Differential scope probes?

[zigzagflux]

Those of us testing our tube gear would greatly benefit by using differential probes. Currently, I tend to rely on AC coupling of my scope, and inverting one channel in ADD mode to simulate differential measurements. Not exactly the safest method, neither for my scope, nor for my body.

Problem, of course, is the high cost of a decent Diff Probe. What do you use? Is a DIY probe practical with something like an Analog Devices isolation amplifier? Better to fork out the $500 ?

[Robert McLean]

Could you explain the safety issue you are refering to? I have done the invert and add thing many times, I am trying to remember now if it was ever on tube amps or if it has always been on lower voltage circuits. If the probe ground clips are each connected to the amp ground then the probes and scope are happy. If my household wiring is done correctly then I am happy. The probes need a voltage rating to meet the peak voltages they will see, but so will a differential probe. If there is some further grounding issue, like you need to see 2 voltages referenced to something to other than ground, then an isolation transformer can be used.

Am I missing something here ?

[EC8010]

What were you thinking of using a differential probe for? I'm tentatively ordering one (5.6kV isolation) for work but I'm not sure where I'd use one on a valve audio amplifier.

[SY]

Phase splitters that are load sensitive? Outputs of amps with the secondary used for cathode feedback?

[EC8010]

But a pair of ordinary probes will allow you to compare the output of a phase splitter, and I wouldn't be surprised if most phase splitters would overload a differential probe.

An amplifier with cathode feedback might have DC on both terminals, but that's what the "AC" coupling switch is for.

Give me a better excuse? (I haven't yet confirmed I want/need this probe costing umpty-ump.)

[zigzagflux]

Wow, you wouldn't find a good place to use a differential probe? I think they're the cat's meow, if for no other reason than I don't consume two scope channels to make a single measurement.

My primary purpose is measuring things like choke voltages, xfmr secondaries, elevated filaments, and the like. These are all differential measurements when dealing with a grounded lab scope. The times I have wanted to compare two points, both of which are ungrounded, I have had to resort to pulling the third prong on the DUT. Not the safest practice, I admit.

Maybe an example. Let's say I have a DC filament supply, elevated 30V above common (ground). The raw supply is choke input, fed to a LM317 regulator. I have some noise on the output to the filaments. I want to see if this is at all related to choke ringing, so I would like to view the choke voltage simultaneously. Both of these test points are above earth potential, and I only have a two channel scope.

For that matter, ANY test point that I would like to sync with another test point can be difficult w/out a differential probe. Things I can sync to the line- fine. But sync to anything else, it's tricky. There are enough noisy circuits I've made that this ability is greatly appreciated.

Also, there is the issue of equal gain and offset in each scope channel when wanting to view a test signal under high gain. I am not convinced I am getting accurate results, such as maybe adjusting balance in a LTP. I also am not convinced that what one sees (noise-wise) with adding two channels is an accurate reflection of what exists differentially. It's almost like the single-ended noise, which may get CMRR-rejected by my differential tube stage, is able to get through to my scope trace because of the compromised test method.

Not to mention the additional high voltage benefits that a scope probe gives me. If I had the choice between a differential vs. single ended HV probe, I would grab the differential any day (equal cost unlikely, of course). My existing 1X/10X probes don't carry a voltage rating, and I am nervous going over 300V- though I do when necessary.

I have no doubt there are tricks and methods to address some of these 'problems' such that a d.p. is unnecessary; to me, life is just easier with one at your disposal.

[EC8010]

Originally posted by zigzagflux
Let's say I have a DC filament supply, elevated 30V above common (ground). The raw supply is choke input, fed to a LM317 regulator. I have some noise on the output to the filaments. I want to see if this is at all related to choke ringing, so I would like to view the choke voltage simultaneously. Both of these test points are above earth potential, and I only have a two channel scope.

That sounds like a good use. I think I'll defend my request for a differential probe...

As for the phase splitter thing, you put both (ordinary) probes on a single point, sum them, invert one, adjust fine gain on one of other channel to get zero. Now move the probes to your differential signal, remove the inversion in the 'scope, and you see the difference.

Do we take it that you have a differential probe at work?

[SY]

Mmmm, you do understand that I like to measure distortion spectra? And that I use scope probes for that task?

I've been eying the AD diff amps myself. I don't need huge DC isolation, but I do need very tight and stable AC balance. The SSM balun chips might be another alternative.

[zigzagflux]

Originally posted by EC8010

As for the phase splitter thing, you put both (ordinary) probes on a single point, sum them, invert one, adjust fine gain on one of other channel to get zero.

What a fine idea. I will do that every time in the future.

Do we take it that you have a differential probe at work?

I wish. My previous job was in AC/DC drives, and they used diff probes exclusively. I was spoiled, now I realize.

Originally posted by SY
I've been eying the AD diff amps myself.

As have I. For the $20-$40 range, they seem like a good starting point, and most of the work has been done. Some input resistors for scaling, maybe some protection. I could certainly build my own on perf board - was hoping someone here might have had a design that was proven effective.

[poobah]

Tektronix used to make diff probe tha was powered by a wall wart... fit on about any scope. You do occassionally see them on Ebay. They either go for 500 and up... as they should... or they go for next to nothing. If you are not in a hurry, and can restrain yourself until "know one is looking", you could pick one up cheap.

I might still have a few smoked ones around here... if your real ambitious.

The Analog Devices solution is pretty breezy too... as long as you can live without bandwidth.