Differential probe

I am doing some noise measurements up to 30 MHz but I need a differential high impedance input, instead of the 50 ohm unbalanced the spectrum analyzer has.

There are differential probes available but they cost more than I paid for my spectrum analyzer, which makes DIY an attractive alternative.

An instrumentation amplifier would be the easiest way, but I havent worked with them for a very long time, so I don't know what to expect from the modern chips. Maybe someone here can educate me (before I go to a discrete solution).

AD8253 ist the fastest IA I'm aware of, but still short of your spec. I've recently designed something similar and used three ADA4817-1 to build a fast instrumentation amplifier.

Samuel

Might be best to roll your own from two or three op-amps (as Samuel hinted); I believe I saw a Design Note in one of the online trade mags when I was looking for something else.

Be aware that any probe, active or not, will do its best to find a way to get itself blown up; designing protection circuitry that doesn't compromise bandwidth or max loading requirements may well be one of the hardest parts.

If you don't care (much) about range down to DC you might consider having the first two op-amps in the classic 3-amp configuration driving an RF transformer.

JDB.

We bought a barely used active FET probe on eBay at work. I realize this involves money. We got a $1000. probe for about $200. Then we bought a "special" power supply for the probe and the output of the power supply connects directly to our spectrum analyzer. The power supply is made by Tektronix. It was like $160. or so. Making a differential measurement at 30MHz is not that easy. DW.

Thanks for the comments.

Samuel Groner said:

I've recently designed something similar and used three ADA4817-1 to build a fast instrumentation amplifier.

Click to expand...

The ADA4817-1 seems intresting.

To avoid reinventing the wheel, I searched old HP/Agilent service manuals, but no luck so far. If anyone knows better, please let me know. Or if Samuel wants to share his design...

jdbakker said:

If you don't care (much) about range down to DC you might consider having the first two op-amps in the classic 3-amp configuration driving an RF transformer.

Click to expand...

Actually, I thought that too. But where to find a transformer good enough?

Or if Samuel wants to share his design...

Click to expand...

It's hardly worth the time it would take me to make a PDF an upload it. It uses the standard three-opamp IA topology, keeps feedback network impedances low and trims CMRR of the second stage both resistively and capacitively. As an input clamp I used 1N914B diodes to pre-biased 3.3 V zeners.

Samuel

audiox said:

jdbakker said:

If you don't care (much) about range down to DC you might consider having the first two op-amps in the classic 3-amp configuration driving an RF transformer.

Click to expand...

Actually, I thought that too. But where to find a transformer good enough?

Click to expand...

Please qualify "good enough".

My first stop when shopping for RF transformers and the like is http://www.mini-circuits.com/ .

JDB.
[I suppose that the DSL market has more than a few interesting transformers as well, but I've not done much work in that area]

Samuel Groner said:

It's hardly worth the time it would take me to make a PDF an upload it.

Click to expand...

I thought that you have used countless hours with that as usual...

Here is something (block level only)
http://www.home.agilent.com/agilent/redirector.jspx?action=ref&cname=AGILENT_EDITORIAL&cc=US&lc=eng&ckey=1000002794-1%3Aepsg%3Aman&nid=-34022.536880396&pid=1000000317%3Aepsg%3Apro