What does an isolation transformer REALLY do?

Thread Starter

physgrad

Joined Jan 20, 2010
2
I've got a question trying to sort out what an isolation transformer really does. I'm a grad student working in a physics lab, and I'm trying to get all my instruments (lock-in amplifiers, a temperature controller, helium level monitor and computer) set up without ground loops and without a bunch of electrical noise. I've read some different papers and books and they all seem to differ on what an isolation transformer actually does (some seem to say it is a cure-all for every noise/ground loop problem you could ever have).

My basic understanding is that all an isolation transformer does is disconnect the 'neutral' wire from the ground wire (the utility company has the two lines connected at some point as a safety feature which will cause your circuit breaker to trip if either the hot or neutral wire gets shorted to the case of an instrument). If this is true, I don't see how an isolation transformer does much since any common mode noise between hot and neutral is coupled across the transformer (since the noise will create a changing magnetic flux just like the 60Hz power will). Am I missing something here?
 

KL7AJ

Joined Nov 4, 2008
2,229
I've got a question trying to sort out what an isolation transformer really does. I'm a grad student working in a physics lab, and I'm trying to get all my instruments (lock-in amplifiers, a temperature controller, helium level monitor and computer) set up without ground loops and without a bunch of electrical noise. I've read some different papers and books and they all seem to differ on what an isolation transformer actually does (some seem to say it is a cure-all for every noise/ground loop problem you could ever have).

My basic understanding is that all an isolation transformer does is disconnect the 'neutral' wire from the ground wire (the utility company has the two lines connected at some point as a safety feature which will cause your circuit breaker to trip if either the hot or neutral wire gets shorted to the case of an instrument). If this is true, I don't see how an isolation transformer does much since any common mode noise between hot and neutral is coupled across the transformer (since the noise will create a changing magnetic flux just like the 60Hz power will). Am I missing something here?
You are essentially correct. Isolation transformers remove DC components from signals and can also help with "ground loop" problems. For example, at HIPAS observatory, we had several widely scattered buildings, each fed with different legs of the commerciall power distribuition system. There were large differential voltages between the different buildings, which made transferring low level signals between them a real nightmare, as you can imagine. We used a vast array of video, audio, and pulse isolation transformers to eliminate these problems.

I've found that using many small isolation transformers on the signal lines is much more effective than single high-power isolation transformers on the power-line side. Of course, each situation is different.

eric
 

3ldon

Joined Jan 9, 2010
82
My basic understanding is that all an isolation transformer does is disconnect the 'neutral' wire from the ground wire (the utility company has the two lines connected at some point as a safety feature which will cause your circuit breaker to trip if either the hot or neutral wire gets shorted to the case of an instrument). If this is true, I don't see how an isolation transformer does much since any common mode noise between hot and neutral is coupled across the transformer (since the noise will create a changing magnetic flux just like the 60Hz power will). Am I missing something here?
this common mode noise you speak of is not common mode noise, its differential mode noise. common mode noise is transfered through the capacitance between the two coils, and this can be reduced by finding low capacitance transformers or connecting 2 or three isolation transformers in series, or using a common node choke or EMI filters, like these:
http://www.allelectronics.com/make-a-store/category/485/Filters-RFI-/-EMI/1.html
 
From my experience with vacuum tube audio amps, hifi and guitar, I can tell you that isolation transformers...or any power transformers, actually, that give you a secondary not connected to the earth or ground of the incoming line power...do allow you to do two things that can sometimes be helpful or increase safety:

1. An isolation transformer allows you to establish a separate earth ground...a true earth ground, if you wish...or not...but in any case, whether you establish a new earth ground or not, you are no longer connected to the neutral wire of the incoming line power. Why is this important? Because for various reasons, it is possible that the neutral wire might accidentally become "hot", including hot as in 20 KV from the power line itself! So, if you think your chassis or instrument is grounded, but in fact has become "hot" because of a faulty neutral wiring mistake or disconnect of neutral from true earth ground at the main distribution box, you could be killed by electrocution from anywhere from 120 volts AC up to tens of kilovolts. If this were a perfect world, and neutrals were always connected to earth grounds correctly at the main panel, and the earth ground at the main panel was always a perfectly good earth ground, and neutrals were not accidentally connected, by mistake or by short-circuit, to the hot side of distribution grids/pole transformers/etc, then the isolation transformer would not be needed. But mistakes happen.

2. With an isolation transformer, you have the option of determining for yourself where the ground potential should be for "downstream" circuitry. Sometimes one can greatly reduce noise in circuits by "lifting the ground" above ground. For example, by making the ground of a circuit ten volts to even 90 volts electropositive relative to the ground plane of the previous circuit or equipment or power line, the ability of noise originating from more electronegative sources to transfer into the more electropositive ground is greatly reduced. This effect is seen most strongly by elevating the filaments of vacuum tubes above ground (electropositive to ground), which have an intrinsic diode effect (filament to cathode) but it can apply to all sorts of electron guns and even solid state sensors. A distinction should be made between a circuit ground one wishes to elevate, however, and a chassis ground...a chassis ground must always be to a good, proven, true earth ground, not relying ever on the neutral line of line power, in order to avoid possible electrocution.

I dunno if this is of the slightest help to you, but I saw the thread and thought perhaps this might be of some relevence.

best, charlie
 
Top