EEVblog #655 – Auction Score

Lots of test equipment auction goodness from Precision Mechatronics who went into liquidation.
And is it possible to mod a system PSU with binding posts?
Service manual HERE

Keithley 2400 SMU Source Unit
Agilent 54845A 1.5GHz Infiniium Oscilloscope
Agilent E3646A dual output bench power supply
Agilent 6641A and 6643A DC system power supplies.
HP 3488A GPIB Matrix Switch

About EEVblog

Check Also

EEVblog 1559 – PCB Design: Trace Current Rating

Dave answers a Twitter question: How should I design a PCB trace to carry 80A ...


  1. Hi Dave,

    With regards to the HP6642/43A’s it seems they use the same front panel metal as the 6632’s which I have 2 of. On the 6632 ithere is an option of front binding posts and the service manual for them details the hardware required (take a look it might give you some ideas) – I didn’t bother with the filter components that the official option has, just routed wires from the PCB Molex connector the 6632 has for this purpose to 2 binding posts mounted through the front panel and secured on a piece of blank PCB material I epoxied behind the front panel – the front panel cutouts are too big for binding posts by themselves.
    I can post photos if you need them….
    From the service manual you linkedit seems the 6642/43A’s don’t have a front binding post option and so no PCB connector for them, but should be pretty easy to solder straight to the output terminals inside the case.

  2. Nice score 🙂 Congrats on 2400, it’s a nice machinery. You can update firmware to C32 (which is latest last time I checked on KEI website), it fixes some bugs.

    Would love to see teardown of your’s 2400, would be interesting to compare to [url=]one I have as well[/url].

    Also if you still have those Agilent SMU’s, would be cool to put a comparsion 🙂

  3. Nice score, sure would love to see a teardown of that scope.

  4. Hey dave, good score. I wouldn’t mind knowing what auction house you goto, I’ve been searching pickles since your last auction vid in search for a good cheap oscilloscope and haven’t come across any other then a few vintage ones. Thanks

  5. Teardowns please?

  6. Thanks, will take a look.

  7. Hi again!

    A couple of other points:

    Noisy Fan
    You need to check if the noise is coming from the fan itself or the volume of air being moved – replacing with a more modern fan of a similar rating I guess would not solve the air flow noise. If the problem is the fan then it could be noisy bearings. Can’t remember what kind of fan my 6632B’s have, but other HP equipment of this period used high quality Papst which could have their bearings replaced (remove the sticker, take off the split washer and they should fall to pieces). Clean, replace the standard bearings and they should be as good as new and a lot better solution that a new fan with a sleeve bearing.

    Front panel
    If you are going to but sockets on the front, the easiest way is to remove the front panel. Whilst you are in there, it’s a good idea to take the display board out and check the rotary encoders – on both my 6632’s they were very floppy caused by the metal tabs holding the top half of the encoders to the base becoming bent through years of use or by being knocked in transit. Just squeeze them back so they are tight and maybe put some epoxy on them – they should feel quite smooth with very little wobble afterwards..

    I’ve attached some pics of the front panel binding posts – the closest in style to HP ones I found were from Radio Shack..

  8. ooops..seems that multiple copies got posted – can you delete the duplicates?

  9. Thanks Dave. Its been wracking my brain.. its funny though, that was the other Auction site that came up in a Google search. I didn’t think to check I thought it was all recon sale stuff like factory seconds.. Thanks again mate..

  10. Hi Chris,

    just some notes – according to the HP manual, the 2nd pair of wires to the front panel are intended to be sense lines, which I find, hm, sensible 😉

    To enable this, you need to short Sense+ to Out+ and Sense- to Out- on the back panel screw terminal block, and clip out some zero-ohm resistors next to the Molex connector (see schematics and my picture).

    I mounted the front panel jacks in a similar way to yours, only I made the board holding them a little larger and cut a small groove into the plastic standoffs holding the control panel PCB, giving the jacks some extra mechanical stability independent of the epoxy.

    BTW, In my 6632B, the display had started to fade, and I thought the VFD itself had gone bad – alas it was just a capacitor, so if anyone ever has a similar problem, look out for the little brown critter in the picture below.


  11. Hi Pete,
    Thanks for the info – of course you are completely correct and it would have been strange if the sense wires couldn’t have been extended to the front panel – no idea how I missed cutting the resistors – on page 78 of the service manual it is stated as clear as day!

    One of my 6632Bs is badged HP, the other Agilent – the HP one does have a much dimmer display which might be age related, but then again…think it is time to open both of them up, cut the resistors and check that cap!

    Good idea to add some extra stability, but I scored the surface of the front panel and pcb material, poured epoxy into the space then put the pcb on top and epoxied around – don’t think it is going to move any time soon 🙂

    Seems I’m not the only one with problems posting multiple images – looks like when you delete they are not really deleted…

    Many thanks,


  12. OMG Dave, you’re the one bidding against me for the keithley SMU’s. Nice wins by the way.

  13. And here I was thinking it was a 2nd hand test gear dealer!

  14. Just sense lines doesn’t make sense (pun intended!), as the traces and number of vias are clearly for high current. Also, the official options says something like slightly degraded CV accuracy which makes sense.

  15. Hi Dave / Pete,
    From the schematic it seems that when the resistors are cut and the front panel output is installed, the rear output can still be used but then the current goes first via the front output, so the “sense” wires in this case need to be able to carry the same current. See my pic..

  16. awessome thanx for showing this to us.

Leave a Reply

Your email address will not be published. Required fields are marked *