a roll of Wurth Elektronik copper tape - the scoundrels last resort?

So You Want To Be An EMC Engineer?

 

“Abandon hope all ye who enter here”

– Sign above the door on any EMC lab.

 

I’ve been asked a couple of times for career advice in relation to EMC. How do I get into EMC in the first place? How do I progress, perhaps moving from testing to design? Where should I take my career?

I’m generally sceptical about people who offer career advice. Much advice tends to be parochial “do this and you will succeed”. It is based entirely on what the person giving the advice thinks you should do (even if they never did it themselves.

Everyone’s upbringing and experience is so different there is no “one size fits all” approach to any career.

I can only share what I have done.

Maybe it will help.

 

Pre-Flight Check # 1: Make sure you are in the right career

Too many people are guided into careers like doctor, lawyer, engineer that might not be the best fit for them.

Make sure that engineering is right for you.

If you aren’t sure (and that’s OK) then writers like Tim Urban (career advice featuring the Yearning Octopus and your mum in disguise – long read but worthwhile) or James Altucher have lots of thought provoking advice for you.

I think being an engineer is more of a vocation than a job. If you cut most engineers through the middle it will say ENGINEER like a stick of Blackpool rock (a very British analogy). The chances are, if you are reading this, you are already in this category.

 

Pre-Flight Check # 2: Be honest about your reasons for wanting to get into EMC

Why are you wanting to get into the world of EMC?

Wanting something impressive on your CV? Think it might be a good way to get to that promotion you’ve been after? Probably will, but if these are your only reasons then you might be frustrated by the learning curve associated with the field.

One good answer is “it sounds really interesting.” If these are your thoughts then you are not wrong. I think it is one of the most fascinating fields of electronics.

In my case I was cheesed off with working in project management where I was spending less time with my soldering iron and more time in bullshit meetings. An opportunity for an EMC engineer came up in the organisation I worked for and without even thinking about it too deeply I said “I’ll do it”.

Best snap decision ever!

 

Pre-Flight Check # 3: You don’t have to be mad ^H^H^H enthusiastic to work here but it helps.

Whenever I solve an EMC problem I will generally do a little dance. It really floats my boat.

I’m lucky because I get to do what I love and people pay me. Most days I feel like I’ve won the lottery just for doing my day job.

If you don’t love the work (and it can be difficult) then its an excercise in frustration.

Try and follow what makes you want to dance in the middle of the lab. This is a fantastic lens for discovering what it is you are meant to be doing with your career.

 

General Skills: EMC is a Holistic Discipline

I spent the first 7 years of my electronics career working on…

  • power supply design
  • microcontroller coding
  • thermal CFD simulation and design
  • basic mechanical design
  • high speed digital design and test
  • system level architecture
  • cost sensitive design
  • project management

…before I became an EMC engineer. Before even realising I wanted to be an EMC engineer.

I still regularly use ALL these skills in my job as an EMC engineer.

Product design decisions made impact EMC performance.

EMC decisions impact product performance (and cost).

The two co-exist and cannot be separated.

Understanding the compromises of product design, the interaction between competing aspects (particularly cost!) is incredibly useful.

 

Go to the place least crowded / Leverage your existing skills

It might be that your team/employer/company has no EMC engineer. Take on that responsibility. This is what I ended up doing and now, 13 years later, I still love what I do.

Perhaps you have an EMC engineer colleague. Arrange to sit on their shoulder and talk to them. Ask lots of questions. Find out what area they don’t have time to work on or what problems they have. Work on that.

You are a member of an EMC team. Again, what areas do the team struggle with? What area consistently causes problems? No one is an expert on the finer points of widget calibration and the effects of temperature. Become that expert.

Find a niche (rhymes with quiche dammit) and fill it. You get to progress and inevitably find something else interesting to work on.

Follow your curiosity!

 

Get good at fixing EMC problems / make mistakes

Another fundamental truth of EMC is that There Will Be Problems.

Problems present a (usually) unique learning opportunity. Every problem I’ve solved has either taught me something or reinforced some piece of existing learning.

Spend a time in the test lab experimenting and getting an understanding of what works and what does not work.

All experiments are useful. Failed experiments or inconclusive data can help you refine your thinking.

This also leads on to mistakes. I make mistakes on a daily basis. They are usually small and easily correctable but sometimes they are bigger. Like the time I fried a piece of customers equipment by supplying 28V instead of 7.4V. Mistakes are hard teachers but you don’t forget the lesson in a hurry.

Importantly, people remember the mistake less than what you did to fix it. Own your mistakes.

 

Understand how HF current flows

In my opinion, this is the key to understanding EMC.

I recorded a presentation which might help your understanding but others have written about it before me and better (Henry Ott for instance).

Once you can visualise this you can understand the WHY behind so much of EMC.

 

Cultivate a Tolerance for Frustration

I would describe being an EMC engineer as alternately frustrating and elating.

You get better at dealing with the frustration of a problem and at solving it quicker.

Sometimes the scope of a problem is outside of your remit of available tools or skills to fix. Learn what you can and try and figure out a way forward.

 

Learn to automate

One of my favourite articles is Don’t Learn To Code, Learn To Automate.

EMC is no different to any other job, there will be repetitive tasks to perform.

Automating tests frees you up to work on other things and makes your work more consistent. Plus it gives you an opportunity to make a cup of tea whilst running a test. Maybe even a biscuit.

Automation doesn’t always go to plan or work out to be time efficient so pick your targets carefully.

 

Study Widely

Attend courses, webinars, lectures, presentations. Eventually some of it will sink in.

Sometimes you aren’t ready to grasp a piece of knowledge because you don’t have the existing framework for it to the idea to fit into.

Be wary of accepting everything at face value. Specific examples are sometimes presented without context or as globally applicable.

 

The learning never stops

I’m still trying to wrap my head around the intricacies of Power Distribution Network design, LabView coding for test automation and how antennas really work.

 

Share knowledge

Give a presentation to your colleagues about an EMC topic.

Explaining something complex to others in a simple fashion is the best marker as to how well you understand it.

I always spend lots of time on any talk I’m giving to try and make it as simple to understand as possible whilst still being useful.

 

Professional Accreditation

You may have the option of working towards accredited engineer status like the Chartered Engineer path through the IET here in the UK for example.

There are also the independent iNarte certifications which are particularly relevant for our field of work.

Some industry sectors or larger corporation might prefer you to have these qualifications. It certainly shows that you have achieved a certain level of competence and have been vetted to a certain extent by a 3rd party.

Find out what is expected or in your industry sector

I have no strong feelings either way on these professional qualifications. I investigated both whilst I was establishing Unit 3 Compliance and decided that I didn’t have the time to commit to them whilst I was setting up the business.

For me, there’s always something more impactful that I can be doing for my business than getting a piece of paper that might only make a small difference to one or two customers. I want to make a big difference for all my customers.

 

Connections and Groups

People to follow on LinkedIn

Groups on LinkedIn. Both of these are fairly active with some knowledgable members.

Other groups to join:

  • The IEEE EMC-PSTC email reflector is excellent with lots of good questions and answers on the subjects of EMC, safety and general compliance
  • IEEE EMC Society of UK and Ireland have bi yearly meetings
  • If you are in the UK, ICMA-TEL have a good email reflector with a diverse range of content including EMC, global market, safety, ROHS. Monthly meetings, mostly in the south of the UK.

 

Bonus: Copper tape is the scoundrel’s last resort

Useful as a diagnostic tool or emergency patch but not as a long term solution 😉

 

Fin.

Thanks for reading this far. If you have any ideas for what else could be included then drop me a mail.

That’s it from me. All the best on your journey.

.James

 

 

 

unit 3 compliance pocket debug probe

Pocket EMC Debug Probe

Hello 🙂

If you are reading this then hopefully you are a lucky owner of a Unit 3 Compliance Pocket EMC Debug Probe and love it so much it goes everywhere with you, in your toolkit, on your keyring, or in your shirt pocket. Dazzle your colleagues*, impress your manager and baffle your children by knowing exactly what to do with it.

* Note: Colleague be-dazzlement is not guaranteed. Some assembly required. May contain nuts.

unit 3 compliance pocket debug probe

 

About The Probe

Near field probing is a useful way of investigating the EMC characteristics of your system in a workbench environment. Most commonly these probes are used to identify problem areas of a design when a specific frequency has been spotted during emissions testing.

Building your own probe set is often a rite of passage for EMC engineers across the land but this board saves you the trouble.

The Pocket Probe has three sensors which I most commonly use for debugging problems with customer’s products. These are covered below in some more detail.

The probe works best with a spectrum analyser to look at the frequency components of the noise. However you can use it with a high speed digital oscilloscope and either look at the FFT of the captured waveform or just the amplitude data in the time domain.

Someone asked about the maximum frequency of this probe. I’ve not characterised it on a VNA and it will have resonances somewhere but it’s main use is as a method of finding the noisy area of the circuit.

 

Construction

If you have a bare board version then there is Some Assembly Required. You’ll need:

  • Edge mount RF connectors – the PCB is designed to take SMA edge mount connectors but you can solder a coaxial cable onto the pads at the expense of ease of swapping the connection between probes. Some of the PCBs have pads that are spaced slightly wider than a standard SMA connector but solder can be used to bridge the gap no problem.
  • For the 10:1 probe part you also need
    • A 450 ohm 0805 resistor – a 470R in parallel with a 10k gets you close enough for government work
    • A 100nF 0805 capacitor – useful DC protection for your spectrum analyser
    • A pin for the probe – an old component leg works as well as a piece of snipped off pin header

The construction is fairly self explanatory, the R and C go on the two component pads in the bottom right hand corner. The probe tip is soldered to the pad on the curved corner.

 

Magnetic Field Probe

This captures the magnetic field emanating from current loops. Orient the loop to match the geometry of the loop you wish to measure for maximum pickup. Spend some time experimenting with orientation position to see how it affects the received signal.

This is useful for

  • Picking up differential mode emissions in cable assemblies – separate the conductors slightly to maximise the effect
  • Finding noisy power switching converters on a PCB
  • Identifying current loops on PCBs

 

Electric Field Probe

At higher frequencies using the capacitive electric field probe often yields better results. This works really well when looking for “hot spots” of PCB radiation. Move the probe over the PCB whilst minimising the distance between the plate and the board. Look for any increase in the noise floor

Getting the probe close to the surface is the hardest part. If the geometry of the probe doesn’t suit the application don’t be afraid to modify it. Solder some copper foil or tape onto the end and wrap it in insulation tape. This gets you closer to the surface of the PCB without the risk of shorting things out. Remember, the closer the better.

 

High Bandwidth Probe

Once you’ve isolated the area of a circuit board that is noisy then direct probing of PCB traces comes into it’s own to identify the culprit.

The chances are most of the scope probes that you have on your bench have an 3dB bandwidth limit of between 100MHz to 350MHz. Radiated emissions problems most often manifest themselves at frequencies between 200 MHz and 1000 MHz. This limits the usefulness of a standard probe for EMC problem solving. HF probes are available but are expensive and less robust.

This is a super effective, robust, low cost version of those probes.

The equivalent circuit is simple, a 430 or 470 ohm resistor (ideally 450 ohm but this isn’t a standard value) and a 1nF C0G/NPO dielectric capacitor are in series with the probe tip. When connected to an instrument with a 50 ohm input impedance this makes a passive 10:1 probe which has a nice balance of high frequency performance and lower circuit loading effects. The resistor gives the 10:1 part and the capacitor prevents you from blowing up the input stage of your spectrum analyser if you are probing onto lines with DC content.

The 10:1 divide ratio is more about reducing the influence of the 50 ohm receiver input on the line being measured rather than being used for an accurate voltage measurement. Feel free to fine tune the resistor value on your board with 0.1% parts if that’s your thing!

It also gives you more repeatable measurements when doing A to B improvements to your circuit than using the other probes.

How to use it? Connect to your measurement instrument and start poking around the traces or pins in the “hot” area that you’ve identified using the above probes. Chances are you’ll narrow down the problem pretty quickly.

Why isn’t there a ground lead? The capacitance between the ground plane of the circuit board, your hand and the cable forms a virtual ground lead.

File the tip to a point if you like for probing onto fine pitch SMT pins.

 

Pre-Amplification

Sometimes the signals are just too small to measure and you need a little more gain. For the most part your problem signal should stand out but there are things you can do if you need more volume.

If your spectrum analyser has a built in pre-amplifier then turn that on – it’s sometimes under the amplitude menu tree. If you don’t have an internal amplifier then the LNA4ALL is a great external amp for this kind of work. It has a 50MHz to 4GHz bandwidth and is available with options to add ESD protection, operate it from 5V, or a case to mount it in. For only a few Euros it’s a great device.

Pair this LNA with a back to back SMA connector to enable mounting the amplifier directly to the probe board for ease of handling.

If your spectrum analyser has a Probe Power connector on the front you can tap power from this to run the probe. I made up a lead to use the PS/2 keyboard port on the front of my spectrum analyser for this same purpose.

 

Immunity

You can also connect the probe to an RF signal generator and use it to inject noise into circuits. This is a useful way to measure the immunity performance and identify susceptible circuits. Putting a 3-6dB attenuator in series (for the plate and loop probes) as a matching pad can reduce the effect of the probe and circuit impedance on the RF generator signal amplitude.

You can use this on any of the three probes on the board, give it a try to see what happens.

 

Recommended Accessories

 

Feedback

Please let me know how you get on with the probe, send me any design suggestions or ask any questions. I’ll be happy to help.

Happy probing!