Isolating a non- isolated USB scope?

[jman43195]

(First post) So I want to get a Hantek 6022BE or an Owon VDS 1022(non I model), but sadly, as a beginner, I don't have much money to spend on a nice bench top scope, or a differential probe. Would using a "USB isolator" allow me to work on 120V mains voltages without it blowing up?

[Kleinstein]

The USB isolator could avoid blowing the computer, but it is a dangerous way - especially for a beginner. The ground side of the probes could still be life voltage and most probes are not isolated. Even if there is some insulation it may not be main rated.

The right way is usually to have the scope grounded and use an isolation transformer for the circuit to test. Even than mains voltage (even if only 120 V) is not really suited for a beginner. 

[jman43195]

So could I view the secondary of a transformer with one side grounded? Sorry, don't know if an isolation transformer has anything special compared to a regular transformer

[nctnico]

(First post) So I want to get a Hantek 6022BE or an Owon VDS 1022(non I model), but sadly, as a beginner, I don't have much money to spend on a nice bench top scope, or a differential probe. Would using a "USB isolator" allow me to work on 120V mains voltages without it blowing up?

DOn't float your scope or equipment under test. Get a CAT rated differential probe and a GFI.

[jman43195]

I don't have $200-500 to dump for just a differential probe right off the bat. That's why I asked if I could isolate that kind of scope.

[shakalnokturn]

Cheap solution to start with is the isolation transformer as stated previously.
Use it on the DUT not on the PC/scope.
Adapt it's power rating to the equipment you're likely to work on. (200VA should get you quite far.)
All it does is break the reference that neutral has to ground on the distribution network, that way you can tie the ground clip of your scope (that Must be ground referenced) to a point that would have otherwise created a short-circuit. (Watch the EEVblog video on the subject...)

By tying a ground clip you're reestablishing a reference to ground, this means that DUT is no longer safe for fingers.

If scoping with more than one channel I recommend you use a ground clip on one probe only. (Avoids doing something that shouldn't be done like choosing 2 different references 150V apart...)
Keep the probe hooks on the probes, or if you must go poking fit some form of insulation around the ground ring at the end of the probe. (Avoids sparks if you slip...)

[shakalnokturn]

The even lower cost solution is without transformer, your scope/probe combination must be able to handle the voltages WRT ground.
You set both channels of the scope to have same sensitivity and a math function to subtract one from other.
You now have a single channel differential input.
Make sure to NOT use any ground clip on either probe, also make sure ground ring is insulated.

[Gyro]

Agreed with the others on the safety warnings. Even on an Isolated VDS1022I, the metal case will be at probe ground clip potential.

Picking up on Kleinstein's point about protecting your computer... It's worth paying a little extra for the VDS1022I isolated version. The non-isolated version can be upgraded by component additions, but you might not break even on component cost. Also note that the DC-DC converters on most cheap external USB isolator boards don't have enough output current capability. The 6022BE performance will suffer badly from being used with a USB isolator as it needs a much higher USB bandwidth.

[Gyro]

The even lower cost solution is without transformer, your scope/probe combination must be able to handle the voltages WRT ground.
You set both channels of the scope to have same sensitivity and a math function to subtract one from other.
You now have a single channel differential input.
Make sure to NOT use any ground clip on either probe, also make sure ground ring is insulated.

A very cost-effective approach with decent quality probes,  however it works much better on analogue or higher performance DSOs than low cost USB ones. I've observed quite a lot of glitching on edges.

[shakalnokturn]

Having isolated USB is good "just in case", can save the scope/PC from roasting if you go wrong.
It also has the disadvantage that you could make your scope live without realising.
If it was my setup I'd re-reference the floating scope to ground with a resistor low enough to trip the GFCI, that way you know you're doing something dangerous and the resistance will in some cases avoid big bangs.

[Gyro]

^ Agreed.

[nctnico]

I don't have $200-500 to dump for just a differential probe right off the bat. That's why I asked if I could isolate that kind of scope.

A diff probe is around the $150 mark nowadays. Look for the MicSig DP10013. Otherwise use the math function on your oscilloscope to subtract two channels. Just don't float the DUT or oscilloscope. It is very easy to make a deadly mistake if you make a large portion of your test setup carrying lethal voltages without the protection of ground and a GFI.

[Brutte]

My way of dealing with 240VAC is to do development on 24VAC.
Once the development is done, I just swap components for higher voltage rating. The advantage is that I can focus on what scope shows instead of what I am touching.
50W transformer 240V/24V costs $10 in here.

[nctnico]

My way of dealing with 240VAC is to do development on 24VAC.
Once the development is done, I just swap components for higher voltage rating. The advantage is that I can focus on what scope shows instead of what I am touching.
50W transformer 240V/24V costs $10 in here.

That is a good method. What I do sometimes when dealing with a defective switchmode PSU is to power the primary side using a bench PSU at a safe voltage (often this requires two supplies; one for the general input and one to supply the controller chip). This way you can safely check whether the waveforms make any sense and the primary side has any chance of working at all.

[David Hess]

I would absolutely use a USB isolator no matter what measurements I was doing with a USB instrument.  It does not require a power line measurement to damage your laptop or computer through its USB port.