HP 33120A repair.... I need help [EPILOG]

[Citizen]

Hello

I bough my  HP 33120A from ebay for about 130 Euros, it was sold "as-is", and i got it today.
It is not working properly.

Let me describe the problem:

33120A is switching on without problems
 Self-Test: PASS
The menu and display are working properly.
The frequency is adjustable as it should be.
Sync-Ouput seems to work just fine  and  follows the output frequency pretty accurate.

 The main output is srewed. Instead of desired frequency i get  permanent   -17 V.
If i try  to andjust the output amplitude the FG crashes randomly: display blinking and doesnt react to anything.
In menus no error is shown.


I am complete newbie in repairing thing, but i have everything i need : oscillocope, multimeter, soldering station:)


Does anyone has an idea what is wrong with this device, or where should i start my journey?=)

I am also looking for HP 33120 schematic, but cant find anything. Can someone pls  post a link?



I am really thankful for any help:)

[c4757p]

Sounds like the negative (PNP) side of the push-pull output amplifier is blown. Shorted, perhaps, and upsetting the power supply (hence the crashes). Look for large-ish transistors near the output and test them out of circuit.

Disclaimer: I've never had this particular function generator, haven't seen inside it, and haven't looked for a schematic. They're all reasonably similar, though.

[c4757p]

Schematics are in the service manual here. Second result on Google for "hp 33120a service manual"! How did you search for that - reference section at your local library??

[Citizen]

Ok, i have done optical examination...

Some things look strange.

1: is it a fuse? And if it is one, why the F..k is it shorted??

2: I found some crusty resistors(?) the  doesnt look pretty healthy at all.Or is it normal for an old circuit?

I have to go now, i will come back in few hours, and measure those bastards.

[c4757p]

Now that you have the service manual, don't just speculate. Find anything that is questionable in the schematic and figure out what you should expect from it and what could cause failure. Even better, if there are troubleshooting procedures, follow them (I'm on my phone so I can't be bothered to check).

[benemorius]

Give the service manual a good long read. An HP manual from that era should still have some useful troubleshooting intuition to glean.

[Paul Price]

 The easiest way to kill a signal generator is to damage the output circuit, check out this area of the output amplifier.

[Citizen]

Ok, i measured the R's  around the damaged output area, here are the results.

Most roasted resistances  shows 35 kOhm  instead of 100 Ohm.
As shown in attahced PDF-File,  all measured values differs from the datasheet. I think it is mainly because they  are all somehow parallel to each other. So now i will calculate according to he schematics, if my measurments are correct.
 

I just dont understand how can you burn a FG like tah, Why the hell you should apply  high voltage on the OUTPUT??

Moreover i think i will replace  the fuse, now it seems to act like jumper:)
Anyone knows, if i can use simple "very fast fuse" instead of original agilent one?

[free_electron]

allright. looks like some dillweed really f-ed up that machine.

that fuse is  0.5 ampere fast blow signal fuse from littlefuse. if you inject DC into the output the fuse would pop , protecting the output.
but some dodo clearly soldered a couple of copper strands across it ( most likely because he blew it a few times and did not have the right fuse )

so now, he fried the output attenuator.

things you will have to replace :
- all resistors in the output attenuator. these need to be KOA resistors or you will never get amplitude flat in the high range.
- the aromat relays in the attenuators

but , before you even start with that you need to fix the output amplifier.
most likely the two big thransisotrs with the circular heatsink on em are fried....

[Citizen]

Ok, seems like  16 Resistors are dead...+ I think those  two transistor are also dead. At least one of them gives short with my diodtester.
So i will buy 2 of them from ebay, and theni will  look for resistors.

[Citizen]

So, if anyone  wonders what happened to this  FG, here are my results.
 I had to replace both Motorolla transistors Q713  and Q178. Then i really  tried to desolder those fried resistors. But the problem was, that most of them were burned in the PCB, so it was almost impossible to remove the chip, and not to damage the trace.
But somehow i managed to  rescue some pads.  On places where pads where removed completely, i just soldered   resistors sandwichwise,they  are parallel anyway.

For one resistor i  had to  use jumper...

instead of 69 Ohm i used 69.8 Ohm.

One more thing: i had to remove the protection diodes CR801 and CR802, because one was dead. The device  seems to work  without problems. Do i need to replace them, or is it not necessary?

All in all the result is satisfying. Even if it looks ugly, the frequency is very accurate, all functions  are working  great. The only thing is, that the amplitude is  about 10% off (except Sinewave). Maybe some calibration will help there.

I will attach some pics, but pls dont shoot, it was my first repair attempt:)

[SeanB]

Put the diodes in, and get a pack of spare fuses as well, and keep them in a sealed package with a desiccant sachet inside the meter. Hard to not peel tracks with having charred remains there, just clean up the board with a fine brush and alcohol, and trim where you cooked the relays with the soldering iron.

[sorbik]

I just dont understand how can you burn a FG like tah, Why the hell you should apply  high voltage on the OUTPUT??

I've seen techs put the output of a 5700 meter calibrator into another 5700 thinking it was a 5790 AC measurement standard (they look pretty identical). Maybe they thought it was a fancy... function meter.. with bnc inputs?

[Citizen]

I somehow cant find a replacement diodes. The datasheet says they are
DIODE SI PN SOT23 100V 750MA

But i cant find anything i google

Can anyone link me the  manufacturer part number?

[arekm]

I somehow cant find a replacement diodes. The datasheet says they are
DIODE SI PN SOT23 100V 750MA

List of parts says:
CR801-CR802 1901-1386 2 DIODE SI PN SOT23 100V 750MA 28480 1901-1386
where you google for agilent part number 1901-1386 and in doc
http://cp.literature.agilent.com/litweb/pdf/5989-5867EN.pdf
you can find that this is:
1901-1386 DIODE-SWITCHING 100V 1A BAS78B OSRAM