Fixing An Agilent Oscilloscope Power Supply

We should all be so lucky as [Salvaged circuitry], who scored a cheap Agilent oscilloscope from an online auction. Of course, its low price had a reason behind it, the ‘scope didn’t work. At fault was its power supply, the repair of which was documented in the video below.

These ‘scopes have relatively straightforward 12 V power supplies, extremely similar to off-the-shelf parts. The video is an interesting primer in switch-mode power supply repair, as the obvious failure of the filter capacitor and a MOSFET is traced further to the PSU controller chip. We see a new capacitor mounted proud of the board to reduce the risk of heat damage, and then some careful solder rework to save some lifted pads.

The result, a working oscilloscope. Maybe we’d have hacked in another 12 V supply, but given that this is a piece of test equipment perhaps it’s best to stay as close to the original spec as possible. As a parting shot he shows us an equivalent power supply, and promises us a side-by-side test in a future video.

These ‘scopes aren’t as popular in our circles as the cheaper Rigol range, but it’s worth remembering that they also have a budget model.

20 thoughts on “Fixing An Agilent Oscilloscope Power Supply

  1. I got rid of an HP spectrum analyzer and an HP scope with dead power supplies. The scope had a dead house numbered IC that I would have had to re-engineer. The spectrum analyzer had a “black box” switching power supply that I would have had to build from scratch, or draw out the schematic and re-engineer. Some old stuff just isn’t worth bothering with, no matter how much it weighs. Nostalgia ain’t what it used to be.

        1. I run his videos at 1.5 speed, and skip over a lot. But I love to see old stuff, especially that I used to have or work on. No desire to collect or use it though. I’ve enjoyed going to the military electronics museum in Glen Burnie MD, and the radar Museum in London, ONT.

    1. Sometimes, if you are lucky the house nubered IC’s are just common IC’s without the common labeling on them. A long time ago when I did a lot of electronics and my nose was burried deep in databooks a lot I was pretty good at figuring out what mystery IC’s really were. It is interesting how often designers do not stray much from the boilerplate designs in the application notes.

      1. Then again, HPAgilent loved to pepper their designs with custom unobtanium ICs and hybrids. It wouldn’t surprise me if something simple and easy to source externally was subject to NIH syndrome in an HP design.

      2. I only realised that relatively recently… was into a bunch of car phone and usb chargers and despite epoxy blobs and sanded chips, most of them had the same standard part and same circuit.

        Another case was a specialist ISA card, where a chip had a really well glued plaque/label on top of one chip, which came off with a lot of heat and scraping and turned out to be a 74LS13

  2. May I recommend first fixing the mechanical parts before soldering them tight in place. That way there is no mechanical tension on the soldered pads, coming loose over time.

    1. This is a valid point. I originally installed the TO-220 packages because I did not want to reinsert the heatsinks then desolder them again in case a different component had failed. The pads holding in the heatsinks are not very substantial. they are even more delicate than the traces going to the leads of the TO-220 packages.

  3. I could have filmed the repair of my PC power supply Chieftec 850W Proton. The biggest thing was to ….buy that specific large capacitor. Originally there was a 680uF 400V Teapo LE, 30mm in diameter, 50mm in height, that I could find exact lifetime info. Yes, it was available widely from other brands , but when you buy a batch of hundred of them or so…. it had 8000 working hours then (at 100 degrees C) . Otherwise I had to buy a 2000h rated one, which I bought finally. For every 10 degrees of temperature drop, lifetime doubles, so let say it works at 50, then it can work 64000h. And one N Mosfet had to be replaced. If I only had filmed everything I do everyday then I would be famous.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.