Service Bench Shielding

Watthour

Electron Rancher - JS3600
Like a lot of us girls, I use my computers as a resource when working on receivers and amps. I've found that when working in tuners, particularly on AM, I experience significant RFI which can skew the alignments. A CPU clocking at well over 1 GHz and its data bus clocking along at over 250 MHz seems to have quite an effect on the tuners. I've moved the nearest computer 6' from the bench, but still experience the phenomenon until it is shut off. Turning off the LED monitor seems to have no effect. Disconnecting the USB mouse and keyboard, and disconnecting the speakers also does nothing, thus I'm thinking it must be the PC itself. I set a grounded steel sheet between the PC and the workbench to no avail.

Has anyone found a method to shield computers effectively to eliminate this? I have two PCs in my work area and have to print manual sheets, then shut them off before undertaking an alignment. It's not the end of the world as we know it, but is inconvenient.
 
I think that shutting them off is your best option while you do sensitive lab work. You need a Faraday shield/cage to stop RFI. When I worked a Motorola Comm they had a copper screened room for pager test/alignment, those pagers were so sensitive. One job I had was to scotch brite the fingers on the doors to remove oxide build up. We used those pagers as our RF noses.
 
FWIW. Possibly your computer RFI is back feeding via the power supply into the power wiring of your work shop and radiating that way ?

Might be simple enough to purchase a quality surge/spike/filter multi socket metal cased ISOBAR unit by Tripp lite and plug your computer into that and see if that helps, but any power line filters used will be only as effective as the low impedance of your workshop`s power circuit`s ground, and might be helpful to power your device/tuner into a separate power filter as well.

I have some good results with split/clamshell type ferrite sleeves inserted right at the IEC power connector, and the Ethernet cable at the computer, as well.
I started chasing this computer RFI/EMI noise problems in the mid nineties, and they were much more RFI/EMI dirtier than they are now.

Also some computer power supplies are better filtered than others, especially more modern quality ones, because of stricter RFI/EMI regulations..
Here is a picture of my 5 kHz to 2,6 Gig. receiver suite with 6 home built non clear panel side case desk tops & 2 lap tops, all in a 12`X12` room, and after upgrading to LED monitors for the desk tops, and using the methods described above, except for noisy/RFI LED lighting in my house, my computers radiate very, very little RFI/EMI now.

Good luck Watthour

Kind regards, Billy FerrisIMG_0199.JPG
 
Thank you for the tips. I have several of the clamp-on ferrite loops that I can try. I have also discovered that the switching supplies in the LED/LCD monitors emit noise, as well as the LED lighting right over the bench. I may have to revisit the power line side. I have three circuits in the office/workshop, and may experiment with filtering the receptacle for the unit(s) being tested instead of all the other circuits in the building.

I learned long ago that I need to shut down my 15W FM transmitter in my garage/workshop when doing any adjustments since the GP antenna is only 40' from my office workbench.
 
Thank you for the tips. I have several of the clamp-on ferrite loops that I can try. I have also discovered that the switching supplies in the LED/LCD monitors emit noise, as well as the LED lighting right over the bench. I may have to revisit the power line side. I have three circuits in the office/workshop, and may experiment with filtering the receptacle for the unit(s) being tested instead of all the other circuits in the building.

I learned long ago that I need to shut down my 15W FM transmitter in my garage/workshop when doing any adjustments since the GP antenna is only 40' from my office workbench.

You`re welcome Watthour, and I wish you success in your pursuit.
These suggestions are considerably less expensive, and more logistically friendly than a effective home built Faraday cage, I would think..
Good luck Sir.
Kind regards, Billy Ferris
 
Well, the room where the work occurs is in my basement, double concrete walls separated by 6' of backfill on two sides, second skin of plywood under the floor joists and just above the drywall ceiling, 2x6 walls with plywood skinned on both sides for the remaining walls with drywall over that, and a solid core door. It was built to be a shelter room in case of severe weather, and I'm not seeing a good way to cage it now. Besides, I believe that most of the offending devices are right in the same room, so it may be fruitless.
 
and more logistically friendly than a effective home built Faraday cage
me too :)
You do not have to physically be in the Faraday cage, only the DUT does.
You have two issues, conducted and radiated emissions to contend with, each having different methods to suppress.
 
me too :)
You do not have to physically be in the Faraday cage, only the DUT does.
You have two issues, conducted and radiated emissions to contend with, each having different methods to suppress.

That's pretty clear, and the power side noise should be easier to address. I may have to acquire a 1 to 1.5 kVA shielded isolation transformer or supply with filters as Bill suggests for powering the DUT.

As for the radiated interference, I'll apply some of the clamp-on iron chokes to several of the cords, but the PC and LED monitor could prove interesting if the noise is not all due to the switching supply.

I've also noticed radiated noise from some of the test instruments, like one o-scope in particular, and a waveform generator, and I'm not sure it's all from power supplies. My bench is going to look pretty strange with a tin foil hat. The metal riser and shields separating the equipment from the work surface on my bench is already grounded, but it may not be very effective as an EMI/RFI barrier.

Workbench08.jpg


I've always been in the habit of isolating o-scopes from ground so I can test hot chassis equipment, but I may have to revisit that, too.

This probably wouldn't even me much of an issue if I wasn't so damned reliant upon a PC as a resource. Sam's manuals never caused this kind of interference, they just cluttered the work area.
 
But it is mainly the AM band that is the issue at hand so you can power off noise sources, print out what you need to do for the task at hand. If it was an all day event then more drastic measures would be needed.
For ultra critical applications, it comes back to magnetic ballasts and incandescent lamps. LED's in themselves are not the problem it is the efficient switch mode power supplies that drive them, but that can be fixed if you build your own LED power supply, you need is a step down transformer and a current regulator approach. A lot of monkey business :)
You can isolate your scope ground or isolated the DUT from the line using a isolation transformer or use probes designed for the application. You need differential probes for these line voltage applications.Lifting grounds are much cheaper.
Doing regulatory work is a challenging aspect of EE. I once had to make special ethernet AUI cables, using very expensive double shielded Beldon cable, to get our equipment to pass regulatory. Had to buy the whole spool to get the stuff we wanted. All the store bought cables failed as they were not built to the exact cable build methods as detailed in the IEEE 802.3 spec. It is a lot of trial and error and head scratching. So we use special cables to pass and the customers end up using the lousy cables that fail = stupid system.
 
Yes, AM is most greatly affected by the noise generated, but I have noticed it on FM, specifically at the lower part of the spectrum.

I have been printing the sheets I need, then powering down. Unfortunately, there always seems to be that ONE section I miss in my planning, and have to go to a different PC on the network to view/print more. The only "equipment" problem there is between my ears.

Interestingly, I have a no-ballast LED strip lamp above the bench which emits noise (RFI). I'm thinking it may be the current regulators for the LED segments (inside the tubes).

In short, I've been getting around the problem, but suspect there might be a way to have my cake and eat it, too.
 
I think that shutting them off is your best option while you do sensitive lab work. You need a Faraday shield/cage to stop RFI. When I worked a Motorola Comm they had a copper screened room for pager test/alignment, those pagers were so sensitive. One job I had was to scotch brite the fingers on the doors to remove oxide build up. We used those pagers as our RF noses.

Could not have said it better!

Thank you for the tips. I have several of the clamp-on ferrite loops that I can try. I have also discovered that the switching supplies in the LED/LCD monitors emit noise, as well as the LED lighting right over the bench. I may have to revisit the power line side. I have three circuits in the office/workshop, and may experiment with filtering the receptacle for the unit(s) being tested instead of all the other circuits in the building.

I learned long ago that I need to shut down my 15W FM transmitter in my garage/workshop when doing any adjustments since the GP antenna is only 40' from my office workbench.

Those clamp on ferrites have to be used correctly. Most power cords are not shielded, so you will need to put a few loops of cord through the ferrite AND it needs to be the right MIX for the frequencies of interference. Take a quick look over at Palomar Engineers for a crash course. Since the AM band is getting hammered, it's probably the pc power supply - most have relatively low switch frequencies. You're lucky that your monitor isn't causing RFI - I've had trouble with a number of them.

Good luck
 
Thanks for the input. My LED monitor is creating RFI (Post #4), and so is the LED lamp over the bench, although to a lesser extent (distance?). I'll study that Palomar information to see if anything sticks.
 
I have the computer in a whole different part of the house, and have to shut them down to get rid of RFI. I also located several LED replacement bulbs which cause issues. AM is getting hard to use, there's just so many sources of interference working against it.
 
After tossing a Kenny with a dead front end on the bench for service I've now verified that the 48" LED strip over my bench is affecting both AM and FM, although FM to a lesser extent. It seems to be broadcasting, not line-side EMI or switching noise. I can actually see it affecting the signal strength.

I may have to "de-convert" the fixture back to fluorescent. I have a pair of fluorescent magnifier lamps on either end of the bench, still using the magenitic ballasts, and they have no such effect.
 
After tossing a Kenny with a dead front end on the bench for service I've now verified that the 48" LED strip over my bench is affecting both AM and FM, although FM to a lesser extent. It seems to be broadcasting, not line-side EMI or switching noise. I can actually see it affecting the signal strength.

I may have to "de-convert" the fixture back to fluorescent. I have a pair of fluorescent magnifier lamps on either end of the bench, still using the magenitic ballasts, and they have no such effect.

Did you try a different power supply for the LEDs? I have one of those LED strips, it runs off 12V. Try a linear supply, or a motorcycle or car battery (if you have one sitting around).

The inexpensive switching supplies can give off a lot of noise, and the long LED strip broadcasts it on the DC side, as does the AC wiring on the other end. I would not bother trying to filter it, simply replace it with something more quiet.

Some computer monitors can also be a source of noise, even when "off". A power strip with a master switch can help for that. I had a CFL bulb at my work bench, it was noisy as heck. Replaced it with an old school incandescent bulb, all was good. I also have battery driven LED flashlights for detailed inspection work inside tuner boxes - never enough light in there when inspecting traces for problems, etc.
 
The LED fixture is a converted 48" fluorescent, ballast removed, line power to the ends of the LED "tubes" which presumably have their inherent current limiters. I strongly suspect they are indeed broadcasting a psuedo-square wave in DC, and well as affecting the line side to some extent.

It's an ongoing exercise.
 
Sometimes it's down to the power supply.

I'm a ham. When I'm active...In active on HF/shortwave. If you thought your PC was bad on AM.... everything screws with the shortwave bands.

It's even worse when you consider my radio is connected to my computer for digital mode stuff.

Despite that...I have zero noise from my PC. None.

I think it's down to good grounding and a good power supply. Don't use your house ground as an RF ground.
 
I'm curious how that works (separate RF ground*), since all the wiring codes, of
course, specifically want all (antenna) grounds tied to the house. I dx the AM &
shortwave bands, so noise floor is a constant concern.
*isolation transformer/s?
 
I'm curious how that works (separate RF ground*), since all the wiring codes, of
course, specifically want all (antenna) grounds tied to the house. I dx the AM &
shortwave bands, so noise floor is a constant concern.
*isolation transformer/s?

It's best understood reading antenna theory and trying to understand it...because the idea applies to basically "monopole" antennas. But there are cases where the grounding you do to satisfy electrical code isn't what you want for a ground that's part of the antenna system. Like if I had a vertical antenna; I would need a "groundplane" under the antenna and extending outward. I can make this by using lengths of wires extending from the base of the antenna. They lay on the ground, but they aren't connected to the same ground as the rest of the electrical system...or at least they shouldn't be as that can cause common-mode current in the coax shield. So typically have a balun or some kind of 1:1 isolator between the groundplane of the antenna and the monopole element. The coax-run that goes in to your house runs through a lightning arrestor.

You really have to want to work with antennas to study it enough to make sense...it's a LOT more important for transmitting than receiving; but not that it doesn't help receiving. But you also have to remember about stuff like impedance. I've built antennas that to an ohm-meter would show a dead short at DC potential; but at 144MHz it's not. I..like you...wondered the same thing about how RF vs DC ground works when I started.

Anyway...noise floor can come down to local conditions as much as it can the antenna system. My first antenna was a LOT noisier than my second and third antennas; which are center-fed doublet/dipoles with twin-lead feedline.
 
Back
Top Bottom