Isolation Transformer

[oakkar7]

Dear All,

It is not a modern tech but still mystery for beginners. After I bought my test gears such oscilloscopes, I noticed that I should have an isolation transformer in my bench. Then, I forgot again with the other projects. One day, after seeing Dave and Todd videos blog about isolation transformer, I started the construction of DIY isolation transformer.

For those interests, here is Dave and Todd explained videos. Thanks both.
http://www.eevblog.com/2012/05/18/eevblog-279-how-not-to-blow-up-your-oscilloscope/
http://www.toddfun.com/2011/04/30/isolation_transformers/

And later, I recently found a post at Keith’s Electronics Blog. There is also original schematic from manufacturer. It isolates even ground pins of each output. Yes, system ground (Chassis ground) is purposely isolated.
http://www.neufeld.newton.ks.us/electronics/?p=500
and schematic,


And here is an AN from APC concerning about isolation in UPS system but worth to read.
http://www.apcmedia.com/salestools/NRAN-7NB2FG_R0_EN.pdf

I agree with Todd suggestion. This is the schematic from commercial isolation transformer. The output winding NEUTRAL is CONNECTED to main GROUND!!!

http://www.carebase.com/Isolation-trans ... -1000B.htm



This is the edited diagram as per suggestion. Disconnected secondary winding (NEUTRAL) from main GROUND.



All blogs and AN have their point of view in isolation but questions still exit.

My Notes:
- DUT (device under test) should be isolated with transformer rather than oscilloscope.

- BUT, Dave scenario-3… if DUT power supply is grounded to chassis (main ground), the problem is still existed. Disconnect DUT ground ?? or use with floating power supply?! or with battery?

- Test device such as oscilloscope should not be floated (must be grounded) without special reason.

- Most forum and articles talked about isolating oscilloscope not DUT. If someone know the best practice and pros, cons, let me know.

- Differential probe, usb isolator are also good but I cannot still effort.

- I need GFCI outlet at my bench.

- My thought is that isolating system ground has some pro and cons. Sometime we need to isolate for full safety or purposely. In most case, test equipment should be grounded for safety.

My construction is here.
http://okelectronic.wordpress.com/2012/12/04/diy-isolation-transformer/

Pls share your thought, experience and best practices.

[Sleepwalker3]

If you buy a professional *purpose built Isolation transformer* (that complies with all the regulations), like I have on my bench, it will have the secondary not connected to anything - i.e. Floating. The Earth will still connect to Earth, just as it should. You should *NEVER* disconnect the Earth on your CRO, nor on your Isolation transformer. If you happen to contact one of the live secondary wires, you will get a minimal shock, even with the isolation transformer, though it will be minimal and likely only barely a tickle. You should note that Earth Leakage circuits will *NOT* trip, unless you have specialised types.

You should also note that having filter units or SMPS's connected to the isolated supply may (and likely will) affect the isolation of that supply, as many have filter caps to Earth, thereby causing a path to Earth if you contact the opposite leg - In other words, you would get more of a belt than if there were no caps on the opposite leg. Do not think that leaving the Earth disconnected on these is a good idea, it's not.

If you really need to measure floating voltage (e.g. you work with high voltage H bridges, High side circuits, EHT floating supplies, etc.), then you should either buy a purpose made floating differential probe, buy a floating isolated channel CRO (like some of -the great Tektronix ones), or use the differential functions on your CRO, providing the levels are within the range of your gear (obviously CRO being a generic term here).

When you work with an Isolation transformer on your bench, you need to be mindful of what you're plugging into the isolated supply (as mentioned above), otherwise it may not be as 'isolated' as you think. Some of my test bench supplies are double isolated (two completely separate galvanic isolation layers) , though the above comments still apply.

If you're working on this sort of equipment for a living (or if you just would like to be around for a while longer), you need proper high quality isolation gear that complies with the regulations, not homemade "I think this will work" stuff.

[oakkar7]

....

When you work with an Isolation transformer on your bench, you need to be mindful of what you're plugging into the isolated supply (as mentioned above), otherwise it may not be as 'isolated' as you think. Some of my test bench supplies are double isolated (two completely separate galvanic isolation layers) , though the above comments still apply.

yes, i also noted this (about to mind what DUT is plugged in). What you means double isolated? there is three windings? Can you explain with a diagram?

thanks all,

[logicnibble]

Hi!
I want to work on an offline power supply (flyback) 50W (100W maximum) prototype from 230V 50Hz and I want to be able to check signals with a Rigol DSO, and obviously, work safely.
What would be the professional approach?
I'd like to buy an isolation transformer but I don't know which one would satisfy my needs, nor where to buy one from Europe.

[peterbj]

Hi!
I want to work on an offline power supply (flyback) 50W (100W maximum) prototype from 230V 50Hz and I want to be able to check signals with a Rigol DSO, and obviously, work safely.
What would be the professional approach?
I'd like to buy an isolation transformer but I don't know which one would satisfy my needs, nor where to buy one from Europe.

I'm not sure about the exact approach but I do know that some have gotten the parts from ebay.

[Sleepwalker3]

yes, i also noted this (about to mind what DUT is plugged in). What you means double isolated? there is three windings? Can you explain with a diagram?

Sorry for the delay in responding, I've been off DP for quite a while and am not in much these days either. I meant that I have a main isolation transformer for my bench, but I also run an additional transformer for stepping down to 110V, but that also provides another level of isolation. If you have say a powerboard with a filter hooked off the main Iso transformer, then it will be leaking current to Earth via the small filter caps (better to use a cheap one without a filter and without MOVs to Earth), if you have a separate transformer for powering just the gear, that ads another layer of Isolation, which can be useful in some circumstances (e.g. when your main Isolation circuit has some level of leakage to Earth). Likewise if you are using an SMPS connected to your Isolated supply - Unless it's a medical grade certified one with minimal leakage, you're likely going to have a bit of leakage, so another transformer can be handy (or multiple isolated secondaries like the pics above). In my case, the second transformer runs off the first, but in most cases it wouldn't matter if it ran directly off the mains, as it still isolates.

[Sleepwalker3]

Hi!
I want to work on an offline power supply (flyback) 50W (100W maximum) prototype from 230V 50Hz and I want to be able to check signals with a Rigol DSO, and obviously, work safely.
What would be the professional approach?
I'd like to buy an isolation transformer but I don't know which one would satisfy my needs, nor where to buy one from Europe.

Probably best to contact a Euro transformer manufacturer. I got mine is Australia and could give you the details, but obviously that would cost a lot to freight and I'm sure there are plenty of good manufacturers over there. I'd avoid cheap Chinese stuff, this is a safety device for you, so play it safe and get one that meets with the various safety standards. If probing a *LIVE* bare SMPS on the bench, be careful, where heavy duty rubber gloves where necessary (eg. if you happen to rest your hand on the unit to be able to steady the CRO probe) and have a bit of insulating material you can use as well. I like to have those little non-slip mats for vases and things that you can get from supermarkets or $2 stores. Sit the unit on on of them and they stop it sliding when you're probing.

Never think you can't get a shock because you have an Isolation transformer, you can *always* get a shock that can easily kill you, so go slow, think about what you want to do before doing it and take regular breaks. Good idea to wear safety glasses too, if you slip with the probes, things can go 'BANG' in your face.

[yo0]

from the OP

And later, I recently found a post at Keith’s Electronics Blog. There is also original schematic from manufacturer. It isolates even ground pins of each output. Yes, system ground (Chassis ground) is purposely isolated.

you notice ground conections in outputs are scraped or covered in the schematic? check carefully hi-res picture in OP link.


best regards

Pio

[Sleepwalker3]

Yeah, looks like they changed their minds and made them no longer connected. I don't like that arrangement in most cases, I prefer it to be Earthed right the way through.

[oakkar7]

from the OP

And later, I recently found a post at Keith’s Electronics Blog. There is also original schematic from manufacturer. It isolates even ground pins of each output. Yes, system ground (Chassis ground) is purposely isolated.

you notice ground conections in outputs are scraped or covered in the schematic? check carefully hi-res picture in OP link.


best regards

Pio

Woo, woo, your eyes are sharp. I don't really notice that before.
Honestly, I think first that I understood the logic behind this simple transformer. But, after reading and searching many forums, resources, I confused again.
After then, I realized that isolation should be consider case by case, not only transformer alone.