Just remeber to check all the 'Backdates' at the rear of the manual if you find something that doesn't match the schematic or parts lists. Some HP semiconductors have replacements available from
formatting link
You can make a good ax in four hours. You can sharpen a good ax in
15 minutes.
I need a front end loader, after my shop roof was damaged. I've been using a coal shovel to scoop up service manuals destroyed by water. Next will be a rake, folled by a scraper & a push broom to clean uo the debris where termites ate into the damp paper. There were a bunch of new boxes of greenbar paper destroyed, as well.
Hey, I like all that old HP, Tektronix, Fluke, Boonton, etc. equipment. It is still serviceable, unlike all the new stuff. And pennies on the pound. Although it's lots of pounds.
I'm sorry tm, with an attitude like that, you won't be allowed to continue posting in this group. You need to get an attitude to hang here. Oh yeah, you need to never trim any posts too! lol
I didn't realize he was being literal. But then maybe part of that was getting the grinding wheel in really good condition. lol
That has got to suck! Not as bad as a fire though. With those you get fire, smoke *and* water damage, the tri-fecta! I saw a web page once where a guy had a terminal (it was a long time ago) start a fire. He had all sorts of problems and gave a lot of good info on how to protect things. Mostly it was about preventing smoke and water damage.
I have to admit I haven't had time to look at the schematic of this unit, but the more I think about it, the less I think it would be the diode function generator (DFG). That just turns the triangle wave into a sine, right? This fault reverses the direction of the sine wave in mid course and changes the frequency as a consequence.
But then I am not that familiar with the operation of the DFG in any detail. If the triangle wave were to change and not go as far negative as it should, that would produce the waveform I am describing, right? In fact, you could consider this to be a sine wave that is clipped in time rather than amplitude. It has a phase section clipped out of it, maybe 210° to 330°. Without disturbing the circuit at all, the point of clipping changes on its own a bit too. So a freq counter won't give a stable display. I guess I should look into the time base circuit... after checking all the supply voltages of course.
Weird, I understand this is a basically analog device (Not a DDS). Then it looks like some inverter falling out of the signal chain. Intermitent power to some section.... maybe? Does it change if you load down the output? Have you opened it up and wiggled things?
No, I haven't loaded the output. I have probed and prodded with my fingers. I think there was a point where it had some impact. I just haven't taken the time to look at the docs yet. I'm thinking the ramp generator is messing up and triggering the switch to the up ramp too soon.
I didn't write them all down, but they are all within 0.2 volts of nominal and I don't see any ripple on the supplies.
On the other hand, the fact that the problem is only on the two upper ranges indicates it is within the "High Frequency Level Detector". I measured some voltages there and the lower half seems to be way out of whack. There are two symmetrical circuits to generate what would seem to be a reference voltage for the +-5 volt level detector. But I'm not sure that is how it works really. According to the schematic notes, changing the switch from the lower ranges to the two higher ranges changes the "references" from +10 / -10 volts to -10.6 / +10.6 volts. What?
Something else I don't get is one transistor in each half is reverse biased, yes, the CB junction is reverse biased, but the BE junction is also reverse biased by 6 volts! The emitter resistor is connected to
+10 volts, the collector is connected to -10 volts and the emitter is supposed to be at +10.6 volts! What??? The transistor is a PNP and the base is supposed to be at +16.8 volts. How is this supposed to work? Have I forgotten my transistor notation? PNP is arrow pointing to base which should be more negative than the emitter, right?
Anyway, the half that is whacked out has some random fluctuations that I can see on the scope which don't appear on the supplies. I figure one of the components is bad but I can't trace it to anything in particular. The circuit all interacts somewhat and the "noise" signal appears through nearly all of it.
Parts of the large schematic pages are missing. I assume somewhere they tell you the conditions for taking the measurements, but I haven't found them yet. They do have a troubleshooting section that clearly tells you to look in this circuit.
All the conditions for the measurements are on PDF page 71, schematic 2, at the lower right side of the page. Look for all the waveforms just to the right of the notes section.
What you describe cries out leaky capacitor. Since it is only doing it on the upper two ranges, that sure points to that circuit.
All of the transistors in that circuit are 2N3904 (N) or 3906 (P) so a shotgun approach might be worth it in case it is a leaky transistor.
Thanks. My copy doesn't have complete pages of the schematics. They are cut off at the bottom and when they span two pages there is a gap in the middle. I think I got this from the Agilent site. Is there a better copy available somewhere else?
BTW, to make it more viewable I rearranged the schematics a bit and rotated the vertical pages. I see notes on page 70 of 87 and this page also has assemblies A2 and A3 and waveforms 12 through 16. Is this the page you mean?
It includes notes 1 through 3 which is cut off a bit. I don't have notes 4 and 5 at all. On my page 68 I see notes, but all of the data on the right is cut off and note 5 is missing.
Wait! I found an intact set of notes on page 77. I will recheck the DC measurements with the settings indicated. This may bring the good section into alignment with the schematic values, but I'm sure the bad half will still be bad. Problem is I just don't know how transistors are expected to behave when both junctions are reverse biased like the schematic says they should be.
I think your shotgunning idea is good, I just don't relish the idea of taking this thing apart, but I guess it can't be too bad. When they have front panel switches soldered to the board it makes me twinge. I'm probably just being overly worried, I expect it is made to be repaired as well as assembled, unlike today's stuff.
Heck, I might want to get a second one so I can use it to make a sweep generator!
Join the Yahoo group for HP Test Equipment. LOTS of engrs and techs have tons of experience with the "boat anchor" HP's. I'm sure more than one has been down this road with your particular gen.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.