medical grade isolation transformers

[djacobow]

How are they better than just regular old isolation transformers? Lower leakage? Someone claims that the medical grade devices do not connect ground and neutral the way so many consumer grade isolation transformers aimed at musicians seem to.

I'm going to use this to power mains connected DUTs so that I can probe them with my scope, so I'll be lifting that neutral/ground connection and ground pass through if it exists anyway.

I can get a non-medical grade 250VA TrippLite device at Amazon for < $100, shipped. Medical grade is more than 2x. Are they really buying a higher class of transformer, and if so, 2x as expensive?

[saturation]

The specifications for medical grade safety and isolation UL 60601-1 is in general, allowable leakage currents are far less than consumer grade, and the UL spec listed various limits for different applications and how to achieve the same.  You can get general purpose medical grade isolation transformers on eBay; a used medical grade is about the same price as new consumer grade.

[djacobow]

Thanks!

Shipping cost may be decisive. Amazon Prime is a powerful thing.

[saturation]

Welcome.  Summary of the new leakage current limits and new names:

http://www.ebme.co.uk/articles/electrical-safety/335-electrical-safety-tests

A pdf comparing consumer versus medical grade:

http://www.slpower.com/reference/An113%20Leakage%20Current.pdf

The pdf uses the old standard.  The newer standards have some overall relaxed limits.

Hospitals tend to discard isolation transformers once the warranty runs out, refurbishing rooms or equipment, so you can pick up many very good ones' via eBay cheaply, despite shipping costs from the weight.  Cost varies with power output.

$55 delivered:

http://www.ebay.com/itm/Dale-Technology-Isolation-Transformer-Medical-Grade-model-IT200-1-5-Amps-/121311079821?pt=LH_DefaultDomain_0&hash=item1c3eb42d8d

[djacobow]

OK, I got a cheap non-medical isolation transformer, and have had a chance to take it apart. What I found inside is pretty interesting, and somewhat confusing. I've also done some reading, which has confused me further.


The following are connected together:

   - the input ground pin
   - the chassis
   - the output ground pin
   - the secondary neutral
   - a "mystery" green wire going into the xformer on the 2ndary side

So, this is clearly not what I want. Now, my first idea to modify this to be useful as a bench tool was to separate the ground pin on the output and separate the secondary neutral from ground. However, this doesn't work because the outlet provided for the output actually links the ground screw and the mounting screws to the chassis.

So, my second idea was just to separate the input ground and leave everything else as is. This gives me true isolation on the secondary, and it leaves the secondary "floating ground" and neutral at the same potential. BUT it also leaves the chassis of the transformer at the secondary "floating ground."

Is this the best possible arrangement? It seems not to me. I don't like that a fault between primary and the transformer chassis would go undetected until I touched it. I'm also not sure if the neutral and ground on the output should be shorted. But, from the perspective of a DUT plugged into a "real" outlet, this is exactly the situation (they are connected back at the service entrance) so maybe it's fine.

Second, what is that mystery wire going into the transformer? It looks as if it is tap on the winding, but I know it is not. Is it to some kind of cage around the secondary to manage capacitive coupling?

Finally, I'm considering connecting the secondary "floating ground" to the real input ground through a 10 Meg resistor, maybe a small cap, too. Is this a bad idea? Pointless?

Suggestions welcome!

Regards,
Dave J

[BravoV]

Had similar issue while ago and discussed and got answer in my thread here -> Modding an isolation transformer, please advise, hope this helps.

[Gallymimus]

OK, I got a cheap non-medical isolation transformer, and have had a chance to take it apart. What I found inside is pretty interesting, and somewhat confusing. I've also done some reading, which has confused me further.


The following are connected together:

   - the input ground pin
   - the chassis
   - the output ground pin
   - the secondary neutral
   - a "mystery" green wire going into the xformer on the 2ndary side

So, this is clearly not what I want. Now, my first idea to modify this to be useful as a bench tool was to separate the ground pin on the output and separate the secondary neutral from ground. However, this doesn't work because the outlet provided for the output actually links the ground screw and the mounting screws to the chassis.

So, my second idea was just to separate the input ground and leave everything else as is. This gives me true isolation on the secondary, and it leaves the secondary "floating ground" and neutral at the same potential. BUT it also leaves the chassis of the transformer at the secondary "floating ground."

Is this the best possible arrangement? It seems not to me. I don't like that a fault between primary and the transformer chassis would go undetected until I touched it. I'm also not sure if the neutral and ground on the output should be shorted. But, from the perspective of a DUT plugged into a "real" outlet, this is exactly the situation (they are connected back at the service entrance) so maybe it's fine.

Second, what is that mystery wire going into the transformer? It looks as if it is tap on the winding, but I know it is not. Is it to some kind of cage around the secondary to manage capacitive coupling?

Finally, I'm considering connecting the secondary "floating ground" to the real input ground through a 10 Meg resistor, maybe a small cap, too. Is this a bad idea? Pointless?

Suggestions welcome!

Regards,
Dave J

I dunno man, clearly this isn't a full isolation transformer.  I'd try again and buy another one rather than dinking with this.  Consider your safety and the safety of your test equipment.  You really don't want an inappropriate amount of leakage or a mediocre quality isolation barrier (not to mentions your primary short to ground concern).  BTW neutral should not be connected to ground if you intend to have a true isolation transformer as the transformer output would no longer be floating.

BTW I have a couple of these which I use for neuroprosthetic development.  They really aren't that expensive.
http://www.amazon.com/Tripp-Lite-IS250HG-Isolation-Transformer/dp/B0000ET7Q6

one interesting (but annoying) features of these transformers is that they generate a LOT of heat just idle.  Probably a good 10 - 20W just being plugged in.  Upside is the leakage is truly nice and low as it should be.

[saturation]

The reduce total transformer losses is one reason a toroidal is preferred if you can afford it and is better long term in installed devices.  Not to mention in non-toroidal transformers, the bigger it gets the more annoying the hum.

http://www.amgistoroids.com/toroidal-technology.html

BTW I have a couple of these which I use for neuroprosthetic development.  They really aren't that expensive.
http://www.amazon.com/Tripp-Lite-IS250HG-Isolation-Transformer/dp/B0000ET7Q6

one interesting (but annoying) features of these transformers is that they generate a LOT of heat just idle.  Probably a good 10 - 20W just being plugged in.  Upside is the leakage is truly nice and low as it should be.

[amyk]

One thing to be aware of with toroidal transformers is they have very high inrush currents.

[G7PSK]

The green wire disappearing into the secondary winding will be to an earthed shield. You could remove the center tap earth lead from the secondary and fit a new outlet socket, that way you have a fully isolated secondary without an earth.

[djacobow]

Thanks, Bravo. That thread is indeed helpful. If I read it correctly, you were advised to simply separate the secondary neutral from ground and leave it at that. That gives you a fully isolated output but with everything else still earthed, including the ground pin on the output.

I'm thinking about the implications of that for a DUT that uses the ground pin for its own chassis. It seems to me that a fault in the DUT that bridged neutral to ground, or hot to ground would go undetected while using this transformer on the bench. Later, when you plugged the device back into a normal outlet, the hot/ground fault would trip the breaker, and the neut/ground fault would do nothing, so, I guess... no harm no foul there.

Had similar issue while ago and discussed and got answer in my thread here -> Modding an isolation transformer, please advise, hope this helps.

[djacobow]

OK, so I understand what you're saying about fitting a new socket that does not connect to earth. (I might choose to just add one, leaving the earthed on in place) but I'm not sure I follow the "you could remove the center tap earth lead from the secondary."

I thought you just suggested that the green lead going into the transformer wasn't a tap on the secondary at all, but a shield. (It could not have been a center tap on the secondary because as shipped it was connected to one of the secondary outputs; that would have shorted half the secondary.)

The green wire disappearing into the secondary winding will be to an earthed shield. You could remove the center tap earth lead from the secondary and fit a new outlet socket, that way you have a fully isolated secondary without an earth.

[djacobow]

Agree, this is not exactly the beast I want, but, leakage aside, I'm not sure it can't be made to function that way. I may still get a medical grade transformer, but I'm a little leery that when I get a medical grade transformer and open it up, it'll be wired just the same way. It's not like anybody provides a schematic or a sensible description of how the unit is wired!

I dunno man, clearly this isn't a full isolation transformer.  I'd try again and buy another one rather than dinking with this.  Consider your safety and the safety of your test equipment.  You really don't want an inappropriate amount of leakage or a mediocre quality isolation barrier (not to mentions your primary short to ground concern).  BTW neutral should not be connected to ground if you intend to have a true isolation transformer as the transformer output would no longer be floating.

BTW I have a couple of these which I use for neuroprosthetic development.  They really aren't that expensive.
http://www.amazon.com/Tripp-Lite-IS250HG-Isolation-Transformer/dp/B0000ET7Q6

one interesting (but annoying) features of these transformers is that they generate a LOT of heat just idle.  Probably a good 10 - 20W just being plugged in.  Upside is the leakage is truly nice and low as it should be.