Double-conversion online topology UPS Sine wave quality

[hendrix]

Hello,

I have a question about checking the quality of the sine wave coming out from the double-conversion online UPS that I need to use for a project. I would like to do this with an oscilloscope and I know its dangerous (230v mains).

The question is, if I plug a step down transformer to the output of the UPS (something like 12v secondary maybe), and measure from the secondaries , will the 12v AC sine wave accurately represent whats going on with the 230v ac output of the UPS or will the transformer smooth some of the ripples in the primary?

Thanks.

[johansen]

yes, it will be sufficiently accurate.

is one side of that 230v output grounded?

[amspire]

Mains transformers have poor frequency response, so you probably cannot see how bad the waveform really is.

You can do better by making up a simple current probe, and measuring the current through a resistor across the UPS output.

Start with something like this for a few dollars:

http://www.ebay.com.au/itm/5A-Electric-Power-Mini-Current-Transformer-TA1309-100-/391260986153?hash=item5b18f8db29:g:PuAAAOSwoydWq0lS
Data Sheet:
http://www.yhdc.com/en/product/415/

Put a 200 ohm resistor across the pins - now 5A current through the current transformer = 1v across the 200 ohm resistor. Response 20Hz-20KHz and isolation rated at 6000V for 10 seconds.

Now get some enamelled copper wire and wind 50 turns through the hole - keep the wire away from the pins obviously. Now 100mA through the enamelled wire - 1V output.

Get a 10K 10W Wirewound resistor and put it in series with the enamelled winding  and you now have a safe probe that for 240V in gives you  24 mV out. If you are patent enough to do 500 turns, you get 10 times more sensitivity and you can use a 100K 1W resistor instead.

You can buy proper current probes and high voltage differential probes for scopes that have a much wider bandwidth and sometimes can do DC current but they tend to start at about $400 and go up from there.

If you can find a good split core - something like this:

http://www.ebay.com.au/itm/New-Non-invasive-AC-Sensor-Split-0-30A-Core-Current-Transformer-SCT-013-030-/331751675239?hash=item4d3df0d567:g:tpkAAOSwa-dWk1dF
Data:
http://garden.seeedstudio.com/images/b/bc/SCT013-030V.pdf

it is much easier winding the primary turns. However, I have no idea what frequency response this transformer has. It also has a lower isolation voltage.

[hendrix]

Johansen,

No, I was thinking about using the transformer without grounding it just to be on the safe side.

[hendrix]

Amspire,

Thank you, however, what you are making me build seems awfully like small transformer, is that wrong?

[amspire]

Amspire,

Thank you, however, what you are making me build seems awfully like small transformer, is that wrong?

Yes, a current transformer is just a transformer with a primary consisting of a wire going though the core. Nothing wrong with making the same wire go 50 times though the core instead.

It is just a way to get a cheap transformer that has a very high isolation, excellent frequency response and is very cheap.

The big difference is that it will be running with less then a volt across the primary instead of 240v.

IF you could find a good audio transformer from an old valve amplifier, you could possibly get similar bandwidth, but not the same isolation and safety.