Differential Probe Suitable For Analog Audio Design

[GrievousAngel]

Small signal diff probes can be extremely expensive . . . way over my budget!

Which diff probes have others used that works well for audio performance measurements?

What do y'all consider the min or max specs for meaningful audio measurements?

Has anyone tried the ~$70USD China active diff probe (in metal enclosure) w/ +/- 10V I/O range? Specs appear to be reasonable for audio measurements . . .

BTW . . . I plan to use with the Siglent SDS804X HD or a PicoScope 4262 16bit HiRes (recommended for audio etc below 5MHz bandwidth).

Later, Billy

[David Hess]

I have never noticed that any particular differential probe worked better for audio.  Do not count on good distortion performance through a differential probe.

For audio, I mostly either used two channels in invert and add mode for differential measurements, or I made a simple differential probe on the spot with some operational amplifiers.

[TimFox]

For audio, you might consider using good transformers, e.g.  https://www.tubesandmore.com/products/transformers_chokes?filters=2341a3027  at $48 or $70.
Obviously, not good for DC or RF.
I have had good results using these to measure distortion of push-pull (balanced) outputs, so long as I pay attention to the power ratings of the transformers (distortion increases at higher power), sometimes needing a well-balanced differential attenuator before the transformer.
(Full specifications are available by clicking on that web page.)

[coppercone2]

I also have a transformer, the old HP ones.
HP 11004A

[coppercone2]

How do you feel about a
HP 11005A

https://www.ebay.com/itm/235572002985

[CaptDon]

As a previous poster mentioned, simply use two channels of a scope and configure for differential measurement!! Works great on a 4 channel xcope since you still have two remaining channels for non-differential use!! I have a Tektronix 500 series module that was designed for differential use complete with band width limiters and D.C. offsets per channel. Good up to 1MHz. I very very rarely need a diff measurement in audio work!!

[David Hess]

I have a Tektronix 500 series module that was designed for differential use complete with band width limiters and D.C. offsets per channel. Good up to 1MHz. I very very rarely need a diff measurement in audio work!!

That would be a 1A7, 3A9, or 3A10?

That design lived on in the 5A22 for the 5000 series oscilloscope mainframes, 7A22 for the 7000 series oscilloscope mainframes, and AM502 external amplifier.  They are excellent for audio, and one of the best choices if not the best choice, but have been out of production for decades.

I myself have at least two working 7A22s and one working AM502.  Putting together a working AM502 would probably cost at least $300.

[TimFox]

There are also free-standing PAR (Princeton Applied Research) preamplifiers left over from the good old days, often available on eBay.
See https://testequipment.center/Products/PAR-113  and click on the specifications link.
There are other versions, including battery operation, and some used as plug-ins on PAR lock-in amplifiers.

[CaptDon]

Looked in the module drawer, it is an AM502. I wish the bandwidth would have extended to perhaps 10MHz. I bet that module is 50 years old but still works perfect. It is a must to clean the switch contacts from time to time although they stay in good shape the more the module gets used. It is really versatile!!

[David Hess]

Looked in the module drawer, it is an AM502. I wish the bandwidth would have extended to perhaps 10MHz. I bet that module is 50 years old but still works perfect. It is a must to clean the switch contacts from time to time although they stay in good shape the more the module gets used. It is really versatile!!

It should be possible to remove the 1 MHz bandwidth limit, but I took a quick look at the schematic and there is no obvious limiting circuit.  The high frequency cutoff network removes all capacitors to achieve 1 MHz, so that is not it.  Would have to do some real calculations or measurements to find it.

Linear Systems makes some low noise JFETs which might improve the noise performance.

[PChi]

I built a low voltage differential probe using common mode feedback to meaure the currents in a Class B output stage. After trimming the CMRR at low frequencies was better than 72 dB but as expected degraded with frequency.
At 10 kHz = 66 dB.
At 100 kHz = 56 dB
At 1 MHz = 50 dB
At 10 MHz = 45 dB
Gain 1 up to 10 kHz.
Gain about - 3dB down at 1 MHz.
It could do with some more work to improve performance but was adequate for what I wanted to measure.
I have attempted to add a schematic.