[REPAIRED] Agilent 54832B CH1 vertical errors (flatness)

[Jwalling]

I purchased an Agilent 54832B off the 'Bay that was listed as fully functional ('natch)
but fails the calibration on CH1 (vertical errors, flatness)

I wrote an email to the seller stating:
"Hello,

Unfortunately, this scope does not pass it's built-in calibration routines, it fails on CH1 vertical, flatness. With a 50 Ohm terminator connected to CH1, I can get the scope to trigger over a 800uV range. On Ch2, it will only trigger over a 30uV range which is about right. CH1 is very noisy which I believe is causing the failure.
See the pictures, it was failing in 2015, and still failing now."

The seller wrote back that he's sorry, and how competent his technician is, etc. and has given me permission to break the seals and open the scope to see if I can repair the channel. Given the fault, and before I dive in, has anyone ever had any success repairing this particular problem? I'm hoping it's not the input amp hybrid... but perhaps a noisy OP-AMP or de-coupling cap?

EDIT: added picture.
Thanks.

Jay

[Jwalling]

Ok, so I've opened the victim up and what do I see sitting on top of the CH1 attenuator?

A FREAKIN' CAPACITOR! 

I'd like to know just how the hell this ended up there, and how on earth it managed to stay stuck to the cover of the attenuator through shipping from Denver CO to near Boston MA.!!!?? 

Jay

[Jwalling]

Twas' not to be so simple. Note the highlighted components between CH1 and CH2. The larger resistor is actually cracked as well as burnt.   Looks like some clown input-ed something beyond the ratings of the scope.
I hope the Hybrid is OK. The signal for the most part looks fine with the aux out connected to it, as well as with a 1GHz sine wave, so hope is there.

Luckily, I have some parts boards from a 54831B that I bought for $40 each a few years ago. I'll replace them as well as that small SOT device marked G21. Hopefully that little guy protected the hybrid!

Hmm. Maybe I'll hold off on the SOT device for now, it's not shorted.

Jay

[Jwalling]

Success!    I never did figure out where that damn cap was from, I looked over the entire ACQ board.

Jay

[Jwalling]

I should also mention that the seller has graciously offered a partial refund if I fix it. 
He said that he would much rather give me a small partial refund for fixing it than have to issue me a full refund and send me a return label.

I paid $1250 for it. But here's the thing: If this scope had been at least minimally tested by running the built in calibration routine which only requires a length of coax cable and is simple to do, it would have been listed as "For parts or not working" and never would have fetched $1250. I might have only paid $500 or so and parted it out if I couldn't fix it.

So what would be a fair refund amount?  I'd love to hear any opinions.

Jay

[TheSteve]

I'd be expecting at least $400.00 back. If possible let the seller make the first offer. Detail what you all had to do to repair it and the time it took.

[siggi]

I might have only paid $500 or so and parted it out if I couldn't fix it.

I'd tell the seller this and then perhaps argue to ~split the difference - if the seller does a any amount of trade, it won't hurt to get some goodwill. Comes out to as near as $400 refund as makes no difference. Alternatively you can price your time to this repair and ask for ~2* that amount back.

[Jwalling]

The seller refunded me $300.  That seems pretty fair to me, and I left him great feedback. It also gave me the knowledge to hack a 54831B into the 1GHz 54832B which is worth a lot to me.  If I had sent it back, I would not have learned this - I value knowledge very much.

Thanks TheSteve and Siggi.

Jay

[TheSteve]

Always nice when things work out and the mod is awesome.