Measuring Frequency,waveform and inrush current of 240v

[Shady]

Hi , I am interested in comparing the frequency and waveform of 230/240 outputs from the mains supply, a petrol generator and various inverters. I would also like to measure the max start up current of a 240v aircon compressor. ( I know that the newer fluke clamp meters could do the inrush test, but as its a one off I cant justify the cost as I already have two cheaper clamp meters)

At present I dont own and have never used a scope , but am seriously considering buying a rigol DS1052E (would eventually do the firmware hack to the 1102e). Would this scope and the supplied probes be suitable for these tests or can you sugest a set of more suitable probes that could be used with this scope or a more suitable scope.

Am I right in thinking that it would be best to power the scope with a different source than I wish to measure , i.e. use battery bank via inverter to power the scope to measure the mains sockets and vice versa. The mains tests would "only" be to provide a reference with which to compare the other sources to.


These tests are not the prime reason for buying a scope , but if this cheap scope would enable me to do the above tests then that goes a long way to justify finaly buying one.

Shady

[jahonen]

You'll need a CAT-rated voltage probe for mains measurements. Isolating scope power is not enough to make it safe.

This is because even if isolating the scope from the mains prevents direct short circuit if live is connected to scope ground, there is a big hazard for somebody accidentally touching scope ground and grounded object. Scope has only one ground so everything becomes live if single-ended probe is used.

Here is an example of suitable probe for safe mains voltage measurements:



For the currents, you'll need something like this (Fluke 80i-110s, good for 150A peak):



I have used that current probe to measure inrush current limiter effectiveness in my power amplifier project. It works quite nice, as you can see here.

Unfortunately, those costs easily more than what you probably would pay for clamp meter.

Regards,
Janne

[alm]

Agree 100% with Janne.

Never use a non-isolated scope (almost all of them, except things like Fluke Scopemeters, Tektronix THS7xx series or Tektronix TPS2xxx series) without proper ground connected. It's not just the BNC ground shell that would become a shock hazard, but the whole scope is designed with the idea that ground is a safe potential, so no attempt was made to isolate the user from it. The case might be connected to ground, or the isolation between the knobs/buttons and ground might be bad.

Depending on whether you want to measure the voltage or current waveform, the correct tool is a high voltage differential probe or current probe. High voltage differential probes are also made by companies like Tektronix, and also cheaper alternatives like Testec, but it's going to be expensive regardless. For current probes, you should be able to get a relatively cheap one because you only need AC, and not a lot of bandwidth. Should be possible for under $100 used, probably more when new.

[Zero999]

As it's low frequency, a simple mains isolation transformer should be able to do the job.

The only thing is, you need to calibrate it because the turns ratio will be slightly lower than the primary/secondary voltage rating would indicate so the secondary voltage will be higher than expected.

Start by using a 12V power transformer run of the mains, which should give a fairly good sinusoidal waveform. Measure the primary and secondary voltage with a DVM and work out the turns ratio, then you should be ready to get going.

A more accurate way would be to use a special voltage measuring transformer which will have a wider bandwidth and lower distortion than a mains transformer.

For current waveform measurement, use a current transformer or a hall effect sensor.

Another option for voltage measurement is to use a current transformer or hall effect but passing the mains cable through it many times and connecting it in series with a resistor but you'll probably find the power dissipation will be unacceptable to get enough current to give an accurate result.

[Zad]

How about a few turns of wire around the live of whatever it is you are trying to measure the current in?

[Simon]

what about two probes, one for live one for neutral and then use the subtract function to get your waveform back

[Zero999]

If you're going to do it that way, you might as well use one probe for live, assuming that the chassis is connected to earth, if not connect the 'scope's 0V to earth.

[Simon]

true but using both would guarantee to see what is on those wires and not run the risk of death, of course if you just want to see roughly what your mains looks like then yea one probe and put the ground to earth and voila'

[alm]

That method works, but the issue is that the CMRR of the vertical amplifier of almost any scope is pretty bad, so any common mode signal will cause distortion in your results. You clip both ground leads together, and don't connect either of them to the circuit under test. It is safe, and probably works better than using a transformer with dubious frequency response (not very much over 60Hz for most power transformers).

[Zero999]

true but using both would guarantee to see what is on those wires and not run the risk of death, of course if you just want to see roughly what your mains looks like then yea one probe and put the ground to earth and voila'

I'm not sure if I'm with you.

The neutral wire should be near 0V (within a couple of volts at most) so the voltage on it shouldn't matter, it'll give you an error of 1% maximum.

What I meant is that the chassis of most 'scopes is connected to earth so to measure the mains, all you need to do is connect the live to the probe but I know that there are some 'scopes which are double insulated (your Chinese 'scope for instance), in which case you need to connect the chassis (the probe 0V) to earth. This is perfectly safe, there's no risk of the scope's chassis becoming live, no neutral to earth short and no risk of a blown breaker, as long as the 'scope probe is rated to >350V it will be fine.

If you're really paranoid then connect the device under test to an isolation transformer but it will have a high impedance so might interfere with the current/voltage waveforms.

[NiHaoMike]

For voltage, use a power transformer with a rated primary voltage higher than the circuit voltage, like a 240V transformer for a 120V line or a 480V transformer for a 240V line. Or maybe use a step up transformer backwards. It should easily stay within the linear range.

For current, a Hall effect sensor is your best bet. Closed-loop nulling type for best accuracy and bandwidth, but that's a little overkill for mains use. Or a Rogowski coil, but those need compensation to be accurate with nonlinear loads.

[Neilm]

The trouble I have found with current probes is the dynamic response. They work well when there is a steady AC signal, but the transient from the inrush current can overload them and you don't see all the data. I had a problem like that at work where the inrush in the design was more than the mains switch could take resulting in a loud "crack" if you turned it on at the wrong part of the waveform. Eventually I had to use a small resistor and an isolated scope to do the measurement. (Tektronics 2024 with proper isolation). Checking it against the Tektronics current probe showed me that the probe was not showing the full magnitude of the pulse - it was not even going into overload when it should have.

Yours

Neil