Agilent E3631A

[k1mgy]

Can anyone with an Agilent E3631A confirm the part installed as U5 in the A to D section? 

The manual does not list a part number for it. 

In my semi-broken E3631A (ebay) U5 has the same part installed as U4 (74HC4051).  I don't believe U5 should be that part.. so it looks like a repair gone wrong.

To access this, slide the unit out of the case, remove the two screws holding the hinged section, and open the top board.   U5 is near the back, to the left of the 5V regulator.

[k1mgy]

Whoops.  Seems when ever I ask a question, the answer appears!

https://doc.xdevs.com/docs/HP_Agilent_Keysight/E3631A/Unit_2/img/001/DSC_4788.jpg

It's a 74HC4053D 3x 3 ch mux/demux

.. and some IPA, a microscope, and just the right angle reveals that the part installed in U5 is indeed the 4053D.  The 3 sure looked like a 1!

[ArthurDent]

Have you watched this repair video of the HP e3631A ??

[k1mgy]

Yes, it was very helpful.

Upon further examination, the COMP pin is toggling, consistently regardless the position of the LIMIT button, and this makes it further down stream.  FLASH, however, is pulled to the negative rail at all times, and CR3 keeps it at -0.65 V going into the microprocessor (thank goodness).  So I think my problem is in the FLASH chain.

I found several resistors in the ADC that were open circuit.. which I suppose is not all that strange for a unit so old.  They were all the 0.1% types with green colouring.  May have been a bad batch?  The rest seem fine.. but haven't checked the caps yet.  Doubt bypassing would cause this kind of failure.

It seems probable that when the LIMIT switch is activated, I'm reading on the front panel the *desired* output.. and when LIMIT is NOT activated, the front panel displays the actual sampled voltage/s.  Makes sense...