Low Voltage Differential Probe

[RF-Engineer]

Can anyone recommend a low-voltage differential probe for under USD $1K?  Does such a thing exist?  100 MHz is fine.  I'm shocked to find that most new differentials are USD $4K and above.  I already own a Micsig HV differential probe but I would like one for small signal measurements with 1X and 10X capability.  Used probes look pathetic, most of which are missing all their connection accessories.  I can use up two scope channels with an invert-add function, but would like the convenience of a single probe. 

Paul

[tautech]

Can anyone recommend a low-voltage differential probe for under USD $1K?  Does such a thing exist?  100 MHz is fine.  I'm shocked to find that most new differentials are USD $4K and above.  I already own a Micsig HV differential probe but I would like one for small signal measurements with 1X and 10X capability.  Used probes look pathetic, most of which are missing all their connection accessories.  I can use up two scope channels with an invert-add function, but would like the convenience of a single probe. 

Paul

Some suggestions here:
https://www.eevblog.com/forum/testgear/reasonable-but-low-cost-differential-probe-for-siglent-sds1104x-e/

[nctnico]

I slapped one together using an AD8421 instrumentation amplifier chip. It is not very difficult to DIY.

[KE5FX]

I slapped one together using an AD8421 instrumentation amplifier chip. It is not very difficult to DIY.

You're not selling the LMH3401-based one any longer?  Seems like that's a good fit, except for the lack of 1x support.

[nctnico]

I slapped one together using an AD8421 instrumentation amplifier chip. It is not very difficult to DIY.

You're not selling the LMH3401-based one any longer?  Seems like that's a good fit, except for the lack of 1x support.

I'm still selling that! However the DIP1400 (LMH3401 based) has a division factor of 10 when terminated into a high impedance input and the input impedance is 2k. I have used it for low voltage measurements but I think a different approach is more suitable for sensitive signals.

[Keith956]

You could make your own. I made one of these and it works very well:

https://hackaday.io/project/169390-a-10x-100mhz-differential-probe

[Hydron]

May also be worth looking into the AD8129/8130 amplifiers. Run at +/- 12V they have a reasonable common mode input range, though the differential input range can be a little limiting (that sentence should give a hint that they're not a typical opamp...). I believe the EEVBlog HVP70 uses two of them (one each of 8129 and 8130 for the different gain ranges).

[David Hess]

The AD8129 or AD8130 would be my choice for the simplest high frequency design however their input voltage range is small enough that input dividers will likely be required.  The prototypical Tektronix P6046 had an input range not much larger than that and was also commonly used with a x10 input divider.  If you must do without the input divider, then a much more complicated design more along the lines of the Tektronix 7A13 is required and it only supports an input range of +/-10 volts with no input divider.

[Hydron]

If you can eat the noise penalty of using a 10x divider followed by a 10x gain stage, then you can do +/- 100V common-mode with +/- 5V differential range by using the AD8129 in 10x gain mode (minimum stable gain, with a 200MHz bandwidth!) after a 10x input attenuator.
If you put a AD8130 in there instead (or have both, and use the power-down pin to switch between them - I think this is the HVP70 method) then you can raise the differential input range by 5x to +/-25V at the cost of an overall 10X division ratio (or a somewhat smaller ratio with a bandwidth hit if gain is increased, but the AD8130 does have an output swing limit to consider as well).

Building something like this (potentially using normal 10x probes, matched as well as possible, as input attenuators) has sat on my to-do list for quite some time, but it's priority hasn't been high enough to have a realistic chance of actually happening soon unfortunately. I already have a couple of the Micsig probes for higher voltage applications (though one is playing up ) and simply haven't badly needed a low voltage version yet.

[nctnico]

Building something like this (potentially using normal 10x probes, matched as well as possible, as input attenuators) has sat on my to-do list for quite some time, but it's priority hasn't been high enough to have a realistic chance of actually happening soon unfortunately.

That is exactly what I did with the AD8421. Because it has a very high input impedance an (approx.) 1M Ohm resistor across the input together with some capacitance and 10M Ohm resistors to ground allow to use a standard 1:1 / 1:10 probe.

[David Hess]

I read the AD8129/AD8130 datasheet wrong.  A +/-10 volt input common mode range is possible without input dividers so that would be suitable for measurements on logic boards without input dividers.

Getting enough differential input voltage range then becomes the problem with +/-2.5 volts maximum and this reduces the maximum input common mode range although this should not be a problem in this application.  That suggests using a x2 differential attenuator on the input to boost the differential input voltage range so logic families up to 3.3 volts are accommodated or x3 to support 5 volt logic.  The gain can be adjusted at the output to make a convenient overall x10 attenuation factor.

For more general purpose use, I would consider the distribution of attenuation used by the old Tektronix P6202A active probe which uses x5 common mode attenuation at the input with x2 attenuation from the double terminated 50 ohm output for a total of x10, but noise would be marginally higher.  The input range would be 50 volts common mode and 12.5 volts differential.