I had a challenging EMC problem solving project in the lab this week.
A customer making a miniaturised 4 cm^3 buck-boost DC/DC converter for Li-Ion battery charging was having radiated emissions issues. The small size meant that adding common mode chokes to filter the input and output connections wasn’t practicable so a more in depth investigation was required.
How bad is it?
Here are the emissions for the EUT without any modifications. The green reference trace is the AC/DC mains power supply being used to power the EUT. It is failing the Class B limit (blue) by some margin.
Initial Isolation and Investigation
To investigate the emission radiation source (not the cause yet), I placed large clip on ferrite cores around the DC input cable and the battery output cable to reduce emissions directly from the cables.
This improves some of the frequencies but not all of them. If the radiation was entirely cable related then this would have dropped the emissions significantly. As it hasn’t, we can conclude that the majority of the emissions are coming from the PCB.
The three peaks we’ll focus on are 180MHz, 300MHz and 500MHz.
The location of the emissions for the 180MHz and 300MHz emissions was initially puzzling. Mostly it was centred around the drain of Q1. If we consider the operation of the circuit, Q1 is turned on permanently in boost mode with Q3 acting as the normal switching element and Q4 acting as a synchronous rectifier. Where is this switching noise coming from?
Those of you familiar with synchronous switching converter operation will be shouting at the screen right now. Of course, the answer is bootstrapping.
The high side N channel MOSFETs Q1 and Q4 need a gate voltage higher than their source voltage + their threshold voltage to turn on. In this kind of circuit, this voltage is derived from the switching node via a bootstrap circuit.
Even though Q1 is nominally on all the time it still needs to perform a switching operation with Q2 to charge up the bootstrap capacitor powering it’s gate driver circuit.
Checking the datasheet, this switching operation takes just 100ns. That’s very fast indeed and explains the source of our switching noise!
The same bootstrap operation is happening to provide the drive voltage for Q4 but because the boost node is continuously switching this voltage is being provided without such a short switching event.
Due to space constraints it wasn’t easy, but I managed to get the microscope out and modify the board to accept a small but high current ferrite bead in series with Q1 drain.
It didn’t take long to narrow down the 500MHz emissions to the boost output diode D1 with a large amount of ringing on the cathode.
The interesting thing about this diode is that it is only conducting for a very brief period in the dead-time between Q3 turning off and Q4 turning on. Dead time between these parts is set at 75ns, again a very short time period. Good for reducing switching losses, disadvantageous for EMC emission.
The part selected for this was a slightly electrically over-rated 40V 1A, SMB packaged part with a reasonable capacitance. Switching 1 amp of current through this part for only a brief period of time before shorting it out and discharging the diode capacitance was causing the ringing to occur.
A ferrite bead was added in series to damp this as the customer wasn’t too keen to head down the rabbit hole of investigating specifying a lower capacitance rated new diode or looking at whether the diode could have been removed altogether at the expense of slightly higher power dissipation in Q4.
Interestingly, this is what the emissions looked like with the diode removed but still with the lower frequency emissions present from the input transistor drain. Note the wideband reduction in emissions above 300MHz.
With both of the ferrite beads in place the emissions profile of the EUT was reduced to meet the Class B limits. With more time the peak at 160MHz could be investigated and further reduced but project time pressures and the customer understandably wanting a “good enough” result meant we concluded this investigation here.
DC/DC converters are often provide a challenging EMC opponent when it comes to radiated emissions. I was glad of the opportunity to work on this project and provide a successful result for the customer. This is the kind of work that I love.
The advantage of being an EMC-consultant-with-a-test-lab combined is that this kind of work can be compressed into hours of work rather than days/weeks oscillating between your lab and the test lab. Problems Fixed Fast!
I hope you found this piece useful, get in touch via the usual channels if you have any questions.
Ladies and gentlemen, I present this week’s episode of “Crimes Against Cables”
Example 1: “I had some leftover components to use”
I’ve seen plenty of interesting EMC “solutions” over the last several years to deal with radiation from cables.
A common one is to separate the shield ground from the signal ground with some combination of components (beads, capacitors, resistors). This approach appears to be particularly common on industrial touch screen display modules for some reason.
This is (in 99% of cases) a bad idea. I’m not sure what you are hoping to achieve by this and, probably, neither are you 😉
In fact I dedicated a small part of a recent talk to discussing grounds and grounding – you might want to check it out.
Example 2: How to Break a Shield
Another notable poor example was an otherwise well crafted piece of military equipment. Shielded connectors and cables all over, it looked like it would be survive some serious electromagnetic abuse (as anything being tested to MIL-STD-461 should).
This ends up being not only an emissions problem but an immunity one as well as the cables are just as capable of conducting noise into the shielded case.
This sort of thing can be solved with something like an EESeal type component or by a secondary external screen over the entire assembly.
Example 3: Plastic Fantastic
I’ve even seen ferrite cores that were just a moulded plastic lump to appear like cores. Maybe it was a “special” plastic? I never found out, it didn’t help the emissions either.
But this next one was a first even for me.
Example 4 – The Strangest Decision Yet
I was performing a full set of EMC tests to EN 55032 and EN 55035 for a customer. The product had a HDMI interface so obviously there were radiated emissions problems.
The first step as a diagnostic was to use some copper tape to make a connection between the connector shell and the metal back plate – the anodised chassis and EMI gasket material provided was not making a good contact.
This didn’t help so I buzzed the connection with the multimeter to make sure I had some continuity and… nothing.
No connection between the connector shell and PCB Ground.
OK, so there must be a capacitor in series with the shield connection. Fetch the capacitance meter and… 1.2pF.
The board designer had neglected to connect the shield of the HDMI to PCB Ground. It’s a new one for me!
The addition of copper foil to bridge the connector pins to nearby solved the emissions problem but left me wondering why someone thought that was a good idea.
I’m going to leave you with this closing thought:
I’ve yet to come across an EMC problem where floating or not connecting a shield ground has improved the situation.
https://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2020/11/DSC_2492-scaled.jpg14402560James Pawsonhttp://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2017/01/unit3compliance_400x400-80x80.pngJames Pawson2020-11-27 22:19:122020-11-27 22:22:55HDMI? More like HDM-WHY? Thoughts on Cable Shield Grounding
A new customer came to me with their product that was having problems during testing at another laboratory. There were radiated emissions problems (mostly solved with improvements to the ground plane scheme on the PCB) and a very interesting (and challenging) ESD problem which I’ll cover in this blog.
Here was the device exhibiting the problem, a Diodes Inc AP22802AW5-7 “power distribution load switch”. Input VBAT from a stick of AA batteries, SW_PWR from a rotary switch, and output to the rest of the circuit.
The ESD problem was described by the customer:
The EUT stopped working when 4kV contact discharges were applied on discharge point shown. I removed the batteries and I put them [in] again and there was not any response from the sample (no otuput and the green LED remained OFF).
[A second sample] was then tested with the same result, although this time not on the first discharge
Upon inspection both devices had failed due to the load switch (AP22802AW5-7Diodes), with one failing open and one failing short and both becoming very warm.
ESD diode placed on input and output of load switch (with no effect)
ESD diodes placed on all [discharge points] (with no effect)
ESD diode places on VCC close to pullup resistors for [discharge points] with no effect
First thing first was to get the product set up on the ESD table (with a bit of added blur to protect the innocent).
It was very easy to re-create the problem observed at the original test lab with the second contact discharge to the EUT exposed contact point causing the unit to shut down.
In each case, the power switch was failing low resistance from IN to GND. The initial theory was that the device was being damaged by the high voltage punching through the silicon layers leaving a conductive path.
Eliminate the possible
I made a series of experiments to determine the coupling path into the problematic device. Working on the principle that, because of the 15cm distance between discharge point and problem device, that conduction might have been the problem.
Capacitors on Vin and EN
plus disconnect EN line
plus ferrite beads and capacitors on Vin, Vout and EN
plus local TVS diodes on pins of device
plus ferrite beads in series with [EUT input] lines
Whilst none of these experiments were successful they certainly helped eliminate conduction as the coupling path.
Because of the very high frequency content of the ESD pulse, capacitive coupling is likely going to be the dominant coupling method. Whilst it could couple into the device directly, there was more opportunity for the pulse to couple into the traces connected to the device first. Filtering the inputs eliminates two coupling possibilities
Change of sample
The PCB was starting to get a bit tired from the repeated hot air SMT de-soldering and re-soldering so I swapped to another supplied sample. To be able to operate the unit out of the casing I swapped to a linear DC bench supply instead of the AA batteries.
This proved to be an interesting mode as it allowed me to kill the power quickly. The next set of experiments were in an attempt to reduce the effect of capacitive coupling to the problem device.
Improved ground stitching / connection
Changing supply voltage
Indirect HCP discharge – not to EUT but to the Horizontal Coupling Plane albeit with a vertical ESD gun to increase capacitive coupling to EUT.
Reduction of coupling into Vin terminal by removing components and copper
Addition of copper foil shield over the top of the device
Failure mode discovery
Setting the current limit on the DC supply to a fairly low value (about 20% higher than nominal current draw) was a good idea.
When applying the ESD strikes the supply went into foldback as the EUT power input went low resistance. I discovered that quickly turning off the power and then turning it back on effectively reset the failure mode of the device. This proved to be repeatable over several discharges: zap – foldback – power cycle – EUT OK.
Here’s the VI curve of an undamaged device. It’s a bipolar voltage between VIN and GND. On the left of centre is the standard forward biased body diode. On the right is the reverse biased breakdown of around 8V.
Now for a damaged device. In this case the current changes quickly for a small applied voltage and there is no non-linear characteristic. Essentially, a short circuit.
Turning up the maximum voltage that the curve tracer can apply and dialling down the series impedance allowed me to simulate the over voltage fault condition and create a latch up condition. This latch up wasn’t permanent due to the bipolar sine wave nature of the curve tracer applied voltage.
However turning up the voltage enough to cause excess power dissipation inside the device did result in the same failure mode using the curve tracer.
I have never encountered a device that is this unusually sensitive to ESD events before. A nearby 2kV discharge on the PCB top layer ground plane was enough to cause the latch up condition.
I noted in the report to the customer that this device had been changed to “not recommend for new designs” by Diodes Inc. I wonder if they identified this condition in the device and withdrew it for that reason.
The customer resolved the issue by replacing the device with a different part and we all lived happily ever after.
https://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2020/09/power-switch-circuit-block.png386749James Pawsonhttp://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2017/01/unit3compliance_400x400-80x80.pngJames Pawson2020-09-07 21:57:282020-09-07 21:57:28ESD Latch Up Behaviour in Diodes Inc. Power Switch Parts
The LVD a.c. voltage limits are defined in terms of rms. From 2014/35/EU
This Directive shall apply to electrical equipment designed for use with a voltage rating of between 50 and 1 000 V for alternating current and between 75 and 1 500 V for direct current
So that’s 50Vrms and 1000Vrms.
A customer was asking about a low current (sub 1mA) 40Vac rms source and if the LVD applied.
This voltage is less than the 50Vrms threshold mentioned above so would be technically exempt from the LVD. Unless of course the equipment contained a radio module in which case the RED makes the LVD applies with no lower voltage limit.
I think that there is still a risk that needs to be assessed here.
Just because you are exempt from the directive doesn’t mean you are exempt from making your product as safe as possible.
Looking at the main table from 62368-1 for categorising shock risk, 40Vrms/50Hz means it is an ES2 hazard. The voltage source would have to be under this limit for normal operation, abnormal operation (e.g. blocked vent, stalled motor, controls set incorrectly) and single fault (open/short circuit) conditions.
This means, even though the LVD is not strictly not applicable, it would be wise to put in a Basic Safeguard between the user and the exposed voltage.
Additional: The provisions of the General Product Safety Directive (2001/95/EC) would apply to any product falling outside of any specific safety standard. The Harmonised Standards for this Directive include EN 60065 and EN 60950-1. Since both of these have been superseded by EN 62368-1 it would be reasonable to use this standard instead.
Thanks to Charlie Blackham from Sulis Consultants for the tip.
This safeguard could be an enclosure, insulation, an interlock or barrier.
Instructions or PPE aren’t sufficient as they are considered supplementary safeguards.
But what about the current limits?
That’s just considering the voltage source purely from a voltage perspective. If it can’t drive enough current into a 2000 ohm load for more than 2 seconds to form a hazard then that might change the classification.
This current is measured using one of the appropriate networks from EN 60990 such as the one below
But I know what I’m doing…
The requirement for safeguards depends on if you classify the user as a normal person or an instructed person
Skilled person > instructed person > normal person
22.214.171.124 instructed person person instructed or supervised by a skilled person as to energy sources and who can
responsibly use equipment safeguards and precautionary safeguards with respect to those
ordinary person person who is neither a skilled person nor an instructed person
skilled person person with relevant education or experience to enable him or her to identify hazards and to
take appropriate actions to reduce the risks of injury to themselves and others
The level of safeguard required between the user and the ES2 hazard is defined in EN 62368-1
For a normal person we must use a basic safeguard
But for an instructed person we may use a precautionary safeguard
A Precautionary safegard (defined in 0.5.5.3) could take the form of instructions or training, but the addition of warning stickers, PPE could also be considered part of this.
This is why I like EN 62368-1 over some of the older safety standards (I’m looking at you, 60950)
Rather than a prescriptive “thou shalt use 2.5mm clearance or be smote verily” it helps and guides you through all the steps into understanding why or how a requirement is derived.
Also the companion EN 62368-2 explanatory document contains even more background and context. I wouldn’t recommend applying -1 without having -2 to hand.
Stay SAFE kids.
https://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2020/06/table4.png3781237James Pawsonhttp://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2017/01/unit3compliance_400x400-80x80.pngJames Pawson2020-06-18 11:02:402020-06-26 08:49:23LVD Voltage Limits are RMS + Thoughts on Marginal Voltages
“EMC Testing (TBC): Must comply with IEC 60601-1-2:2014, Medical electrical equipment — Part 1-2: General requirements for basic safety and essential performance — Collateral Standard: Electromagnetic disturbances — Requirements and tests”
So lets take a look at a the EMC tests that might be required for a typical Rapidly Manufactured Ventilator project.
This is with the view of meeting the Essential Performance / Basic Safety requirements of EN 60601 whilst addressing the highest risk items first. This is prioritising speed of testing instead of performing a belt and braces, test everything approach that would be the common approach for Medical Devices.
These ventilators are going to be used in a hospital / clinical care environment under medical supervised use and not in a home environment.
60601-1-2 classifies a hospital as a Class A emissions environment for Radiated and Conducted emissions. This means that less time needs to be spent fighting to get the emissions below Class B.
Risk items to radiated emissions could include any brushed DC pump motors as these are notoriously noisy. Ferrite cores may be required around motor cables to mitigate this noise.
Following the design guidelines further down this article for any PCBs is recommended and will greatly assist with reducing EMC radiated emissions.
Most RMVs will be using an off the shelf power supply already approved to EN 60601 for Safety and EMC. AC Power conducted emissions should therefore look after itself and won’t be a significant worry for testing.
For Harmonic Distortion and Flicker, there is an interesting note in EN 60601-1-2 in Annex A
“It is assumed that ME EQUIPMENT and ME SYSTEMS used in hospitals (and large clinics) are not connected to the PUBLIC MAINS NETWORK.”
If this is the case, then Harmonics and Flicker requirements and tests need not apply as these only relate to the public mains network.
Most ventilator systems have no external electrical ports apart from the power supply. They are mostly self contained units. This greatly speeds up and simplifies the testing, and reduces the risk of problems with Signal Input / Signal Output Ports (SIP/SOP Ports in the standard, analogous to a Signal Port from other EMC standards).
Mains borne interference (EFT, Surge, Conducted RF) should be handled by the EN 60601 pre-approved power supply without issue. It will still need checking but ultimately the risk is low.
Dips and Interrupts and the hold up time of the power supply is something that would need considering at the Risk Analysis level to derive the correct immunity criteria for each of the individual tests.
If this needs improving then selecting a slightly larger power supply than nominally required could help. More likely, additional bulk decoupling on the main power rail (e.g. 1000uF) will help maintain the system DC voltage under these conditions.
Immunity Performance Criteria
Caveat: This section is me thinking aloud as I have no domain specific knowledge for Risk Management and Medical Devices. I’m trying to approach this from a common sense perspective to aid anyone working on a RMV.
The function of the EMC Immunity tests for Medical Devices is to ensure that the Essential Functions continue to operate and that Basic Safety condition is maintained.
Normally the immunity performance criteria would be based on the type of EM phenomena being simulated in the test. This is normally Criteria A for continuous phenomena (radiated or conducted RF immunity) or Criteria B for momentary phenomena (ESD, EFT, Surge, Dips/Interrupts). Criteria C only tends to crop up for longer duration power interruptions.
In the case of a Medical Device, maintaining the Essential Performance is the key parameter. If a momentary EM phenomena causes this to happen then this is a major problem.
Therefore immunity performance criteria must be considered for the key function of the device as well as the duration of the EM phenomena.
Based on this thought process, a sensible starting point for the immunity criteria is:
Assume criteria A (unaffected performance) for:
Key function of assisting patient breathing for all tests. This includes momentary EM phenomena tests = ESD, EFT, Surge, Dips/Interrupts.
Non-critical functions under continuous EM phenomena tests = Radiated and Conducted RF Immunity
Assume Criteria B for:
Key function performance for this means that there should be a function in the RVM firmware that remembers its last current operating state and settings and that it starts up in that state from a power cycle. This creates a requirement for programmable non-volatile memory (some kind of EEPROM) in the RVM.
All momentary EM phenomena tests for non-critical functions e.g. display readout may temporarily distort or flicker so long as it recovers
Assume Criteria C for:
Non-critical functions from momentary power loss e.g. screen/display readout or setting
There are two big risks to the immunity performance: Radiated RF Immunity and ESD.
Radiated RF Immunity
The basic requirement for radiated RF immunity is a flat 3V/m from 80MHz to 2.7GHz. So far so good, this is a fairly easy test to meet.
Now the bad news. Table 9 gives a list of spot frequencies to be tested to simulate close range exposure to common wireless technology standards. The table is summarised here:
Test Level (V/m)
18 Hz pulse, 50%
FM +/- 5kHz dev. 1kHz sine
710, 745, 780
217 Hz pulse, 50%
810, 870, 930
18 Hz pulse, 50%
1720, 1845, 1970
217 Hz pulse, 50%
217 Hz pulse, 50%
5240, 5500, 5785
217 Hz pulse, 50%
As you can see, this has testing up to 28V/m, a significantly higher field strength than 3V/m!
Risks to the EUT
This test loves to mess with analogue sensors. In the case of ventilators, the pressure sensors used frequently have an analogue output to a DAC on the CPU. This presents two risk areas:
Demodulation of noise inside the pressure sensor amplifier. This takes the small transducer signals and amplifies it up to the output voltage. Noise demodulated here would cause the carrier to be superimposed on the pressure readings.
The input of the ADC could be susceptible to noise picked up on the analogue voltage from the pressure sensor, even if the pressure sensor itself is unaffected. This will affect the readings.
Since the airflow and pressure sensors are a key component to the operation of the ventilator, these must be protected at all cost.
Design ideas to mitigate this interference include
Keep traces/connection as short as possible between sensors and ADC
If you can mount them all on the same circuit board then do so
This circuit board will have one layer dedicated to a solid ground plane fill over the entire plane. All ground pins
Cables = antennas that are good at receiving the interference. Minimise use of cables where possible.
Figure out what your minimum bandwith requirements for airflow are and filter the signal appropriately. You probably won’t need to sample the airflow faster than 10kHz so put a low pass filter right next to the ADC input. Something like a 4k7 and a 1nF will give you a 3dB of 34kHz. This will reduce the risk of RF noise being demodulated by the ADC input.
Decouple the supply lines to the pressure sensor well
Add a small filter to the pressure sensor input, perhaps another RC filter as shown above. This will help prevent the pressure sensor from being affected by the test.
It is possible that the pressure sensor will be directly affected by the radiated noise picked up by the sensor body itself and not by the traces. It would be prudent to provide a PCB footprint for a shielding can near the sensor. I have seen this effect on gas sensors in the past.
Assuming that the advice above is followed, the risk to the EUT is manageable.
One of the interesting features of Radiated RF Immunity testing is that of the Problem Band where most issues occur.
Most of the time, the problem band is in the 100MHz to 300MHz area (I’ll cover this in more detail in a future article). Cables tend to be the best antennae at these frequencies and, hopefully, our ventilator only has one cable of interest – the AC power cable. This has plenty of filtering for conducted emissions reduction which should handle this noise.
Probably the two biggest problem frequencies from the spot frequencies above are going to be 385 MHz and 450 MHz.
Then we are into the realms of direct pickup on internal signal cables and PCB traces at higher frequencies. If we’ve laid out our PCB well as highlighted above (short analogue traces, filtering, good ground plane, shielding provision) then this will help mitigate our risks.
The levels of ESD testing are almost twice that of the regular EMC standards with a requirement for 8kV contact and 15kV air discharges.
ESD is very good at upsetting digital systems and it has a particular fondness for edge triggered pins e.g. reset lines and interrupts.
If the reset line for the CPU controlling the RVM is shared with other digital circuit blocks or supervisory controllers then an RC low pass filter at the input to the CPU is highly recommended. This helps prevents unwanted resets.
Checking can be implemented in the Interrupt Service Routine to ensure that an interrupt condition actually exists, effectively de-bouncing the input.
Thankfully the Ingress Protection requirements for the RVM of IP22 and the requirement to provide flat, easily cleanable surfaces will probably dictate the use of some kind sealed membrane keyswitch panel. These have good ESD immunity as no direct contact discharge can take place on an switch where the plastic covering remains in place.
Whatever user interface technology the RVM employs, this will be a key risk area for ESD. If this is on a separate PCB to the main controller, all interfaces will need some kind of filtering. A small capacitor to ground on each of the lines that goes to the keypad would be a good idea. 0603, 100pF usually works well here.
Lastly on the mechanical design, keeping the electronics well away from the enclosure seams will also reduce the risks of creepage of any discharge into the circuit board.
Summary Test Plan
Radiated Emissions, Class A, 30MHz to 1GHz (EN 55011)
Mains Conducted Emissions, Class A, 150kHz to 30MHz (EN 55011)
Text in bold is highlighted as a risk item.
ESD, (EN 61000-4-2), 8kV contact, 2/4/8/15kV air. Test to connectors as well.
Radiated RF Immunity (EN 61000-4-3)
80MHz to 2.7GHz @ 3V/m
Various spot frequencies at up to 27V/m
EFT (EN 61000-4-4), AC Mains Port, 2kV
Surge (EN 61000-4-5), AC Mains Port, 1kV line-to-line, 2kV line-to-ground
Conducted RF Immunity (EN 61000-4-6), AC Mains Port, 3V/m (6V/m in ISM bands)
Dips and Interrupts (EN 61000-4-11), AC Mains Port, various
Not only has this article identified key EMC risks to Rapidly Manufactured Ventilators but also provided some design guidelines to dealing with the problems that might arise.
Some of the guidelines within might be useful to anyone designing a Medical Device. We haven’t covered the requirements for Patient Coupled Ports or SIP/SOP ports from an EMC perspective as they aren’t of too critical a concern for an RVM.
We can see how looking at the standard and pulling out the required tests can help us understand the risks involved in the design.
Experience of knowing how the tests will typically affect the EUT is the key to unlocking good design practices. In my case, this comes from having worked on many designs with problems and the learning that comes from fixing the issues that crop up.
Remember EMC test success comes from good EMC design. For a time critical RVM there is one chance to get it right – no do-overs!
I hope you found this article useful. See you when all this has calmed down. All the best, James.
https://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2020/04/BIPAP.jpg7981024James Pawsonhttp://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2017/01/unit3compliance_400x400-80x80.pngJames Pawson2020-04-18 15:55:522020-04-18 16:05:00Ventilator Projects and EMC Testing (EN 60601-1-2:2014)
1. CE marking probably won’t be necessary for UK for temporary ventilators 2. Still need to apply to MHRA for permission to use in clinical setting 3. This will likely involve the supply of test data to MHRA 4. The requirement for EMC testing is currently To Be Confirmed 5. U3C will give any ventilator design a pro bono EMC design review
In the current Covid-19 pandemic there are many project teams around the world coming up with designs for ventilators. The potential is for many such devices to be required.
The problem that this article identifies is that for normal medical devices, the length of time taken to get products approved is defined in terms of months. This is especially true for ventilators which are, according to several sources, a Class IIb device requiring a robust quality system and Notified Body approval.
The current requirement is for devices to be delivered within weeks.
The procedure covered under the “Exceptional use of non-CE marked medical devices” involves applying to the MHRA directly for a pass to supply this product for use without CE marking. This request will have to include data that shows compliance with the test criteria in Appendix B of the MHRA specification.
Page 11: “3. When the current emergency has passed these devices will NOT be usable for routine care unless they have been CE marked through the Medical Device Regulations. The device must display a prominent indelible label to this effect.“
Page 24: “EMC Testing (TBC): Must comply with IEC 60601-1-2:2014, Medical electrical equipment — Part 1-2: General requirements for basic safety and essential performance — Collateral Standard: Electromagnetic disturbances — Requirements and tests”
Note that the requirement for EMC testing remains TBC = To Be Confirmed. I strongly feel that at least some testing should take place, but functionality of the ventilator has to come first.
Next article will be a breakdown of the EMC test requirements called up in EN 60601-1-2:2014. I’ll also highlight some risk areas that designers of these ventilators will need to to bear in mind.
If anyone has any need for pro bono EMC design review services for ventilator projects or needs some EMC testing turning around quickly then get in touch.
I had an email from a customer that I’m working on some design consultancy work with, saying that one of their prototype products was having some radiated emissions problems at an accredited lab. Could I take a look?
Absolutely, EMC radiated emissions problem solving is my favourite part of the job! Ironically, it is usually the customers least favourite part!
Thankfully I had a slot free the next week so they bundled their kit into the car for the long drive “Up North” from their base in the South West of the UK.
After some tinkering, the equipment was set up in the chamber for some radiated emissions work. The first scan confirmed the problem levels and frequencies that had been observed at the other laboratory.
The problem areas from their last scan were at 35MHz, 80-90MHz and a broad band between 150MHz and 220MHz.
The system was housed inside a nice aluminium case that was being used for CPU heatsinking and environmental protection as well as EMC shielding. A rough diagram of the internals shows a main PCB with a large CPU / memory block in the centre and a variety of cables leaving the PCB and the casing.
The main power cable housing also had two debug connections inside the same housing that weren’t being used in the field but were available for updating software and such like.
As is so often the case, this product was in it’s final stages of the development life cycle, meaning that no major design changes were possible. These EMC problems would have to be resolved using easy to fit additional components. Thankfully I have plenty of things in stock to try out.
There are two important characteristics about these emissions that show us where to look
They are predominantly broadband, an indication of analogue noise e.g. DC/DC converter / power supply. Sometimes this broadband noise is generated by digital switching but this can be less common.
They are all low in frequency, where large or long structures are the most efficient antennae. This usually means cables.
So power noise and cables…. hmmm…. any good ideas?
OK Kids, Let’s Take a Look at the Cables.
In a very sensible move by the designer, both the DC power and Ethernet cables had some common mode filtering on the PCB.
Ethernet magnetics have common mode chokes built into the transformer stack which reduces the noise emitted and increases the susceptibility performance of Ethernet despite the often unshielded twisted pair cables used.
The caveat is that once the cables have left the magnetics that they must be protected from other interference sources. Noise coupling on to these lines is going to be heading straight out of the enclosure using these lines as the antenna. Similarly, if common mode noise gets onto the centre-tap of the output side of the magnetics then this can also cause similar issues.
I have experienced system noise coupling on internally routed Ethernet cables before and it nearly always results in lots of low frequency emissions.
The power cable had a small surface mount Murata filter in place with excellent attenuation at the frequencies of interest.
Both the Ethernet and power cables pass through the shielded enclosure with no connection or filtering to the case. In bypassing the quite nice Faraday cage of the enclosure, any noise current on these lines will inevitably appear as radiated emissions and be picked up by the receive antenna..
Now to find out some more info.
Radiated Emissions Experiments
First, unplugging the Ethernet cable dropped the emissions significantly from 30MHz to 120MHz.
Secondly, some messing around with ferrite cores on the power cable reduced the 150MHz to 220MHz hump down to sensible levels.
This left a single peak at 270MHz that was traced to noise using the coaxial RF cables to the antenna to radiate.
Lets look at each of the points in a bit more detail:
The only practical method of dealing with the Ethernet emissions was to change the bulkhead connector to a metallic screened version and the external cable to a SSTP (Screened Shielded Twisted Pair) type of cable. No exciting analysis here I’m afraid.
Details of the Power Cable Noise Coupling
The most interesting coupling mechanism was happening inside the un-screened bulkhead power connector. Thanks to the power filter on the PCB, there was very little noise being conducted back down the cable from this line. However, the debug connections to the CPU are picking up all kinds of noise and carrying that noise to the connector.
Disconnecting and bundling the debug cables near the connector cuts the radiated emissions down to next to nothing.
What’s most interesting is that the capacitive coupling region between the power cable and the internal debug cables is so small. The connector is only 20mm long and the cables run parallel with each other for barely any distance. And yet there is enough noise current being coupled onto these lines that it causes a radiated emissions problem.
Details of the RF Antennae Noise Coupling
By the time that all of the cables had been filtered or removed, there remained just one emission at 270MHz that was failing the Class B limit. An investigation with RF current probes showed a lack of noise on the main output cables listed above, even when they were screened or filtered appropriately.
A wander round the enclosure with an electric near field probe and spectrum analyser showed a spike in emissions near the RF antenna housing on the side of the EUT.
Checking the antenna feed cables showed them connected to the PCB pretty centrally. Disconnecting the coaxial cables from their mating halves dropped the emissions down to the noise floor.
Even though the noise isn’t in-band for the antennae themselves, they still perform well enough to radiate the noise and cause an emissions problem.
Summary of Fixes Applied
The below diagram shows the fixes applied to the EUT to achieve a Class B pass.
Firstly, a fully screened metal bulkhead Ethernet connector was chosen for use with a shielded cable. This isn’t ideal from the installation point of view but is ultimately unavoidable without more significant modifications to the EUT.
Secondly, a Wurth ferrite was equipped around all three of the cables connected to the power bulkhead connector. As detailed above, it is necessary to put the ferrite around all three cables and not just the power to reduce the noise entering the capacitive coupling region around the connector.
Thirdly, a small ferrite was placed around each of the UFL cables at the point at which the antenna cables left the housing. This is a fairly common modification for radiated emissions, one I’ve employed several times before, and there are numerous suppliers of ferrites of various lengths with just the right inside diameter for the type of thin coaxial cable used with UFL connectors.
Any time your cable passes through a shielded enclosure with no RF termination at that point, you can pretty much guarantee its going to need some filtering.
Nothing particularly in depth in this analysis of the EUT, but I did find the coupling in and around the power connector particularly interesting.
At the end of the day, the best outcome was a happier customer with a path forward for their product.
ESD protection is essential to control the Electro-Static Discharge event from damaging sensitive circuitry within a product. But its location within the system needs to be considered carefully and is sometimes not obvious at the schematic level.
I’d like to share with you a great example of this that I found whilst working on a customer’s system. I probably wouldn’t have spotted this without testing but I will certainly have it in mind for future design reviews.
In this instance, the Equipment Under Test is formed from a 2 part metal chassis consisting a large base and a hinged lid. On the lid there is a membrane keypad that interfaces via a Flat Flexible Cable (FFC) to the front panel PCB. There is a second ribbon cable from front panel PCB to CPU board carrying the button presses to the processor.
The ESD protection is on the front panel cable, next to the point where the unit is likely to be touched – the keys. So far, so good.
The base and lid are connected elsewhere via the typical long piece of green and yellow wire for electrical safety purposes. The inductance of this connection (long wire, single point) means that it has minimal effect at the high frequencies present in an ESD waveform. Also, the case halves are separated by a rubber environmental seal meaning there is no contact around the edge of the case.
EUT + ESD = ???
So what happens when the EUT is subject to an ESD event? There is no discharge to the plastic membrane keypad on the top and discharges to the Vertical Coupling Plane don’t have any effect. However, when a discharge is made to the seam between the lid and base, something interesting happens.
Because of the conditions mentioned earlier (large seam with a significant, remote impedance connecting the lid to the base) the pulse is free to couple to the internal cable assembly as shown below.
Because the ESD protection is on the front panel display board it is unable to prevent the flow of high frequency current down the cable and into the CPU.
The effect of the discharge is to cause the entire system to reset and eventually the GPIO lines responsible for monitoring the front panel keys were damaged to the point of non functionality.
On the face of it, the designers had acted sensibly; the ESD protection was right next to the interface that was likely to be touched by the user. However, the design of the case and the routing of the cable proved to be a problem – something that was not anticipated.
With the addition of some simple capacitive filtering or ESD protection at the point at which the cable enters the CPU board this problem was overcome.
There are lessons for us all here that I would summarise as:
Consider every cable as a risk, even internal ones
Watch out for cables crossing enclosure seams or apertures where coupling is a risk. Not a dissimilar situation to a PCB trace crossing a split in a ground plane – and we all know how bad those can be, right?
Consider how the PCBs and cables will be integrated within the system through a mechanical design review (with your EMC hat on)
It doesn’t matter how well designed you think your system is, testing is necessary to find these problems
I’ve always been impressed with Apple’s approach to reducing problems caused by EMC/EMI. Making top of the line technology in a compact case means minimising risk and maximising performance.
Let’s look at an example of well considered EMC design and why it is so useful.
Even the EMI shielding solutions are stylish
Because their products are charged at top dollar prices, they can afford to (or can’t afford NOT to) put in features like this.
The RAM on the new Mac Mini (thanks to iFixit for the great photos) has its own removable cage, secured to a PCB level counterpart with screws and, no doubt, a decent fit along the edges. What’s interesting is that this shielding system will have multiple functions.
Primarily, it will be used to reduce the EMC radiated emissions from the product. The Apple products I’ve had in my anechoic chamber have all been very quiet and this is why I hold Apple in some regard for their EMC design.
Apple will no doubt have tested their design with multiple RAM vendors to satisfy themselves that the design meets the requirements of international EMC standards.
However, were the user to install some non-Apple verified memory modules then the risk of emissions could increase. One can well imagine that Apple will have considered this in their EMC Risk Assessment.
The secondary benefit is more subtle. Take a look at this image.
Original image courtesy of iFixit, markup by author
The memory modules and their screening can are highlighted in red. Next to it, highlighted in green, is a smaller board level shielding and a UFL antenna connector. (There are another two connectors out of sight underneath the case)
That’s right, Apple have put the most noisy part of the system (RAM) right next to one of the most noise-sensitive (Wi-Fi). What?
This is not an uncommon problem, especially when trying to compress so much functionality into such a small space.
The Mac Mini is only 165mm square (that’s 6.5″ if you are watching in black and white). The case includes an integrated mains power supply making proximity between electromagnetically incompatible systems unavoidable.
Modern RAM speeds are fast and the Mac mini is no exception. Everymac lists the latest Core i7 model with a DDR4 memory speed of 2.66GHz. That’s uncomfortably close to the Wi-Fi operating band of 2.4 to 2.5GHz.
The interference spectra of a DRAM interface fundamental frequency is generally quite wide band.
If you turn on any form of Spread Spectrum Clocking (SSC) to reduce the peak energy then it can spread over tens or hundreds of MHz. Either way, that puts the edges of the memory fundamental in band for the 802.11 a/b/g/n/ac interface on the Mac mini.
The harmonic emissions of the memory are also prevalent and it’s easy for these to fall in-band of a wireless interface like Wi-Fi. For instance the second harmonic of 2.66GHz is at 5.32GHz in the channel 64/68 region for 5GHz Wi-Fi. Big problems.
Improve Performance? The Can Can.
The effect of in band interference on a Wi-Fi interface can be subtle.
At it’s most gentle, there’s a reduction in both performance and range. The modulation, coding type and channel width of the Wi-Fi sets the robustness of the interface to interference.
At the other end of the scale, whole channels can be blocked out entirely.
This intra-system, or platform level interference is pernicious and can be difficult to isolate and track down. Low noise floor spectrum real-time analysers are extremely useful tools here.
Ultimately, segregating the noise source from the receiver, is the only real solution. This can be achieved by physically separating the aggressor and victim (not possible here) or by shielding.
For some companies, the fallout in performance of a couple of Wi-Fi channels is no big deal.
If you are Apple however, then you can’t afford to have dissatisfied customers complaining about poor Wi-Fi speeds. As always, the EMC budget has to be congruent with the product budget and the desired performance.
The Last Line Of Defence
Check out the textured surface between the mounting holes for the lid (blue highlight on the above photo). That will be an EMI seal to ensure good contact between lid and case. Not only a nice touch but an important one.
The Wi-Fi antenna is mounted on the outside of the shield so this circular lid actually screens the antenna further from the noisy internal circuitry of the mini.
Well done Apple. I’d love to see your Wi-Fi range testing results… please?
A customer had purchased some power supplies from Amazon UK to get started with the development on their product. And why not? There are lots of cheap products available and everyone has a budget to meet. The chances are that they’ll get damaged, lost or broken anyway.
They were happy with the (perceived) quality of the PSU so approached the manufacturer directly for bulk pricing for volume production. However, the Amazon sample made it’s way to Unit 3 Compliance for EMC pre-compliance testing where the fun began…
Externally, the only way to tell the difference between the compliant and non-compliant versions is a slight difference in the length of the barrel connector and a slightly different shape of strain relief grommet.
These devices are being marketed as the same device on the outside and yet are completely different on the inside!
I’ve not been able to subsequently find this exact power supply on Amazon but there are similar looking variants still available.
A Real Problem
Crucially, it’s not just EMC that is being sacrificed. This “race to the bottom” of extracting every last penny from products has more serious consequences.
At a previous employer, an inspection was performed on 50 power supplies (again, bought from Amazon) that one of the project teams had purchased for powering various development platforms within the company. This revealed some serious safety problems (creepage and clearance) resulting in the entire batch being quarantined and scrapped for recycling.
Another aspect to consider – if the manufacturer has two different, almost indistinguishable products then how does your supply chain guarantee that you will receive the correct one? What is to stop the manufacturer from swapping out the more expensive compliant power supply halfway through production?
The principle of caveat emptor still applies. Disingenuous product markings are being used to falsely indicate compliance.
What To Do?
The obvious way round this is only to buy small quantity power supplies from trusted suppliers. I know from working with other customers that suppliers like RS and Farnell / Element 14 take compliance seriously. Buying from these sources is more expensive financially but what price do you put on your own safety?
If you are relying on buying a pre-approved power supply always ask for the EMC and safety test reports and the Declaration of Conformity. A supplier who cannot readily supply these readily should be disregarded.
Compare the details in the reports with the physical sample in front of you. Especially for safety reports, photos of the unit are generally included, inside and out. Look for any differences between the two.
Differences in EMC performance are not obvious. The only way to be sure of the quoted performance is to perform some quick tests, conducted and radiated emissions being the two main ones.
How We Can Help.
Here at Unit 3 Compliance we can give you some peace of mind that your power supply isn’t going to cause you any issues. Some of the things we do include:
Provide full EMC testing for all off the shelf products
Electrical safety analysis and testing
Help you understand the compromises and
We can review test reports and compare to physical samples with an experienced eye
Every incoming customer power supply is given a HiPot test as standard to help catch any problems
https://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2018/10/power-supply-comparison-header-1.png308841James Pawsonhttp://www.unit3compliance.co.uk/wordpress/wp-content/uploads/2017/01/unit3compliance_400x400-80x80.pngJames Pawson2018-10-26 09:49:372018-10-29 07:58:08Off The Shelf and Non-Compliant Power Supplies (from Amazon)
Unit 3 Compliance Ltd
2, Wellington Business Park
New Lane, Bradford