N2796A active probe response problem

[EE-digger]

I just picked up an N2796A (2 GHz) single ended active probe to use with a 3000T series scope.  Looks almost new, all accessories included.

The probe passes calibration and a quick check of frequency response on a 54846A scope looked good, compared to an 1152A probe.

The problem is that low frequency response (65kHz shown) has significant overshoot as shown in the magenta trace.  The yellow trace is an 1153A connected to the same signal.  The overshoot become less significant beyond 1MHz or so but I'd like to depend on the waveform accuracy for LF work.

Under the AutoProbe box label are three holes.  Two have pots beneath them.  The one closest to the cable appears to be offset and adjusted well.  The one closest to the scope showed very little effect so it may be a supply adjustment.  It had only slight vertical offset effect and no impact at high frequency or low frequency.

Is there any hope for correcting this overshoot?  On these newer probes, are any response characteristics stored in the probe so that perhaps only a re-programming may be necessary?

 

[EE-digger]

Bummer, Keysight says the probe is bad, no adjustments for this response, as I suspected.  $1089 for a swap   

Just another reminder not to buy from Ebay without a return option.  A dead probe from a high tech dumpster looks just like a good probe.

[TheSteve]

Bummer, guess there will always be more deals on probes.

[wraper]

Just another reminder not to buy from Ebay without a return option.  A dead probe from a high tech dumpster looks just like a good probe.

What do you mean by "without a return option"? Anything can be returned, regardless of what the listing says. If seller does not agree, file "not as described" claim.

[wraper]

Just another reminder not to buy from Ebay without a return option.  A dead probe from a high tech dumpster looks just like a good probe.

What do you mean by "without a return option"? Anything can be returned, regardless of what the listing says. If seller does not agree, file "not as described" claim.

Not if the item is listed under for parts or not working.

If it was sold as not working, of course you cannot claim INAD with a reason that it's not working. However you can still return it as not as described nonetheless. "No returns accepted" in practice does not mean anything.

[EE-digger]

You are all correct.  This probe was described as "used" and I verified with the seller that he would take a return if it did not function properly.

Hopefully it goes smoothly.

I was really tempted to keep it in spite of its flaw but as I tested bandwidth in earnest this evening, I discovered that it is messed up.  It has a number of peaks and valleys and was about useless by 1.3GHz.  The 1152 probes sitting next to it look like little gems in comparison.

Since the probe had no knife wounds, bullet entry holes or tire marks, I have to wonder how it died.  I'm suspicious of ESD discharges.  I can envision varying levels of discharge taking out chip or discrete components.  Offhand, I don't know what the ESD tolerance is on this family.

[TheSteve]

My N2795A.

Edit - btw none of my other active probes show any overshoot. I suspect it has to do with the front end damping. I've also never noticed it before as I've never tried using an active probe at such low frequencies.

[EE-digger]

Thanks Steve.  I was wondering how your probe looked.

When I have my analog hat on, I've used active probes many times for low frequency design because impedance is high down there, while capacitance is low.  I was attracted to the 2796A as it gives a decent input and offset range, plus high impedance.  And it also gets the scope ready for the future 

Now I'm thinking about going back to the Tek probes.  They are cheap and plentiful compared to Agilent stuff.  My P6245 probe heads are half the size of Agilent's.

I do like the 1153A however.  The one I got was complete with attenuator heads, etc.  I have a Linear Tech LTM8049 eval module which I milled down to the core circuit.  It will fit nicely into the 1153A AutoProbe box and give the dual 5v to 6v boosts it needs.

[TheSteve]

I picked up an 1158a for cheap a while back, no accessories though, still waiting for those to pop up at the right price. It's a great size. It requires the probe tip to have a damping resistor in it or the waveform will over/under shoot. My 1134a also wants series resistors. Yes, I have a problem collecting probes I'll never use...

[EE-digger]

I had an 1134A at my previous office.  One of the cables broke on the browser head and I discovered how much two little pieces of coax can be worth.

Check your PM for a guy who has various pieces.  I think I paid $60 each for the little blue resistor tips.