Isolated voltage measurement

[MarkoAnte]

Hi,
I'm looking for a way to successfully measure different points in a 3 phase pfc and dc/dc converter. Till now I have been using Testec TT-SI-9002.
But all of this devices (we saw Dave take apart a LeCroy one) seem to only have a high value resistor divider inside. As long as I'm measuring only 50/60 hz ac its ok, but as soon as I turn on the power switcher it pick up all sort of crap as soon as I connect 1 lead to anywhere.

Are there any better solutions that somebody knows about? Maybe something that is actually isolated and /or EMI hardend. I am thinking of making my own isolated one I may even have it as a project for university - killing 2 birds with one stone. Just have a ADC on the isolated side and then isolate the paralel communication form the adc and convert it to a waveform on the mains referenced side. I'm using AMC1200 to measure DC and ac voltage and current and I get more usable readings off that than form the Testec TT-SI-9002.

Any input is apreciated.

Best regards,
Marko

[PartialDischarge]

What BW do you need to measure?

[MarkoAnte]

The switching frequency is 60 kHz but that is a square wave. I think somewhere about 1-5 Mhz to get some information about transient effect from switching. The AMC1200 would be to slow. I was thinking about using something like a 50 - 100 Mhz adc with paralel data output. Isolate that and get it on the other side. And just use a R-2R ladder to get a voltage from that and then use some analog filtering and buffering and put it on a BNC to scope.  If the r2r lader would be fast enough? It would be definitely a critical layout placement with low inductance smd resistors.

[David Hess]

Use a better high voltage differential probe.  I am surprised the Testec TT-SI-9002 was not sufficient.  Maybe it is out of calibration or damaged.

[tautech]

Have a search in the forum for Siglents ISFE for input isolation for scopes
Differential probes are you other simple option

[oldway]

http://www.analog.com/media/en/training-seminars/tutorials/MT-071.pdf

[2N3055]

I agree with David..  That probe should be OK.. It may be damaged..

Also if you want to measure switchers, 25MHz is about minimum to do it right...
Less bandwidth is not enough to see the edges and faster transients.. And those are the important ones..

It might be that your probe is fine.. It might be that switcher is too noisy, so what are you seeing is really there..
There could be some capacitive common mode coupling, or maybe conductive coupling..

You need to power DUT and scope from different sockets, and put filtering on both, to prevent injection trough power cords...

Sometimes just rearranging stuff on the desk makes it better..

Hope you solve this.

[mk_]

Hi,
I'm looking for a way to successfully measure different points in a 3 phase pfc and dc/dc converter. Till now I have been using Testec TT-SI-9002.
But all of this devices (we saw Dave take apart a LeCroy one) seem to only have a high value resistor divider inside. As long as I'm measuring only 50/60 hz ac its ok, but as soon as I turn on the power switcher it pick up all sort of crap as soon as I connect 1 lead to anywhere.

Are there any better solutions that somebody knows about? Maybe something that is actually isolated and /or EMI hardend. I am thinking of making my own isolated one I may even have it as a project for university - killing 2 birds with one stone. Just have a ADC on the isolated side and then isolate the paralel communication form the adc and convert it to a waveform on the mains referenced side. I'm using AMC1200 to measure DC and ac voltage and current and I get more usable readings off that than form the Testec TT-SI-9002.

Any input is apreciated.

Best regards,
Marko

As said before - this Probe should be ok, we use it all the time @ really noisy environments. But anyway - take a look for Testec TT-SI 9110, which has higher bandwith and is a little bit more robust.

Anyway, measuring in high voltage/high-frequency-systems needs much more then a solide probe... the risetime of a 60khz-switcher @ high voltage  coupled throu some pf coupling-capacity eliminates low voltage measurements without a good differential probe... there is a lot of black magic in measuring low voltages signals in this situations....

your ACM... well..  sorry to say that - if you can`t verify your Testec I do not think that you can build a better probe from scratch... it`s not that difficult but... you should search for differetial probes here in this forum, som nice designs and knowhow can befound.

Michael 

[MarkoAnte]

Sorry for the long w8 for a replay.

It might be that your probe is fine.. It might be that switcher is too noisy, so what are you seeing is really there..

I don't think so. Even if I just connect the 2 probes to the same point I still get a noisy 0V.

There could be some capacitive common mode coupling, or maybe conductive coupling..

This was my thinking also. That's why I was looking for a real insulated differential probe.

You need to power DUT and scope from different sockets, and put filtering on both, to prevent injection trough power cords...

My DUT is 3 phase. It is a separate 3 phase cable that goes to the power box and then back to me. Its just that the power box is just on the other side of the wall and there is no 4 pahse to hook my self on to .  So in all there are about 2 meters of 3 pahse cable to the power box and then 1 meter back to me.

I will have to try to measure it with my OWON  scope that has batery power operation. But its so bad. So so bad.  To bad there is no barf emoji here 

[2N3055]

Did you try to short diff probe inputs and than connect it to signal generator, at various frequencies... It seems to me that your probe has bad CMMR...

Also, did you try to look a FFT of the noise to get spectrum? It might give you some clue..

Before you go any further, did you try loop antenna? You can try just shorting ground wire to the tip of your normal 10x probe, and sniff around wiring and switcher..
I'm sorry, I don't know much about you, so don't know if you tried all the basic troubleshooting yet.. 

[David Hess]

Sorry for the long w8 for a replay.

It might be that your probe is fine.. It might be that switcher is too noisy, so what are you seeing is really there..

I don't think so. Even if I just connect the 2 probes to the same point I still get a noisy 0V.

Do a couple of different tests.  Connect the probes to themselves and nothing else, connect them to ground on the device, connect them to a stable point at the extreme of common mode voltage that you want to measure, and then connect them to a point which is swinging over the common mode range that you are trying to measure.

Also experiment with connecting the probe ground leads to each other and nothing else.

What kind of sensitivity are you trying to achieve?  Modern differential probes are often noisier than old ones because they rely on more front end attenuation instead of bootstrapping their input circuits which allows for a high common mode range without attenuation.

There could be some capacitive common mode coupling, or maybe conductive coupling..

This was my thinking also. That's why I was looking for a real insulated differential probe.

While I do not really recommend this for safety reasons, a power line isolation transformer can be used on the oscilloscope and then the oscilloscope ground can be connected to the common point for measurements on the device.

I will have to try to measure it with my OWON  scope that has battery power operation. But its so bad. So so bad.  To bad there is no barf emoji here 

They make DSOs with isolated *single* ended inputs which are very useful for 3-phase and other off-line measurements.  Using a battery powered DSOs is the next best thing.

[MarkoAnte]

@2n3055 I will try to short diff probe inputs and than connect it to signal generator. But I don't have any at high voltage generators only the one in my scope +-5V I think. I don't as much want to see where the noise comes form in my switcher but more to see the switching action of the transistors. The switcher is a 3 phase 10KW PFC and it works in hard switching of all of the transistors and its inevitable that its going to make some noise. But I have worked with switchers in the past and its seams that I made the switcher(and or the control pcb) better than all the previous ones because I had 0 problems with noise in my control PCB till now. (knocks on wood ) But the quantification of the nose and ringing of the switched waveform is what I wanted to see.

@David Hess

...connect them to a stable point at the extreme of common mode voltage that you want to measure, and then connect them to a point which is swinging over the common mode range that you are trying to measure.

this is what I have problems with.

I have SDS7102 form OWON. It was my first scope. And it drives me insane. The firmware is broken. The fan is broken and makes noise if you look at it funny (I have send it for repair and the problem was back out of waranty this time). I then got a old tex scope, a rigol scope and then I sold thoes and got my self a Keysight   MSO 3000 and every time I have to use the OWON because of the battery feature (only reason why I sill did not scrap it) and hear the fan making noise again I die a bit inside. But yea in my old job we had a fluke portable oscilloscope and it was way better at this kind of measurements.