Author Topic: Mains-Powered Equipment Safety Plan  (Read 391 times)

0 Members and 1 Guest are viewing this topic.

Offline amosborne

  • Contributor
  • Posts: 7
  • Country: us
Mains-Powered Equipment Safety Plan
« on: April 19, 2021, 04:50:01 pm »
Preamble

I have some test equipment (an HP6253a, for example) that I'd like to probe around inside while powered on in order to assist debugging. Since these equipment are powered by mains (US), this activity is inherently dangerous and can kill you under the right circumstances. A while back I took a high-voltage safety training course through my employer; with that training in mind as well as my other readings around this forum and the internet I have written down the following safety plan. Would the kinds folks on the forum take a look at my safety plan and let me know if I've missed something important?

Safety Plan

Safety is achieved during any activity through both passive and active safety measures. Passive safety measures should be considered a last resort backup should any active safety measures fail.

Passive Safety Measures

A test setup diagram is provided as an image below. Mains power is received by a 1:1 isolation transformer which is properly rated for the amount of power the device under test requires. The isolation transformer provides a single isolated output through a cheater plug with the third wire ground lifted to isolate the device under test from earth ground. The device under test shown is a notional example of a power supply; in this case both the internal circuitry and the metal chassis are floating. The device under test may or may not refer the internal circuitry to the metal chassis; in this diagram they are floating with respect to each other.

This test setup allows for a single fault without the risk of an electric shock from the device under test. A person inadvertently touching a single point of the internal circuitry will not close a loop for current to flow and therefore an electric shock will not occur. If the metal chassis is floating with respect to the internal circuitry then this test setup is double-fault tolerant in the specific case of a person touching both the metal chassis and a single point of the internal circuitry.

Additional passive safety measures could be taken but are not required, such as using properly rated insulating electrical work gloves or standing on a properly rated electrical isolation mat. Different applications or test environments may call for different or additional passive safety measures.

If a person is wearing an ESD wrist-strap, it must have a 1MOhm-10MOhm bleed resistor installed (rather than a direct short to earth ground).

Active Safety Measures

  • Never work on a live piece of hardware without fully understanding it, the dangers it presents, and where those dangers lie. Even while powered off, the internal circuitry may contain capacitors which hold dangerous level of charge for long periods of time.
  • Always work on live equipment with one hand in your pocket. A single hand can still get shocked if two points in the internal circuitry are touched simultaneously, but at least the current is through the hand rather than the heart. A piece of equipment is live whenever it is plugged into mains.
  • Always verify by test that there is electrical isolation between isolated-live to earth, isolated-neutral to earth, and chassis to earth. Never reference the device under test to earth ground; this means no single-ended oscilloscope probes.
  • Always keep a clean and organized workspace. Always work slowly and deliberately.
  • Never work alone.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 2168
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #1 on: April 19, 2021, 05:23:24 pm »
I hope you aren't easily offended!  I disagree with a lot of that.  Simple rules rarely solve complex problems.  The only part I completely agree with is understanding your DUT--and how various failures might present additional dangers. 

First, I'm generally a proponent of isolation transformers and sometimes I get flak from other members for it for various reasons.  However, I would never recommend using them in all cases (or even as a default) and certainly wouldn't recommend lifting the ground on a routine basis.  I disagree with the notion that you should not reference mains powered devices to earth--in fact grounding them in the same manner as they normally would be should be your default.  This is especially true for very high voltage situations where trying to isolate things can result in inadvertently creating extreme dangers.

For most potentially hazardous measurements, a 'hands-off' approach is usually the safest, with insulating gloves a good alternative.  Keep in mind that an electric shock is not the only hazard present when working on powered electronics.  Understanding how much energy is present is one important thing to consider--one of the features of service-type isolation transformers is that they generally limit the current to much less than what you have at a typical wall socket.  If you don't need isolation for any particular purpose, you might be served better by a circuit consisting of a GFI and the smallest fast-blow fuse that works for your DUT.

Also, I always work alone because it minimizes distractions, although occasionally I will wait until someone is present for very specific things.  As for hand-in-pocket, why are you touching things?  If there's that much energy, get appropriate test leads that can be connected before the DUT is energized.  If you do get a shock, it probably won't kill you--but you may smash your head because you can't get your hand out of your pocket quickly enough to break your fall.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline TimNJ

  • Super Contributor
  • ***
  • Posts: 1267
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #2 on: April 19, 2021, 07:46:33 pm »
If there's that much energy, get appropriate test leads that can be connected before the DUT is energized...

Although sometimes it's much more convenient to poke around at a live circuit with test leads, probes, etc., I agree that one of the best things you can do is to have your body completely detached from the DUT. Figure out how to safely de-energize the circuit (i.e. bleed off big caps) and once de-energized, attach your probes, stand back, turn it on, and observe. Besides being safer, you're less likely to make a dumb probe slip mistake, which may be even more important when working on old test equipment that might have some un-obtanium parts.
 

Offline WattsThat

  • Frequent Contributor
  • **
  • Posts: 456
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #3 on: April 19, 2021, 07:57:41 pm »
What was the voltage level of this “high voltage” safety training that you received? I’m just trying to get some perspective on why you think any of the stated procedure would be applicable.
 

Offline amosborne

  • Contributor
  • Posts: 7
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #4 on: April 19, 2021, 08:19:31 pm »
I'm not easily offended - I did post this here for review after all. The high-voltage training I received covered from 50V up to maybe 500V, if I recall correctly. Certainly not kV range.

I agree that the hands-off approach is best and that is my first line of attack. But there are definitely times when poking around with a multimeter while the circuit is live is certainly more convenient, as long as my own safety can be maintained. Maybe the recommendation is just do everything you can to never even have to get close - I can work with that if that is the best practice.

I have definitely seen past discussion on similar topics about lifting the ground versus using a GFI, and I don't think there was ever a conclusive statement one way or the other. Using a GFI seems like it would increase the risk that you get shocked... but in the event of a shock would potentially reduce the damage. I went back and forth on this on myself... I could probably still be convinced for the other approach.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 14487
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #5 on: April 19, 2021, 09:02:51 pm »
I can think of no negative of using a GFCI to power the DUT. I can also not think of any good reason to lift the ground. Ground is there to protect you from a fault condition. Don't over think this, the most important thing is to practice safe habits, don't work while you're tired or distracted and always maintain situational awareness. Don't worry too much about following a bunch of specific rules or procedures.
 

Offline penfold

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: gb
Re: Mains-Powered Equipment Safety Plan
« Reply #6 on: April 19, 2021, 09:04:36 pm »
Never work on a live piece of hardware without fully understanding it, the dangers it presents, and where those dangers lie. Even while powered off, the internal circuitry may contain capacitors which hold dangerous level of charge for long periods of time.

Fair enough to assess as many risks as possible and find some mitigation for each... but please never assume you fully understand the risks or nature of the equipment because the one you forget is the one that'll get you ;)

Another thing is to use some insulated sheeting (I believe they make a heavy-gauge PVC for that purpose), to selectively insulate anything you don't want to touch or slip the probe onto, use it as a cover if you leave it DUT unattended or a mat to keep the DUT of the work surface.

As far as isolating transformers go, I'm generally an advocate. Unless you're working in an earth free environment (nothing "earthed" within reach of your bench.. including oscilloscopes) I would avoid floating metalwork and connect it to earth: a suitable safety isolating transformer shouldn't fail to a dangerous condition so shouldn't need the double fault tolerance of disconnecting the earth (so long as it is "earth" earth and not in any way connected to the transformer secondary). GFI/RCDs definitely.

Any test equipment you use to prove a circuit is of a safe voltage to touch should be tested before and after you do your check.
 

Offline amosborne

  • Contributor
  • Posts: 7
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #7 on: April 20, 2021, 01:28:09 am »
Thanks everyone this is good feedback. It does seem like I overthought the lifted ground.

I'm still a little mystified how a GFI and an isolation transformer can coexist. If the secondary neutral is not tied to earth, surely the GFI will have no way to detect a fault?

Unless you just mean a GFI on the primary side, simply as a best practice for working with lab equipment in general.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 2168
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #8 on: April 20, 2021, 01:58:26 am »
Thanks everyone this is good feedback. It does seem like I overthought the lifted ground.

I'm still a little mystified how a GFI and an isolation transformer can coexist. If the secondary neutral is not tied to earth, surely the GFI will have no way to detect a fault?

Unless you just mean a GFI on the primary side, simply as a best practice for working with lab equipment in general.

In theory you could have a GFI after the isolation transformer and it could function for certain faults, but it really isn't all that helpful overall and would give a false sense of security more than anything else.  A GFI before the transformer is obviously useless.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 14487
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #9 on: April 20, 2021, 02:21:56 am »
A GFCI will provide some degree of protection when you're not using an isolation transformer. An isolation transformer will provide protection under a different set of circumstances. I do not typically use either one when I'm troubleshooting something unless I'm dealing with specific circumstances, such as a "hot chassis" radio, TV or need to poke at the primary side of a SMPS.
 

Offline penfold

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: gb
Re: Mains-Powered Equipment Safety Plan
« Reply #10 on: April 20, 2021, 01:13:53 pm »
I'm still a little mystified how a GFI and an isolation transformer can coexist. If the secondary neutral is not tied to earth, surely the GFI will have no way to detect a fault?

It's a bit academic but on the primary side of the transformer a GFI will trip at a much lower fault current should the iso transformer fail with some impedance to earth (when there's an earthed screen between primary and secondary), so rather than blowing a fuse at some fault current it will trip a little more gently.

On the secondary side, I agree its somewhat pointless to use one on the secondary side (since you're removing any significant conductance between L+N to earth), but if some fault exists in your DUT (may accidentally connect something wrong or slip with the probe) that gives conductivity between isolated-L+N to earth, then it gives you protection against possible double fault conditions... an insulation monitoring device would ideal but costly.
 

Offline amosborne

  • Contributor
  • Posts: 7
  • Country: us
Re: Mains-Powered Equipment Safety Plan
« Reply #11 on: April 20, 2021, 03:31:16 pm »
I'm still a little mystified how a GFI and an isolation transformer can coexist. If the secondary neutral is not tied to earth, surely the GFI will have no way to detect a fault?

On the secondary side, I agree its somewhat pointless to use one on the secondary side (since you're removing any significant conductance between L+N to earth), but if some fault exists in your DUT (may accidentally connect something wrong or slip with the probe) that gives conductivity between isolated-L+N to earth, then it gives you protection against possible double fault conditions... an insulation monitoring device would ideal but costly.

This makes sense to me. I did do some reading about insulation monitoring systems like they use in hospitals; I failed to find anything equivalent for less industrial applications. Business idea?

Anyways, to summarize I think there were basically two options for me to pick from:
  • Use an isolation transformer and do not use a GFI (it would only be helpful in a specific double fault scenario). Keep the metal work earth grounded, not lifted (as in my original plan).
  • Use a GFI and do not use an isolation transformer.

In either case, I should endeavor to be as hands off as possible. i.e. Adjust probes while powered off, then stand back and observe while powering on.

As for which option to pick of the two, I do not know if there is any other criteria to consider. I'm personally leaning towards just going with GFI, because a GFI outlet is certainly cheaper than a new transformer.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf