What cheap EMC pre-complience eqipment to get for ~3500$

[jyrgen]

Hi, at my work we are looking into buying equipment to do emc pre-complience. We do just 1 or 2 new products every year so we have a limited budget (at the moment around 3500$). Most of our products concist of a small electrical motor and a circuit board, maximum 15x15x15cm (6x6x6inch) in size. Below I will list the standards and measurements that we have got from an EMC lab on several products. The goal is to cover as many of the measurements as possible, so we have an idea if the products would pass or not. I will then list what we have thought of to buy. I should also mention that I don't want to build equipment on my own unless it is easy and not time consuming. Any feedback on this would be useful.

Standards:

• EN 61000-6-1 Immunity for residential, commercial and light-industrial environments
• EN 61000-6-2 Immunity for industrial environments
• EN 61000-6-3 Emission standard for residential, commercial and light-industrial environments

Measurements:

• Radiated RF emission 30MHz-1GHz
• Conducted RF emission 150KHz-30MHz
• Radiated RF immunity 80-2700MHz
• Conducted RF immunity 150KHz-80MHz
• ESD
• EFT/Burst
• Surge
• Power frequency magnetic field
• Voltage dips & interruptions

Radiated RF emission 30MHz-1GHz
Rigol EMC bundle (DSA815 spectrum analyzer, near field probes and S1210 EMI software)

Conducted RF emission 150KHz-30MHz
In this category we would like to measure:

• 230V mains powered product
• 24VAC (isolated transformer to mains)
• 24VDC powered product
• Low voltage analog and digital I/O

Equipment below, not sure if 24VAC and I/O ports can be measured with these?

• 2pcs Tekbox TBOH01 5UH LISN (DC power port)

• Tekbox TBLC08 50UH LISN (230V mains port)

[/list]


Radiated RF immunity 80-2700MHz
Tekbox modulated amplifier. Would the TEM cell be a better option?


Conducted RF immunity 150KHz-80MHz
According to Tekbox manual for TBOH01 5UH LISN it could be used to feed RF noise as well.

ESD
This is pretty cheap to rent when needed.

EFT/Burst
No equpment in price range found.

Surge
No equpment in price range found.

Power frequency magnetic field
No equpment in price range found.

Voltage dips & interruptions
No equpment in price range found.

[Neilm]

Radiated immunity testing is hard to do on the cheap - you need a screened area to you don't break the law on radio jamming. When you do have a screened area, you then have to work out the field strength in the area which means haing something to measure it. You also have to consider what reflections are happening in the area, and are you over testing.

ESD requires very careful setup to get reasonable and repeatable results. I did once try with the coupling plans not correctly setup and the results did not match what the correct setup gave.

If your product connects to the mains you are unlikely to get anything that will cover burst and surge.

Radiated emissions are a guide to how your equipment will respond when immunity testing (but not perfect). This can be done in the near field with a spectrum analyser and a good set of probes. I have seen platforms that you would rest the PCB and it had an array of probes in it. The analyser was connected to each in turn and it measured the board performance showing the user what part of the board radiated at what frequency. It then did some maths and made a guess at how it would perform in far field tests. I didn't have much of a play with it, but it showed matching peaks in the same place as my test results. The only problem is if there are cables involved this will not work - the cables could be the antenna.

I would have a look for second hand equipment - equipment hire companies often sell old hire equipment and you can be reasonably sure it will be both safe and working.

[nctnico]

Emissions and immunity are almost impossible hard to test without a real test chamber so you are left with make-do solutions. Some people use small plastic bags with coins to generate broad spectrum pulses. For crude ESD testing I use a battery powered gas lighter with a ground lead. Be sure to keep other equipment away though when doing ESD testing. It is a useful tool if a product fails ESD and needs to be improved.

[SvanGool]

Your test-institute might rent you their facility and equipment for an hour to do the critical (suspected) parts of the major tests, with or without an engineer. That is what I did several times in the past.
If well prepared, you can do quite some tests for a part of your budget. I agree with the others that it will be very difficult to create e.g. your own anechoic RF chamber.

[mikeselectricstuff]

EFT/Burst
No equpment in price range found.

Could potentially be improvised - all you need is a spark discharge to generate the fast transient, teh issue is getting a rough calibration. Should be doable fairly easily.

Surge
No equpment in price range found.

Big capacitor

Power frequency magnetic field
No equpment in price range found.

DIY Generating & measuring this shouldn't be inherently expensive - coils, GMR sensors etc.

Voltage dips & interruptions
No equpment in price range found.

Depending on power level required, should be fairly amenable to improvisation. Maybe some SSRs for interruptions, plus some resistors for dips?

[T3sl4co1l]

I have some ideas about EFT/ESD generation, though I need to do a lot more simulation and experimentation.  If it works out, it'll be pretty cheap, which is cool!

Tim

[coppice]

I have some ideas about EFT/ESD generation, though I need to do a lot more simulation and experimentation.  If it works out, it'll be pretty cheap, which is cool!

Tim

Most commercial ESD testers still use the simple mechanically shuttled capacitor approach. It needs some careful engineering to make sure it stays safe, but it's hardly advanced technology.

It's easy to generate a few EFT pulses, and get the rise and fall times in spec. It might be a bit more of a challenge to generate the 2.5k impulses per second streams to spec. I've never looked at the techniques commercial EFT testers use.

Surge is also a low tech problem. Surge testers are little more than a low power high voltage supply charging a correctly sized cap. The surge is simply a matter of connecting the cap to the UUT?

[T3sl4co1l]

It's easy to generate a few EFT pulses, and get the rise and fall times in spec. It might be a bit more of a challenge to generate the 2.5k impulses per second streams to spec. I've never looked at the techniques commercial EFT testers use.

I'm guessing hydrogen thyratron, but who knows 

The standard also supports 100kHz pulse rep.  Not gonna pull that off with an SCR.

Surge is also a low tech problem. Surge testers are little more than a low power high voltage supply charging a correctly sized cap. The surge is simply a matter of connecting the cap to the UUT?

Sorta.  You can estimate the RLC network parameters from the waveform.  Speed is slow enough you can switch it with SCRs, and the impedance is about right.  (You do need a moderately beefy SCR stack to handle, say, 2kV and 1kA, with a lifetime hopefully of thousands of cycles.)  Coupling can be tricky: a transformer or inductor that can handle that much flux (i.e. ~kV for tens of us) is quite large, and has to be designed to supply load current with acceptable losses.

Tim

[KE5FX]

Problem?

[T3sl4co1l]

I'll take it, if you can get me a calibration certificate on it.

Tim

[PartialDischarge]

For ESD testing nothing beats a piezoelectric kitchen lighter. If an equipment passes this test, it'll do fine in the lab, I know from experience

Also a totally non-certified equipment to take into account is this, a bit harsh actually. Watch the phones get toasted at the end

[Tomorokoshi]

Other methods for totally non-calibrated, relatively on-the-cheap testing for radiated immunity and radiated emissions:

1. A high-power walkie-talkie radio. Something like this:
http://www.motorolasolutions.com/en_us/products/two-way-radios/mototrbo/portable-radios/cp200d.html#tabproductinfo

Note: it may very well have higher fields than would be done in the standard test. However, that may expose weak areas.

2. A good multi-band radio, covering AM, FM, SW, etc. Tune in and listed for obvious sources of noise.

I've worked on products that passed the radiated immunity test at industrial levels (10 V/m), but failed in the field, because they used those Motorola radios in front of the machine while working on it.

[nctnico]

I've worked on products that passed the radiated immunity test at industrial levels (10 V/m), but failed in the field, because they used those Motorola radios in front of the machine while working on it.

That happened to me too. Not everything out there adheres to the emission limits.

[mikeselectricstuff]

I've worked on products that passed the radiated immunity test at industrial levels (10 V/m), but failed in the field, because they used those Motorola radios in front of the machine while working on it.

That happened to me too. Not everything out there adheres to the emission limits.

There is a common misconception that emission and immunity limits have anything to do with each other, which they don't, as the limits are orders of magnitude apart.
The "missing link" between them is radio equipment.
Emissions limits are to stop interference to radio reception, Immunity limits are to prevent interference from local intentional radiation from transmitters like phones, handheld radios etc.

If your product's emissions cause immunity issues with something else then you have a very serious problem which no amount of shielding or ferrite is going to fix!

[nctnico]

I should have been more clear: some non-radio equipment can emit a lot due to wear and there won't be much you can do about it without upsetting the (end) customer because it is always the new equipment which is faulty if it doesn't work.

[T3sl4co1l]

If your product's emissions cause immunity issues with something else then you have a very serious problem which no amount of shielding or ferrite is going to fix!

Well, no.  Try switching rectified 480V in 50ns some time.

Shielding does wonderful things.  I once designed a machine with LVDS interfaces to driver boards... such a poor choice, but with some judicious grounding and cable routing choices, it worked.

Tim

[KE5FX]

Shielding does wonderful things.  I once designed a machine with LVDS interfaces to driver boards... such a poor choice, but with some judicious grounding and cable routing choices, it worked.

Tim

What's wrong with an LVDS interface?

[T3sl4co1l]

Well, the 3V common mode range, for one.

It's not exactly noise immune, in an industrial sense.

3.3V LVDS lines mere inches away from 700V switching edges, no problem right?

Tim

[KE5FX]

Well, the 3V common mode range, for one.

It's not exactly noise immune, in an industrial sense.

3.3V LVDS lines mere inches away from 700V switching edges, no problem right?

Tim

(Shrug) Not uncommon to see LVDS used in the vicinity of LCD backlight drivers, which are noisy enough.  I don't know that you'd have any better luck with USB or SATA/PCIe or any other conventional bus, in a situation where LVDS qua LVDS is failing to work reliably.

In retrospect, if you had it to do over again, what kind of bus should you have used in that application?

[T3sl4co1l]

I would've used RS-422/485 transceivers, probably with isolation.  (I would've used that to begin with, but for whatever reason, LVDS was the order from on high...  )

Tim

[jyrgen]

Thanks for your feedback. So to summerize the opinion so far:

• Emission and immunity seams to hard to test without a test chamer. In that case I should focus on other tests and not buy the equipment I mentioned in the initial post.
• There are a bunch of ideas on how to build for example EFT/surge test. Even though it would be fun, As I mentioned in the post I don't want to build these things because I don't have time with it. If it takes a couple of hours that is OK but not more.
• The ideas with making equipment with for example high power walkie talkei radio does not feel to appealing. It does not feel repeatable and to determine how much it will radiate.
• Is there any used equipment that could do any of the tests that are cheap enough?

   

[mikeselectricstuff]

Have you looked at renting?

Also, you can usually book time at an EMC house to do informal testing,  where you have access to all the gear and an engineer to do whatever tests you want.
 
If you are well organised, you can identify, test (and tweak and retest) critical areas on multiple products in a day, so the per-unit cost can be quite low. It's a while since I did this but from memory it was a few hundred pounds for half a day.

[coppice]

Renting can be reasonably priced for ESD, EFT and surge testing. Anything involving an RF chamber is usually expensive and hard to find. Even in a place like Shenzhen there are very few places that will rent out RF test chambers

[nctnico]

It is not just an RF chamber. All tests require not only the right equipment but also the right setup (shielding, non conductive tables, etc, etc). If you want to make accurate measurements without a huge investment then there is no alternative than to rent a lab for a couple of hours.

[coppice]

It is not just an RF chamber. All tests require not only the right equipment but also the right setup (shielding, non conductive tables, etc, etc). If you want to make accurate measurements without a huge investment then there is no alternative than to rent a lab for a couple of hours.

I've never seen anyone rent out an RF chamber without the whole package or equipment and personnel needed for the standard suites of tests..

[nctnico]

It is not just an RF chamber. All tests require not only the right equipment but also the right setup (shielding, non conductive tables, etc, etc). If you want to make accurate measurements without a huge investment then there is no alternative than to rent a lab for a couple of hours.

I've never seen anyone rent out an RF chamber without the whole package or equipment and personnel needed for the standard suites of tests..

I wasn't writing about RF (emissions/immunity). EFT, Surge, ESD, conducted emissions all need very specific physical setups and these can also change when the requirements are updated.

[coppice]

EFT, Surge, ESD, conducted emissions all need very specific physical setups and these can also change when the requirements are updated.

True, but the fact that these setups are so specific and well defined in standards means they are also trivial to reproduce. Very little thought is required. You just follow the formulae.

[T3sl4co1l]

It is not just an RF chamber. All tests require not only the right equipment but also the right setup (shielding, non conductive tables, etc, etc). If you want to make accurate measurements without a huge investment then there is no alternative than to rent a lab for a couple of hours.

I've never seen anyone rent out an RF chamber without the whole package or equipment and personnel needed for the standard suites of tests..

I would suppose, if you went to a testing lab and requested a custom quote for chamber time only, no support, you might get a good deal.    Or laughed off. 

Low frequency (conducted) and transient stuff is easy enough to test with a ground plane.  You will easily see if there are nearby sources of interference (radio stations, maybe some particularly obnoxious power supplies?), which you'll have to subtract from your test.

Radiated can be done with an open-air test site, but you have to subtract even more ambient sources (potentially).  Or carefully select a site, and antenna direction (assuming you have sufficiently directional antennas), to get it quiet enough to actually perform the test.  And you need a good enough antenna / selection of antennas (I usually see conical dipole for 30-200MHz, log periodic for 200-1000MHz).  This definitely makes it more challenging to do radiated emissions outside of a test lab.  (And, not having a way to calibrate, say, 3 V/m emitted field intensity, and avoiding beaming that at just any radio user that happens to be nearby, is even more tricky!)

Tim

[coppice]

Low frequency (conducted) and transient stuff is easy enough to test with a ground plane.  You will easily see if there are nearby sources of interference (radio stations, maybe some particularly obnoxious power supplies?), which you'll have to subtract from your test.

Radiated can be done with an open-air test site, but you have to subtract even more ambient sources (potentially).  Or carefully select a site, and antenna direction (assuming you have sufficiently directional antennas), to get it quiet enough to actually perform the test.  And you need a good enough antenna / selection of antennas (I usually see conical dipole for 30-200MHz, log periodic for 200-1000MHz).  This definitely makes it more challenging to do radiated emissions outside of a test lab.  (And, not having a way to calibrate, say, 3 V/m emitted field intensity, and avoiding beaming that at just any radio user that happens to be nearby, is even more tricky!)

We used to do a lot of rough emission testing with the UUT on a table in the middle of a field, with pretty good results. Its RF susceptibility testing that is the biggest problem. Pumping out RF between 30MHz and 1GHz to achieve 30V/m or more at the UUT is going to annoy somebody unless you have a big field on the Montana scale of things.