Author Topic: The Trials and Rewards of EMI / EMC  (Read 23287 times)

0 Members and 2 Guests are viewing this topic.

Offline PChi

  • Frequent Contributor
  • **
  • Posts: 265
  • Country: gb
Re: The Trials and Rewards of EMI / EMC
« Reply #25 on: July 03, 2012, 08:39:12 am »
I agree with the other posters and don't think that a screened chamber without an RF absorbing lining is much use for absolute measurements. But it can find out at what frequencies the device under test is emitting.

The open air can then be used to give some idea of values. (possibly not in the UK at the moment though with all the rain). Relative measurements with equipment that has passed could help give an idea of pass and fail levels.
 

Offline NukerDoggieTopic starter

  • Contributor
  • Posts: 42
Re: The Trials and Rewards of EMI / EMC
« Reply #26 on: July 03, 2012, 02:57:04 pm »
PChi,

Yes, exactly - relative measurements are of great value to me at this stage, before my first trip to the cert lab.

Someday soon I hope to actually get my new biconical calibrated at the lab. But it's going to be of great value to me even before that.

True - my RF Coffin isn't big enough to separate the biconical from the DUT at enough of a distance to qualify as far-field only - I'm going to see a mishmash of far field and near field effects. Ok, but this test setup is still going to help me mitigate, iterate and mitigate my device before my trip to the lab.

I'm also building a set of near field E and H probes, and I have constructed a pretty good wideband RF preamp from a kit (from minikits in Australia) to use with the probes, so I'm setting up to do pretty effective amateur pre-pre-compliance work on my device.
 

Offline Neilm

  • Super Contributor
  • ***
  • Posts: 1562
  • Country: gb
Re: The Trials and Rewards of EMI / EMC
« Reply #27 on: July 03, 2012, 07:33:35 pm »
A "home brew" test might be good enough for doing comparative tests - seeing if what you have done has had any effect. However, it will not be good enough to document in an official test report - these have to have full traceability for test equipment. A good test house report will detail the make of each instrument, its serial number and when it was calibrated. Of course, if you go to a test house and they say it fails then you can do modifications and test whether that has had any effect at the frequency concerned.

You are basically hiring the test chamber, equipment, and an engineer for a day or half day and it isn't horrendously expensive. If you haggle/plead and are flexible you might get some deals when they are light on work.

For RF emissions, the main standards refer to an Open Air Test Site (OATS). I have to be honest - I don't know the medical requirements so maybe they don't.

Yours

Neil
Two things are infinite: the universe and human stupidity; and I'm not sure about the the universe. - Albert Einstein
Tesla referral code https://ts.la/neil53539
 

Offline Rufus

  • Super Contributor
  • ***
  • Posts: 2095
Re: The Trials and Rewards of EMI / EMC
« Reply #28 on: July 04, 2012, 01:52:08 am »
For RF emissions, the main standards refer to an Open Air Test Site (OATS). I have to be honest - I don't know the medical requirements so maybe they don't.

I have been involved (a little) with commercial, military, and medical compliance testing and have never seen any open air testing except for establishing compass safe distances (magnetic field generation). That was only to get away from magnetic fields from metalwork in the buildings.
 

Offline NukerDoggieTopic starter

  • Contributor
  • Posts: 42
Re: The Trials and Rewards of EMI / EMC
« Reply #29 on: July 04, 2012, 02:53:39 am »
Well, in general, with a medical device you're dealing with two federal agencies and their respective sets of regulations.

First - FCC regulations. These RF emissions standards are fundamentally aimed at preventing interference and device susceptibility to such interference.

Second - FDA Regulations. The FDA is charged with governing and maintaining the safety of medical devices, among other things. Part of that mandate relates to devices that emit radiation of any kind, whether ionizing or not, and whether such emissions are considered "safe" for humans.

Thus, the developers of the MRI had to prove by scientific measurements and by trials upon animals and humans that the RF and magnetic energies 'emitted' by the machine were "safe". Here, both radiation levels and duration of exposure enter the equation. So do risk vs benefit calculations. Thus, an X-ray does harm to the DNA, but the benefit often outweighs the risk. But soon the calculation over time for total permitted exposure increases the risk to a point that X-ray may be contraindicated for a particular patient - but that has nothing whatever to do with device approval by the FDA. That falls under the bodies that govern conduct of medical professionals.

In general, a medical device may intentionally irradiate the patient, aimed at producing some benefit. The mfg of the device has to prove to the FDA by scientific methods that such a benefit, and no unacceptable harm, accompanies the use of the device.

So, as such, the FDA does not maintain and enforce the same kinds of standards that the FCC does. FCC emissions standards are precise and unyielding. FDA standards are more embodied in principles, such as that central one mentioned in the previous paragraph.

So, using the MRI as an example again, the device could never pass FCC emissions standards unless it is enclosed in a very expensive Faraday room - and it always is so enclosed. But you can stand in the room near the MRI while your friend or spouse is 'in the tunnel' and, as long as you don't have steelies in your pocket, and there isn't a steel gas cylinder of some kind laying around unlashed, you'll be just fine!

So a medical device can 'get away with some emissions' that a non-medical device can't - as long as the emitting device is only used in a properly shielded room.

In the case of my medical device, I'm intentionally emitting infrared optical energy, but not intentionally emitting RF. I want to prevent RF emissions that are in violation of the FCC standards and, if I do so, I can be reasonably sure those emissions will also be safe for human exposure. However - the FDA isn't going to listen to such logic. It will insist I prove not only that the optical emissions are both safe and effective, but also that the RF emissions are safe. The only way I can prove the latter is to bring to bear on my case the longstanding science that such low levels of RF emissions in intimate contact with the skin have never been proven to be harmful. Then I have to conduct trials on humans and document the results.

A cert from a lab that my RF emissions are within FCC standards is going to be crucial in making my case to the FDA, but they won't accept that alone. Thus, in effect, the bar for medical devices is set quite high - this is as it should be.
 

Offline PChi

  • Frequent Contributor
  • **
  • Posts: 265
  • Country: gb
Re: The Trials and Rewards of EMI / EMC
« Reply #30 on: July 04, 2012, 08:40:18 am »
It's a good few years since I did any EMC testing so the standards have probably changed.
The emissions at 3 m were tested in an RF anechoic chamber.
The emissions at 10 m were performed in an Open Air Test Site. I vaguely remember that the Unit being tested was in a glass fibre hut. I always thought that the ground conductivity could influence the results a little and just the ambient emissions were a problem. I have only seen pictures of a 10 m RF anechoic chamber which must be incredibly expensive so an open air site for lower frequency testing is the only real world alternative.
 

Offline Neilm

  • Super Contributor
  • ***
  • Posts: 1562
  • Country: gb
Re: The Trials and Rewards of EMI / EMC
« Reply #31 on: July 04, 2012, 07:29:19 pm »
It's a good few years since I did any EMC testing so the standards have probably changed.
The emissions at 3 m were tested in an RF anechoic chamber.
The emissions at 10 m were performed in an Open Air Test Site. I vaguely remember that the Unit being tested was in a glass fibre hut. I always thought that the ground conductivity could influence the results a little and just the ambient emissions were a problem. I have only seen pictures of a 10 m RF anechoic chamber which must be incredibly expensive so an open air site for lower frequency testing is the only real world alternative.

The general CISPR standard for emissions does allow use of an anechoic chamber, but the results are then correlated to adjust for the differences between the chamber and the OATS.

@NukerDoggie - if you want to sell it outside the US you will have to ensure it passes the relevant tests for the target markets - CE marking for the EU for instance.  You might well find that the standards to meet there differ significantly between the States and the EU.

Neil   
Two things are infinite: the universe and human stupidity; and I'm not sure about the the universe. - Albert Einstein
Tesla referral code https://ts.la/neil53539
 

Offline NukerDoggieTopic starter

  • Contributor
  • Posts: 42
Re: The Trials and Rewards of EMI / EMC
« Reply #32 on: July 10, 2012, 02:00:43 am »
Here's a few pics of my homemade H-field probes, before I insulate them by repeated dipping in opaque-colored epoxy resin. These are 1cm, 2cm and 3cm dia., and constructed from RG-405 coax. I've installed ferrites to cover the range from below 40 Mhz to well over 300 Mhz, then covered them with heatshrink tubing to simulate handles. I'll dip the probes in epoxy from the loops up to the start of the 'handles'.

Can't wait to use these with a wideband RF preamp that's low-noise and 25db gain (cost me only $75 as a pre-built, pre-tested kit from MiniKits, AU) along with the Rigol DSA-815. I'll post pics of the finished probes and the near-field test setup for my medical DUT, and Rigol screen captures in about a week or so.
« Last Edit: July 15, 2012, 02:48:19 pm by NukerDoggie »
 

Offline Benno

  • Newbie
  • Posts: 4
  • Country: nl
Re: The Trials and Rewards of EMI / EMC
« Reply #33 on: August 31, 2012, 12:29:26 pm »
Hi NukerDoggie,

very nice to see all your efforts. I have been playing with in house measurements too, mainly to learn from my errors.

As said earlier, one thing you can have problems with are your cables. You can make a measurement adapter for that too very easy. I used a simple small clamp bought in a DIY shop and used the inner part of a clamp-on ferrite. Around one part of the ferrite I put a few windings and connected that to a coax and BNC.

It is not calibrated in any way, but it helps you to see the effect of changes you make. I use a TTI PSA1301T spectrum analyser at the moment for this, but love what I see about the Rigol. Attached is an image of my clamp.

I read in the replies of the youtube video about the biconical that you have drawings and instructions. Would you be willing to share them with me or post them here? You can contact me via email if you like via my profile.

Biggest problem with all this DIY is that you have no reference, but hey it is fun to play with and learn.

Benno.


 

Offline Benno

  • Newbie
  • Posts: 4
  • Country: nl
Re: The Trials and Rewards of EMI / EMC
« Reply #34 on: August 31, 2012, 12:35:25 pm »
A very usefull site also is design4EMC http://design-4-emc.com/.

He is a blogging EMC consultant and publishes a lot of usefull info. His EMC troubleshooting kit series is very worth reading an includes a lot of tips how to do things easy in the lab.
« Last Edit: August 31, 2012, 01:02:52 pm by Benno »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf