though there is an absence of the photo, this court is unable to accept this explanation as valid for the reason that these new issues, namely, frequency induced resistance by reason of parasitic inductance, skin effect, noise and possible oscillation, etc, did not emerge earlier.

160.  Further, in cross-examination, Professor Cheng acknowledged that invariably in any testing or calibration equipment, it brought into the picture those characteristics also.  He also considered that accuracy is only a relative thing and the issues is about whether the test results can be used practically.  For example, it is not in dispute that all parasitic inductance/capacitance existed in all components and they may compensate each other negating the overall effect.  However, he did not in his report criticise failure in the testing procedure adopted by the prosecution in not having these characteristics predetermined.

161.  Mr Chan (PW7) dealt with the new issues in his further report.  In order to demonstrate, he has made new calculations based on certain assumptions on the overall parasitic inductance of the circuit, the resulting voltage was reduced by 50 per cent. 

162.  Professor Poon maintained her criticism of this assumption of overall parasitic inductance used by PW7 in his new calculation saying it is a biased one leaning in favour of an inductive circuit.  She says given such circumstances, her conclusion on the effect of Exhibit P1 and Exhibit P2 on the human body are still valid and she has already provided a buffer in her calculation to cater for any test uncertainty.

163.  Mr Chan did not offer actually any different explanation as to why he selected that particular figure.  Although, the defence in cross-examination of Mr Chan did offer certain figures, there is no basis before this court as to why those figures are offered.  The cross-examination could only result in the admission by Mr Chan that he is not experienced in these sort of high voltage measurement projects.

164.  On the skin effect issue, the court again preferred the expert view of Professor Poon to that of Professor Cheng.  In fact, Mr Chan shares the same view that there is no reason why the skin effect would unevenly converge to cause an impact to the test result and, further, as said by Mr Chan, any effect would be minimal and insignificant. 

165.  In fact, Professor Cheng in cross-examination agrees that there is no special reason why the skin effect would converge only at the

2 Ω resistors alone.

166.  Then coming to the oscillation issue, Mr Chan said there was no oscillation during the test.  They did not observe any resonance.  He said that to produce oscillation on a certain frequency requires careful design and can hardly occur by chance.  He said that the same resistance box was used to test many devices operating from 7 megahertz to 200 megahertz and that the circuit has multiple resonance frequencies across 7 to 200 megahertz is unimaginable.  Professor Poon agrees that for a simple circuit as the one in question, it is impossible to have multiple resonant points as Professor Cheng suggested.

167.  Lastly, coming to the noise issue, Professor Poon said that the set-up of the said circuit has also catered for such risks and again the test results do not suggest noise playing a part in possibly making the results inaccurate. This court prefers and accepts the expert opinion of Professor Poon on the above issues to that of Professor Cheng.  Her analysis is logical and convincing and her evidence is forthcoming, even in cross-examination.  Her evidence is totally unshaken. 

168.  This court is of the view that the report produced by Professor Cheng on the study of wireless power transfer design for the purpose of transfer of high energy did not, as said by the prosecution, assist the court on the noise issue.

169.  Professor Cheng is a very experienced electrical engineering expert and this court doubts why he would have missed these new issues from his report in the first place if these factors could really and significantly affect the test results.  As said, this court does not accept his explanation based on delayed submission of the photo, Exhibit P19B.  Professor Cheng has further raised concepts of proximity effects, second order effect or matrix effect in court near the end of the trial.

170.  Again, this court considers that had all these been real challenges to the accuracy of the test results, he would have already disclosed much earlier these elements or factors.  Whilst he is an expert in electrical engineering, in the context of our case here, the approach of Professor Cheng appears to be pedantic and academic.  This Court is not with him on the issues that he raised. They would not cast doubt on the accuracy of the test results of the exhibit evaluation reports.

171.  In all, on issue 1, in spite of all the issues raised by the defence, the court finds that the testing method and testing equipment used in the tests on Exhibit P1 and Exhibit P2 is reliable and the test results accurate for the purpose of this trial.

Issue 2

172.  Now coming to issue 2, Professor Poon is the only biomedical engineering expert, in my view, most suited and qualified to tell the effect of the device when applied on humans.  There are, of course, live evidence and reports of the two forensic pathologists as well as the reports of Dr Lam before this court.

173.  Professor Poon has concluded and has in cross-examination agreed that the functions of the EUTs in the present case are more similar to generic stun guns and the drive-stun mode of Taser X26, rather than the probe mode of Taser X26.  Defence criticised Professor Poon for not mentioning the drive-stun mode throughout the first three expert reports.  Professor Poon explained that she had already taken the drive-stun mode of Taser X26 into consideration.  She claimed that in all her reports she had all along been discussing what would have happened to have a person when the drive-stun mode is used.

174.  This court considers these criticisms unfounded.  In her reports, Professor Poon has set out how conducted energy devices (CEDs), operate.  She did refer to a list of references involving Taser X26.  In her report she set out the common principles with CEDs and the underlying mechanisms of syncope relating to physiological responses produced by CEDs on the subject. Irrespective of the operation mode of Taser, there are high voltage and low current simulators.  She told of potential mechanisms leading to the stunning and/or the disabling effect of CED which include not just the transient paralysis and disability of stimulating the motor neurons that control the major muscles. This is in paragraph 8 on her report dated 25 June 2018.

175.  It is also said that one or more of the phenomenon can be triggered and they might be inter-related and can work together to lead to the ultimate stunning and disabling effect.  In fact, the mechanism as divulged in her reports works for all CEDs irrespective of whether they are a stun gun, Taser gun in probe or drive-stun mode.  Depending on the electrical characteristic of each device, also depending on the location applied on human, duration of application, etc, the effect on human would be different.

176.  Professor Poon did in her report dated 28 November 2018 (Exhibit P16) point out that certain references she used did not address the incapacitation effect of CED that operates in the drive-stun mode or with the electrode separation below 4 inches.  She was then referring to reference 5 [2](see last sentence in paragraph 8).  As said by Professor Poon, studies of high voltage electric excitation on the human body are few because they are thought to be dangerous and cause pain on human subjects. 

177.  Professor Poon was also criticised to have omitted a certain part of the references material when the same does not suit her.  For example, she failed to include in her report the findings by an author of a reference that, “Electrodes 5 cm applied directly or over the vastus lateralis muscle, a muscle in the thigh, does not inhibit voluntary function of the muscle during stimulation or afterwards.”

178.  Professor Poon in court says that she does disagree that the electrical current resulting from the drive-stun mode is localised to the muscles in that area.  She also disagrees that the drive-stun mode does not cause muscular incapacitation.  She says that the conclusion of the author only referred to disputation of excited delirium, which the defence disagrees.  This conclusion appears on page 22 of the reference to under, “Section 6 - Excited Delirium.” While the author did not qualify the conclusion and recommendations in this section as relating solely to excited delirium, this court does not find Professor Poon’s disagreement to this conclusion unreasonable.

179.  This court takes the view that depending on, again, the electrical characteristic of a suspected stun device, the test results measurements of the CED, the various factors required for consideration would include but not limited to, as said earlier, the location of application, the duration of application, the length of the electrode separation.  There is nothing to prevent, in my view, a CED, even in the drive-stun mode, to cause stunning or incapacitation. 

180.  One must not forget that for the Taser research, the user should have always been using the Taser device in compliance with the manufacturer’s instructions avoiding certain areas of application on human subjects.  Where in the present case this is not a concern for the issues before the court.  In fact, as testified by Professor Poon, a lot of studies were done on Taser X26 and the CEDs were applied on human subjects avoiding certain areas of the body and instructed by the manufacturers.  This research was done as people were concerned of the effects of this Taser and CED on human whether they would result in ventricular fibrillation and/or death.

181.  The defence has not challenged the conclusion of Professor Poon relying on, inter alia, Exhibit D3 of which Mark Kroll and Exhibit D6 of which Mark Kroll was a co-author.  Mr Kroll is an advisor and director of Taser International.  He has failed to disclose his interests and expertise.  Exhibit D3 says that with the lack of skin penetration the current flow is primarily through the dermis and fat layer between the laterals and there is no significant penetration beyond the subdermal (or subcutaneous) fat layer. Since there is insufficient depth of current flow to capture muscles, the drive-stun mode serves only as a compliance technique.  He says, further, that with a controlled electrical weapon applied directly over the human phrenic nerves that control breathing, there is no effect.

182.  Exhibit D3 quoted the National Institute of Justice, a five-year review of CEWs, and found that a risk of ventricular dysrhythmia is exceedingly low in the drive-stun mode of CED because the density of the current on the issue is much lower than in this mode.  In Exhibit D6, it was said that the study was aimed at understanding the attenuating effect and estimated the residual amount of current flowing towards deeper layers of tissue and finding the extent of region where the current density or the electric field strength exceeded threshold known to cause ventricular fibrillation or to cause cardiac capture.

183.  The paper conducted that the fat and the anisotropic skeletal muscle layers significantly attenuated currents delivered by Taser or CED operating on drive-stun mode in only 9 per cent of the total CEWs current flow beyond the mid-point of skeletal muscle layer.  Current density in drive-stun mode did not exceed threshold for VF induction or cardiac capture outside the subdermal fat regions. 

184.  PW3 criticised Exhibit D3 as inconclusive and unreliable because it is not an international journal that has gone through peer review.  It is noted that PW5 agrees with those conclusions in Exhibit D3.

185.  As for Exhibit D6 which has gone through a peer review, PW3 says that the aim of the research aim is finding out whether the drive-stun mode would cause more serious conditions, ie ventricular fibrillation or cardiac capture.  She says that the LEM model is too simple ignoring the existence of blood vessels cells in a simple simulation.  She also says there are many papers which shows CEWs did cause a disruption to the heart rhythm which may cause dizziness.  She refers to a sentence in Exhibit D6 where it says, “The resulting sensation is often sufficiently strong to assist law enforcement in gaining compliance from subjects.”  She says that this sentence supports that the CEW would cause temporary incapacitation.

186.  PW3 is further criticised for comparing Exhibit P1 and Exhibit P2 to the electrical surgical knives and neuromuscular implants to support a view that even CEDs with electrode separation of millimetre separation can stun or disable a person.  PW3 when cross-examined explained that if the CED can break down the skin with high voltage they share the same similarities as the neuromuscular implants, including exciting the cells.  For the surgical knives, PW3 said there are some electrical knives with one or two electrodes and the mechanism behind the electrical surgical knives is slightly similar to Exhibit P1 and P2.  They operate at different high frequencies which would allow her to compare voltage and current output but their waveform or electrical knives are different and their frequencies might not be as high as the present exhibits.

187.  Despite the various thorough submissions by defence counsel to challenge the expert view of Professor Poon, this court, considering what is before me both from Professor Cheng and the other forensic pathologists, accepts the explanation by Professor Poon and as well as her evidence on this issue.  One must not forget that all tests performed on human subjects using Taser are limited and follow the manufacturer’s instructions in avoiding certain areas. Further, this court accepts that, as said by the prosecution witness, that human skin would break under, inter alia, high voltage conditions and as such, depending on the location of application, duration of application, this court does not doubt the conclusion of Professor Poon drawn on the issue.

188.  In fact, none of the authors in these references refer to scenarios involving the penetration of the skin.  Their views all refer to scenarios where there is no penetration of the skin.  And, of course, penetration and breaking of skin depending on the depth of the flow of the current, the mechanism drawn from the implant is useful for the purpose of analysing the effect of the present Exhibits P1 and P2, when used on human and this Court agree to the expert view of Professor Poon. 

189.  In agreeing to the conclusion drawn by Professor Poon, of course, the court has already considered all the opposing evidence.  This court is particular impressed by Professor Poon’s expertise both in electrical engineering and biomedical engineering, in particular, her experience in the design of medical electronic devices that measure or apply electrical current on human bodies, such as electrical knives or implants that provides neuromuscular electrical stimulation and the research and investigation of blood pressure and measurement devices.

190.  The attacks on or the doubts attempted to be thrown by the defence on the prosecution’s case only concentrate on the incapacitation aspect.  There is not much on this stunning aspect resulting from the fainting as a result of the blood pressure problem and this Court must say that the experience and the expertise of Professor Poon obviously far exceeded that of the other witnesses and her evidence is preferred. 

191.  In fact, both Dr Ng and Dr Chiao also agree that a biomedical engineering expert is the more appropriate person to comment on the issues. 

192.  This court, therefore, agrees with her analysis and conclusion on this issue.

Issue 3

193.  Now, coming to issue 3.  There is no dispute that D1 and D2 knew the subject devices can produce electric arc and can cause papers and tobacco to burn.  In fact, these lighters were sold as electric arc lighters.

194.  Professor Poon in her evidence admitted that the fact that the lighters can produce electric arc does not mean they are stunning device, it all depends on the actual electrical characteristics of each device. 

195.  Defence counsel submitted only two of the devices seized by the police satisfied the electrical characteristic of being a stunning device.  Some of the other devices seized, though can produce electrical arc, give zero voltage readings when tested.

196.  It is also a fact that when one examined the lot of devices, their design and outward appearance looked very similar and, as the police officers say, they could not really tell with their bare eyes which are the ones that fit into the stunning device and which are the ones that do not.

197.  Clearly, tests must be performed on the devices in order to determine whether they are, in fact, stunning.

198.  Both defendants elected not to give evidence in court.  No adverse inference would be drawn against them.  They have, however, previously given evidence in their cautioned interviews with the police.  D1, as a sales lady of the shop for some time, said the subject devices were brought by D2, her boss, and D2 did demonstrate to her how to use the device as an electric arc lighter.  She had no idea of the stunning or disabling nature of these devices. 

199.  D2 said that he was told by the one selling these devices that the current output of the devices was similar to that of a mobile phone and that they would not pose any danger.  He produced the name card of the seller of these devices in the mainland.

200.  This court noted that when the devices, Exhibits P1 and P12, were sold to the undercover police officers, at no time did D1 mention any function of the devices other than it as an electric arc lighter.  Nothing was said regarding a stunning or disabling function.  Even the little tag in the display window showing these devices only described them as electric arc lighter. There is no evidence before this court that these electric arc lighters were possessed up for sale to the extra function of stunning or disabling.  There is no evidence before this court that these electric arc lighters possessing the extra function of stunning or disabling were sold at a price over and above those which do not possess such function.

201.  In fact, when D2 used Exhibit P2 in her demonstration as to how this device performed or operated.  Later she, in fact, sold Exhibit P12 to the undercover officer.  Exhibit P12 is not a stunning device and is therefore different from Exhibit P2. 

202.  Now, did she intentionally sell him something different from what she demonstrated?  There is no evidence before this court in this respect and, therefore, it is possible that she thought the two devices were the same.  In fact, according to the evidence of D1, all the electric arc lighters were sold at the standard price of $398 each.  There is no difference in price for the sale of the lighter with or without the stunning device function.

203.  Following the Court of Appeal decision in HKSAR v Mohammed Khan Shamim, once defendants are found to be in possession of the subject devices, section 24(2) of the Firearms Ammunition Ordinance, Cap 238, imposes on them an evidential burden that they do not know the stunning or disabling nature of the subject devices. 

204.  The prosecution has referred me to the case of HKSAR v Cheuk Yuet Hing, HCCC 488/2015, where the defendant was found in possession of a stunning device, again, in the form of a lighter.  But in that case the court rejected the defendant’s explanation that he thought the device was just a lighter.  Because the court accepted that the defendant had in a cautioned interview told the police that the same was to be used for self-defence.  In our case here there is no such evidence before this court that any of the two defendants knew the devices had a function other than that of a lighter, the nature of which is not apparent from the design of the device.

205.  With their evidence in the cautioned interview, although there is this persuasive burden and with the observations I made regarding the manner of sale and how these goods were displayed, both defendants, in my view, have managed to marginally discharge their evidential burden and the prosecution, therefore, has failed to prove this issue 3, that these two defendants knew of the stunning or disabling nature of the devices, Exhibits P1 and P2.  For these reasons, they are both acquitted of the charges and the charges are dismissed.