My recent scrounge trip to California (the Bay Area) netted me, among other things, a nice-looking HP 8751A network analyzer. This was one of the units HP apparently did in Japan for either design or manufacture, not sure which.
Anyway, the symptoms pretty much screamed 'Power Supply.' When turned on, the fan wouldn't run, and the thing would make a decently loud CLICK (as from multiple relays energizing) accompanied by a soft humming. Sometimes, the screen would show its 'performing self-test' message, but it would just sit there forever. Digging into the thing for the first time, without the benefit of any type of documentation or schematics, was ... educational. Yes, that's a good word. My initial conclusions and repair efforts were:
1). Assume the electrolytic caps on the pre-regulator board have gone bad (it turned out, in retrospect, they were probably OK). Replace all, under the idea of 'shotgun' troubleshooting.
2). For some odd reason, HP made extensive use of carbon composition resistors on one of the two boards in the pre-regulator section. The other was all the more up-to-date film type (carbon or metal, not sure). Testing showed several of the carbon comp resistors had drifted significantly in value. Replace all comp-type with film types on general principles.
3). Replace fan. Obviously (though this led to some other interesting findings).
Get everything put back together, turn it on. Same problem!
Poke around some more, this time with the help of some schematics Simmconn Labs was nice enough to provide (thanks, Xu!)
Next discovery: The FAN_CTL lead can cause the entire instrument to immediately shut down on power-up. Fortunately, the provided documentation also revealed a test point which, when grounded, would disable the shutdown circuit. Lacking anything else to try, and feeling particularly daring, I grounded said point, donned safety goggles, leaned off to one side and pushed the power button. Amazingly enough, the unit powered up normally, passed its self-test and showed its default display on the screen!
My next assumption was that any fan with a tachometer output would work just fine as a replacement. Imagine my surprise when I installed such a fan and powered on -- Only to have the unit buzz like an overexcited beehive!
Then it hit me. The thing was buzzing because the FAN_CTL lead was expecting a steady logic state, not a pulse train like a tachometer output gives. A little more research on the original fan confirmed that HP had not used a type which was readily available (of course not!) No, they picked one which gave its 'normal operation' signal as an open-collector output pulled to ground when the rotor was spinning.
I didn't have one on hand. Nor could I easily find one from my normal supply channels, or the surplus arena. I ended up installing a basic 24V fan, and soldering a short jumper between the FAN_CTL pin and the ground pin, both on the fan connector.
Reassemble, power up, everything works including the S-parameter accessory test set. Given the end result was a terrific network analyzer setup for a total investment of around $400, including repair parts, and an equally terrific education in how to troubleshoot power supply issues in this class of HP instrument, I think I came out of it pretty well.
I recall another post on the forum about another 8751A with the same symptoms. I hope whoever had that unit, if they're still working on it, finds this post helpful.
Keep the peace(es).
Home
Help
Search
Login
Register
