I've read the answers from a similar question but didn't find it to answer my question.

I'm trying to make a phase control circuit to control an AC motor, and would like to monitor the output.

The input can sometimes be normal mains AC, but other times from a battery/230VAC inverter. The inverter seems to have two opposing hot lines (e.g. -115VAC + +115VAC = 230VAC) rather than neutral/hot wire.

Can I measure directly?

I've got a DSO1054Z, which has a maximum input voltage of "CAT I 300 Vrms, CAT II 100 Vrms, transient overvoltage 1000 Vpk With RP2200 10:1 probe: CAT II 300 Vrms".

I understand from this article that CAT II would be recommended for a standard wall outlet.

The oscilloscope is only rated at 100 Vrms CAT II.

So I assume the answer is: no

Can I measure through a voltage divider?

I'm considering using this circuit to reduce the voltage (from 230 VAC to 2.27 VAC). Since there are two hot wires, the voltage divider has to go in both directions. Also, the high resistance should reduce the short-circuit current to about 0.23mA, greatly reducing the outcome of a worst-case scenario.

I'm assuming the answer is: yes, the probe will now be safe.


simulate this circuit – Schematic created using CircuitLab


I'm thinking to use the differential measurement on the oscilloscope to add the two signals.

  • \$\begingroup\$ I've updated the schematic. I'm a bit confused because I'm used to working with DC circuits. I'm guessing that one wire would normally be neutral, but in my case with the inverter, to measure 0V, I'd have to be in between the two wires (the center of the voltage divider). The earth wire might be at zero potential, but I'm thinking that it shouldn't be used as a circuit path for safety reasons, but can be used by the oscilloscope as a reference point (probably already connected if powered from the same circuit) \$\endgroup\$ Jan 15, 2017 at 9:02
  • \$\begingroup\$ That CAT rating referes to the short circuit current capability of the grid you are connecting it to. Can your inverter supply several kA? If not, you only need to worry about the voltage rating. \$\endgroup\$
    – winny
    Jan 15, 2017 at 12:37
  • \$\begingroup\$ The inverter can provide 1000w. More than that and it goes into failure mode. But I might also experiment with mains. But I have a 30mA leak current breaker + 6A current breaker. So you're saying the voltage rating mentioned, 300 Vrms CAT I, should support mains voltage? \$\endgroup\$ Jan 15, 2017 at 13:37
  • 1
    \$\begingroup\$ Please put @ before a username or that person won't be notified. Short circuit current is what matters for CAT rating. Your inverter might produce a few hundred before it blows, so "CAT 1". You use the term "mains voltage", but again, your inverter output and actual mains are two very different things. \$\endgroup\$
    – winny
    Jan 16, 2017 at 6:27
  • \$\begingroup\$ Is the inverter ground connected to the mains ground? Can it be, according to the instruction manual? Also, you might want to check the voltage differences with a suitably-rated multimeter before attaching the oscilloscope. \$\endgroup\$ Jan 16, 2017 at 19:34

2 Answers 2


The difference between your outlet and your inverter is what category it belongs to. CAT I is your average secondary of consumer equipment and similar. Your 1000 W inverter belongs here or is uncategorized, especially if you run it stand-alone from a battery, so you only need to worry about the CAT I or no category voltage rating of your probes and oscilloscopes.

When you move up to measureing mains, you need to wory about which overvoltage catagory and the short circuit current rating at that point. Inside your house, you are at CAT III (UK style fuse-in-the-plug) or CAT II (circuit breaker for each outlet in the fuse panel), fuse panel outside your house CAT III and so on. This puts more requirements on the isolation and fault breaking capabilities of your probes and DMMs. Here is a a picture I borrowed of the interwebs: CAT

As for your voltage divider, that's a safe and good way to divide the voltage down to suit your probes and scopes. You are however on your own if you where to connect it to CAT II or above since you need to guarantee the overvoltage and breaking capability of the divider yourself. A fuse would be the first step in doing so. Professionals use categorized probes for it but they are in CAT II+ envoriements, you are at CAT I or uncategorized when you measure your inverter.

As for your resistor divider, you need to take caution about the voltage rating of the resistors. Many people assume 1 Megaohm 0.5W resistor would be good for 700 V just because it won't overpower it, when it in fact is only rated for 150 V or so. You already mention redundancy, which is very good. A 2x3 resistor arrangement will give you 300 V rating even in case of a fault. For a lab experiment this might be excessive but for a permanently installed divider to mains or mains-level voltage, you need to consider the fault cases and what will fail if so.

EDIT: Your CAT 1 300 Vrms probe will divide the voltage for you, so the input voltage limitation on your scope will not be a problem. Also, if you are building a voltage divider as a permanent solution, remeber to cross each step of the ladder so it becomes one big matrix, not two dividers side by side. This creates your redundancy except for very special cases not even worth cosidering.

  • \$\begingroup\$ Maybe a reminder is also in order about the voltage rating of the resistors chosen for the divider. In this case it shouldn't be a problem, but can be an overlooked resistor parameter. \$\endgroup\$
    – user133493
    Jan 16, 2017 at 12:50
  • \$\begingroup\$ @replete Fixed! \$\endgroup\$
    – winny
    Jan 16, 2017 at 13:10
  • \$\begingroup\$ @winny Super feedback. I checked the datasheet for my resistors, and the working voltage is 200V, overload voltage is 400v, for a power rating of 1/8w. But as you mentioned, if one resistor fails (for whatever reason), the next will suddenly carry all the load. \$\endgroup\$ Jan 16, 2017 at 18:53
  • \$\begingroup\$ @user95482301 You´re welcome. I've updated the answer. If you feel it answers your questions, please mark it as answered. \$\endgroup\$
    – winny
    Jan 18, 2017 at 11:22

You circuit as drawn will reduce the voltage and protect the probes from short circuits. The only thing I see that could go wrong with that is a some kind of failure like a mechanical issue. If a piece of metal shorts the 1M resistor or the 10k resistor somehow becomes disconnected you no longer have a voltage divider. Just being aware of this and the fact you're probably not leaving this hooked up makes this unlikely. You could build some redundancy in to it with a 2 stage divider if you really wanted to. I suspect the full voltage wouldn't bother your oscilloscope but I agree by the rules you posted you're not supposed to try it out.

In reply to your comment I wouldn't worry about using ground as a current path in this case. It's not even enough current to trip a GFCI. Another thing is it's temporary. There's nothing dangerous about checking the voltage of a wire in your house against grounded metal with a volt meter. You might even discover that something that is supposed to be grounded isn't.

If individual leg to ground is important to your use then keep in mind these readings will be a little off if one of the leg voltages is different than the other.

  • \$\begingroup\$ Thanks. I've updated the schematic to use two resistors in series to allow for some protection. \$\endgroup\$ Jan 15, 2017 at 19:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.