Matador

Line Isolation transformer ground
« on: May 16, 2017, 12:54:52 PM »
I have a 500VA AC line isolation transformer I'd like to integrate into my lab bench setup, mostly for testing suspect guitar amps.  I'll connect my variac to its output as well, and mount it in a spare metal case. My question is how to treat the ground on the isolated side of the transformer.

I see two ways of doing this:
A) Bond the ground of the output side to the chassis of the isolation transformer, which is bonded to the ground of the incoming line connection (via the metal chassis).

This seems to be a popular (maybe required?) way of doing it, as it ensures that equipment downstream with a ground-bonded chassis will have the a similar chassis potential as equipment that is not running through the isolated secondary.

Bill Whitlock's grounding article seems to show this very connection:



However this tickled a memory I had when working with boat shore power systems, where exposed metalwork can be galvanically stripped away by DC currents flowing in salt water when boats shared a common safety ground.  Sure enough, for isolation transformers the recommendation seems to be to bond the earth and neutral wires on the secondary, much like what would happen in the panel on a residential installation:



I realize that galvanic isolation is the requirement for boats, in that the poor sod who parks his aluminum inboard-outboard at the docks doesn't want to see it dissolved away in 6 months, but is there any advantage for testing / noise in this configuration?


JohnRoberts

Re: Line Isolation transformer ground
« Reply #1 on: May 16, 2017, 01:50:53 PM »
I am not an electrician but my understanding is that safety ground must be bonded to neutral only one time (at the panel).

I think there is a variant for sub-panels where neutral and safety ground are bonded at the sub panel.

It is generally not code to connect ground to neutral any distance from the panel  since an open neutral wire could energize the ground.

JR
John Roberts
http://circularscience.com
Tune it, or don't play it...

[silent:arts]

Re: Line Isolation transformer ground
« Reply #2 on: May 16, 2017, 03:04:55 PM »
I am not an electrician either but all isolation transformers in lab benches I know are earth free.
SSLtech: No bypass, I don't believe in it.
Gus: I think this build might have problems with DIY builders that don't know things like ohms law.

Matador

Re: Line Isolation transformer ground
« Reply #3 on: May 16, 2017, 05:08:41 PM »
I am not an electrician but my understanding is that safety ground must be bonded to neutral only one time (at the panel).


Except in this case we essentially have another, completely separate incoming line, that happens to swing the same potential as that at our panel.  The secondary becomes a new hot and neutral, like what comes from the power pole in my backyard (well, one phase of it to be exact), right?

I am not an electrician either but all isolation transformers in lab benches I know are earth free.

Do you know what 'earth free' means exactly?  That the ground of the third pin is left floating? 

[silent:arts]

Re: Line Isolation transformer ground
« Reply #4 on: May 16, 2017, 05:13:46 PM »
Do you know what 'earth free' means exactly?  That the ground of the third pin is left floating?
Yes, it is a two pin connection only. Attached an example pic of the special Schuko-Connector they use.
« Last Edit: May 16, 2017, 05:16:54 PM by [silent:arts] »
SSLtech: No bypass, I don't believe in it.
Gus: I think this build might have problems with DIY builders that don't know things like ohms law.

abbey road d enfer

Re: Line Isolation transformer ground
« Reply #5 on: May 16, 2017, 05:51:28 PM »
I have a 500VA AC line isolation transformer I'd like to integrate into my lab bench setup
You seem to have two goals in doing this. One, you want to check "suspect guitar amps". I don't understand how an iso xfmr will help you. Iso xfmrs provide a floating output, so touching either of the live or neutral wires will not be dangerous. If you want to check an amp is actually "suspect", you want to have a non-floating supply.
You also seem to imply that it would be beneficial for noise measurement. This is more correct. What can happen, particularly when measuring unbalanced equipment, disconnecting the test set earth from the DUT earth may break a ground loop. Actually, you don't need an iso xfmr to do that, but using one will make it safe again.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
"The important thing is not to convince, but to give pause for thought." (B. Werber)
Star ground is for electricians.

JohnRoberts

Re: Line Isolation transformer ground
« Reply #6 on: May 18, 2017, 11:05:45 AM »
You seem to have two goals in doing this. One, you want to check "suspect guitar amps". I don't understand how an iso xfmr will help you. Iso xfmrs provide a floating output, so touching either of the live or neutral wires will not be dangerous. If you want to check an amp is actually "suspect", you want to have a non-floating supply.
You also seem to imply that it would be beneficial for noise measurement. This is more correct. What can happen, particularly when measuring unbalanced equipment, disconnecting the test set earth from the DUT earth may break a ground loop. Actually, you don't need an iso xfmr to do that, but using one will make it safe again.
Another consideration when bench testing 2-wire products, some will ASSume that neutral is grounded and cap couple to neutral for chassis shield ground.

For testing it seems you want as real as possible.

JR 
John Roberts
http://circularscience.com
Tune it, or don't play it...

Audio1Man

Re: Line Isolation transformer ground
« Reply #7 on: May 18, 2017, 12:57:36 PM »
Hi Matador
Most products come with 2 wire or 3 wire AC power connector. I do the test any 3 wire new device by verifying the chassis bond resistance before applying any power. If it is a two wire device I check the line cord for any continuity to chassis before applying any power.

Appling power to the new device I use my GFI box to power it up.  My bench device power is through an AC Power meter and I do use an isolation transformer when testing any device’s like a switching power supply that has built in line isolation.

Some Guitar amplifiers do have a “GROUND REVERSE SWITCH” that places a capacitor to chassis from the AC power line Hot or Neutral. This cap may trip the GFI. The old spec for leakage current was 5ma now 0.5ma.

Duke :)

JohnRoberts

Re: Line Isolation transformer ground
« Reply #8 on: May 18, 2017, 02:04:40 PM »
Hi Matador
Most products come with 2 wire or 3 wire AC power connector. I do the test any 3 wire new device by verifying the chassis bond resistance before applying any power. If it is a two wire device I check the line cord for any continuity to chassis before applying any power.

Appling power to the new device I use my GFI box to power it up.  My bench device power is through an AC Power meter and I do use an isolation transformer when testing any device’s like a switching power supply that has built in line isolation.

Some Guitar amplifiers do have a “GROUND REVERSE SWITCH” that places a capacitor to chassis from the AC power line Hot or Neutral. This cap may trip the GFI. The old spec for leakage current was 5ma now 0.5ma.

Duke :)
When did GFCI trip current drop a factor of 10x (was 4-6mA)?  Many power strips and assorted products will fail that stricter standard.

JR
John Roberts
http://circularscience.com
Tune it, or don't play it...

Audio1Man

Re: Line Isolation transformer ground
« Reply #9 on: May 18, 2017, 02:14:58 PM »
Hi John

The leakage current that I was talking about is from the CAPACITOR in the amplifier.  I have seen old products that the leakage current was 5-8ma. This current will trip the GFI.
Duke.


Matador

Re: Line Isolation transformer ground
« Reply #10 on: May 18, 2017, 06:08:37 PM »
I didn't think this all the way through:  I was originally thinking, if I have a device with a low impedance from hot to chassis, that powering with a floating secondary and lifted ground would allow me to ascertain it without worrying about getting shocked.  In other words, there isn't any fault condition on any equipment downstream of the isolation box that could kill me, all other things being equal.

In that case,  the chassis would be at full line potential, but with no earth return, if touched the entire DUT would float to whatever potential I was at so I could make measurements and debug.  With no isolation, I blow a fuse or trip a breaker faster than my own body could be a path for current (hopefully).

A line short to neutral on the isolated side would be seen as full secondary current on the isolation box, which means that the primary side of the isolation box should still (hopefully) trip a breaker or blow a fuse.  But any faults that are on the secondary side of the DUT (which is downstream of the isolation box) should be safely 'testable', if my thinking holds.

abbey road d enfer

Re: Line Isolation transformer ground
« Reply #11 on: May 18, 2017, 11:51:25 PM »
Yes, clearly there are two lines, which do not converge. One is safety; and that definitely advocates the use of an iso xfmr; the other is fault-testing, and that one requires a non-floating supply. Part of being a professional is taking risks, hopefully calculated  :)
Who's right or wrong is irrelevant. What matters is what's right or wrong.
"The important thing is not to convince, but to give pause for thought." (B. Werber)
Star ground is for electricians.

Speedskater

Re: Line Isolation transformer ground
« Reply #12 on: May 21, 2017, 05:39:34 PM »
I went back and looked at the Bill Whitlock paper that the schematic in post #1 was from.
The NEC rules for portable cord connected isolation transformers is the opposite from permanently hard wired isolation transformers in regard to Safety Ground connections.
Kevin