Author Topic: HP 33120A Problem (Read 3777 times)

Henry Finley · « **on:** February 21, 2022, 02:14:24 am »

I have here my buddy's HP 33120A function generator that he has so graciously loaned me. I would like to return it to him with its ongoing problem repaired. The thing works now for the purpose, but has an error problem. When you first turn it on cold, all seems well. After 10 minutes or so, the display blacks out, requiring a re-boot. When it reboots, the screen says RS 232 for a moment, then proceeds on to normal usage, with ERROR below the numerals. It doesn't seem to give any real problems in usage, but it will not connect to a computer (not that he intends to use one). I'd like to surprise him with his machine returned in correct condition. Any ideas? Thank you.

nctnico · « **Reply #1 on:** February 21, 2022, 02:21:32 am »

My idea: don't mess with it without asking first. I would certainly not like people messing with my equipment without me knowing. Even with the best intentions. There is probably a good reason your friend hasn't attempted to repair it. What if you make matters worse?

Henry Finley · « **Reply #2 on:** February 21, 2022, 02:28:07 am »

He hasn't attempted to repair it because he doesn't really care. But it's his, he trusts me, and knows that if I did screw it up I would pay him for it. That's not the point. The machine is also missing its handle. He told me if I could find one to let him know. There are no handles to be found. He has a lot of equipment that I can't afford to collect. It I return the machine working correctly OR his money, he will entrust me with future borrowing all-the-more. When I borrow, it is quickly returned, and in cleaner shape than when it went out the door. He knows that. It's called gratitude and good will. that's the way I do things.

bob91343 · « **Reply #3 on:** February 21, 2022, 03:01:52 am »

How about this:

https://www.ebay.com/itm/121629807571

Kean · « **Reply #4 on:** February 21, 2022, 08:00:57 am »

Are you able to trigger a self test immediately after the problem occurs. You can run a full self-test by holding the SHIFT key while powering on.
This may give an error code to check against the service manual. You might want to try it even before you reproduce the issue to see if it reports anything.

Edit to add:
Release the SHIFT key after it beeps during startup. You can also run the full or individual tests via the SYS menu, and also review any stored errors.
And the display of "RS-232" at startup is normal assuming that is the selected interface. I think for GPIB it will display the address instead.

nctnico · « **Reply #5 on:** February 21, 2022, 11:14:49 am »

@Henry: if you want to take a stab at a repair: I'd start by taking it apart and reseating (pull out / push in) all connectors and look for bad solder joints. Likely you can find a service manual with power supply voltages; it is good custom to check all supply voltage to determine these are within specs. Also look for fuses and check if these are OK.

Kean · « **Reply #6 on:** February 21, 2022, 11:41:45 am »

The 33120A service guide is available on Keysight website (and readily available elsewhere via a google search)
https://www.keysight.com/au/en/product/33120A/function--arbitrary-waveform-generator-15-mhz.html#resources

Also see the attached FAQ that I saved from the old Agilent site around 2013.

Henry Finley · « **Reply #7 on:** February 23, 2022, 12:34:08 am »

Quote from: nctnico on February 21, 2022, 11:14:49 am

@Henry: if you want to take a stab at a repair: I'd start by taking it apart and reseating (pull out / push in) all connectors and look for bad solder joints. Likely you can find a service manual with power supply voltages; it is good custom to check all supply voltage to determine these are within specs. Also look for fuses and check if these are OK.

OK Ive done this test. Sorry for the delay. Work schedule. There are 3 photos here, and 2 more that can't be shot because the screen changes too quickly to get a shot. 1st) hold shift and boot machine. You can see me here holding shift while I shoot the photo with the other hand. As soon as I release shift, it begins its test, whereupon it says PASS and proceeds to normal operation (no photo, too quick to shoot) Photo 2 is normal operation. After 5 or so minutes the screen goes blank, so I cut power. After 30 seconds I repeat the boot+test procedure, but it will not test. It simply says RS-232, along with the little error you see in the third photo. Too quick to photgraph. In the 3rd photo, you see the machine working (apparently correctly), but with the little "error" remaining.

bson · « **Reply #8 on:** February 23, 2022, 02:02:01 am »

It could be the RS-232 receiver chip, like MAX232 or whatever they might use. It might think there's something plugged plugged in on the serial port, read noise, and interpret it as a remote command error. You can view error details somewhere in the system menu, if memory serves.

Kean · « **Reply #9 on:** February 23, 2022, 08:09:33 am »

Quote from: Henry Finley on February 23, 2022, 12:34:08 am

After 5 or so minutes the screen goes blank, so I cut power. After 30 seconds I repeat the boot+test procedure, but it will not test. It simply says RS-232, along with the little error you see in the third photo. Too quick to photgraph. In the 3rd photo, you see the machine working (apparently correctly), but with the little "error" remaining.

OK, when it shows the "error" indicator on the display, I believe you can go into the system menu and review the error codes.
Press SHIFT then ENTER (MENU ON/OFF) to enter the menus.
Press right arrow three times to get to "SYS MENU", then down arrow once to enter that menu.
Press right arrow twice to get to "ERROR", and then down arrow to view the error list.
I think you can use left/right arrows to scroll through the errors if there are multiple.
Press SHIFT then ENTER to exit the menu system, or you can use up arrow to go back up to higher level menu, and continue looking at other options.

I'm pretty certain the "RS-232" message is not relevant to your problem, as that is a normal message at startup if RS-232 is enabled. Not saying there isn't a fault with the RS-232 port, or the related settings.

Kean · « **Reply #10 on:** February 23, 2022, 08:13:24 am »

Quote from: bson on February 23, 2022, 02:02:01 am

It could be the RS-232 receiver chip, like MAX232 or whatever they might use. It might think there's something plugged plugged in on the serial port, read noise, and interpret it as a remote command error. You can view error details somewhere in the system menu, if memory serves.

If this was the case it would normally turn on the "RMT" indicator on the display to show it is under remote control, or show an immediate "ERROR" indicator to show an invalid command. The screen going blank after some minutes of usage after power on will more likely be a thermal related component fault.

Henry Finley · « **Reply #11 on:** February 23, 2022, 11:14:50 pm »

Wow, this thing has 6 errors. I suppose the next thing is to download the book and look them up. although i highly doubt I can do much about them. I'm good on vintage gear, but this IC stuff is another animal. the errors are: 625, 512, 511, 513, 501, -101. thanks fellows. This thing works, or it seems to, with these errors, as far as generating signals for radio alignment and such. It checks out against my frequency counter. I promised Jack I'd stop before I felt like i might foul it up. But so far I have permission to keep trying. If I tear it up i suppose i bought me a new function generator. But that's the breaks when you are a crackpot who can't tolerate when a piece of gear has a problem.

Henry Finley · « **Reply #12 on:** February 24, 2022, 12:21:23 am »

an additional post about something curious. When booting the thing with its errors, I notice when I hit shift, enter to go back to normal operation, It goes back with no little ERROR light. So I go back into the menu, and it says "no errors" And the machine runs fine with no ERROR light. But after turning it off and restarting a few minuted later, it goes back to the original problem. And this time it has 8 errors, not 6 like last time. Crazy

J-R · « **Reply #13 on:** February 24, 2022, 12:42:30 am »

I won't claim to be a repair guru like some others, but there are some basic things you can do easily.

Pull the cover off, visually check for obvious things like bulging caps or blown components, check the voltages listed in the service manual on page 109.

After that, maybe set it to self test in a loop and then get a plastic probe and push on various components on the board to check for loose solder connections.

Kean · « **Reply #14 on:** February 24, 2022, 02:05:24 am »

Quote from: Henry Finley on February 23, 2022, 11:14:50 pm

Wow, this thing has 6 errors. I suppose the next thing is to download the book and look them up. although i highly doubt I can do much about them. I'm good on vintage gear, but this IC stuff is another animal. the errors are: 625, 512, 511, 513, 501, -101. thanks fellows. This thing works, or it seems to, with these errors, as far as generating signals for radio alignment and such. It checks out against my frequency counter. I promised Jack I'd stop before I felt like i might foul it up. But so far I have permission to keep trying. If I tear it up i suppose i bought me a new function generator. But that's the breaks when you are a crackpot who can't tolerate when a piece of gear has a problem.

OK, the service guide is available at https://www.keysight.com/au/en/assets/9018-40999/service-manuals/9018-40999.pdf

625: I/O Processor not responding
501: Isolator UART framing error
511: RS-232 framing error
512: RS-232 overrun error
513: RS-232 parity error

These could be the cause or just a symptom of the fault. The fact the display goes off leads me to think there is a bigger problem and this might just be a symptom, but you did mention that comms didn't work at all so there could be a general issue around the I/O processor section and/or related optical interface.

Like a lot of intermittent issues, this could be tricky to diagnose. I can't suggest anything simple other than the visual check and probing as mentioned by J-R and others.

Kean · « **Reply #15 on:** February 24, 2022, 02:06:51 am »

Quote from: Henry Finley on February 24, 2022, 12:21:23 am

an additional post about something curious. When booting the thing with its errors, I notice when I hit shift, enter to go back to normal operation, It goes back with no little ERROR light. So I go back into the menu, and it says "no errors" And the machine runs fine with no ERROR light. But after turning it off and restarting a few minuted later, it goes back to the original problem. And this time it has 8 errors, not 6 like last time. Crazy

I think viewing the errors just clears the list until the next self test is run (manually or at start up).

Henry Finley · « **Reply #16 on:** February 24, 2022, 02:33:53 am »

Quote from: Kean on February 24, 2022, 02:06:51 am

Quote from: Henry Finley on February 24, 2022, 12:21:23 am
an additional post about something curious. When booting the thing with its errors, I notice when I hit shift, enter to go back to normal operation, It goes back with no little ERROR light. So I go back into the menu, and it says "no errors" And the machine runs fine with no ERROR light. But after turning it off and restarting a few minuted later, it goes back to the original problem. And this time it has 8 errors, not 6 like last time. Crazy

I think viewing the errors just clears the list until the next self test is run (manually or at start up).

I think you're right. this thing acts like I'd expect from something that's been lightning struck. A dadgum Apollo 12 machine. Too bad there's no SCE to aux on this dadgum machine.

Henry Finley · « **Reply #17 on:** February 24, 2022, 02:52:39 am »

Here's what I think. I think this machine has something getting hot in it. We have a boy at work that has one of those gun things you can point at something and get a temperature. I believe I'll ask to bring it home and start pointing it around on this thing. I think it's just throwing up error codes at random with no rhyme or reason. Until that shorted part is found it will keep on working until it's had enough and gives up the ghost. Makes sense. Power supply heats up, loses voltage, and starts throwing up ridiculous errors where there are none.

tautech · « **Reply #18 on:** February 24, 2022, 03:30:38 am »

Quote from: Henry Finley on February 24, 2022, 02:52:39 am

Here's what I think. I think this machine has something getting hot in it. We have a boy at work that has one of those gun things you can point at something and get a temperature. I believe I'll ask to bring it home and start pointing it around on this thing. I think it's just throwing up error codes at random with no rhyme or reason. Until that shorted part is found it will keep on working until it's had enough and gives up the ghost. Makes sense. Power supply heats up, loses voltage, and starts throwing up ridiculous errors where there are none.

Reply #5 tells you to check supply voltages as a first step. < Best advice !
Don't go poking around with an infrared temp gun until you know the most critical part of any equipment; the PSU, is operating correctly.

Henry Finley · « **Reply #19 on:** February 25, 2022, 01:54:58 am »

Quote from: tautech on February 24, 2022, 03:30:38 am

Quote from: Henry Finley on February 24, 2022, 02:52:39 am
Here's what I think. I think this machine has something getting hot in it. We have a boy at work that has one of those gun things you can point at something and get a temperature. I believe I'll ask to bring it home and start pointing it around on this thing. I think it's just throwing up error codes at random with no rhyme or reason. Until that shorted part is found it will keep on working until it's had enough and gives up the ghost. Makes sense. Power supply heats up, loses voltage, and starts throwing up ridiculous errors where there are none.
Reply #5 tells you to check supply voltages as a first step. < Best advice !
Don't go poking around with an infrared temp gun until you know the most critical part of any equipment; the PSU, is operating correctly.

Thanks folks. During the workweek I get very little time to make tests and chronicle the work. What I've done tonight is to hook up my puny volt meter to the chassis and touch 4 marked voltage points. I did not consult the manual to be sure of where to hook the negative lead. I arbtrarily chose the chassis, due to time, just to take at quick stab at it. The voltages I read were 1/2 of the voltages marked on the circuit board., both before and after the above mentioned re-boot of the machine. I further noted that the heat sink of component C got quite hot within minutes, followed by D heating up more slowly, and not getting as hot, but still warmer than what I think is normal. A and B did not exhibit unreasonable heat. I believe these are SCR's, although I'm not sure. Inspection shows 3 leads apiece, but i still speculate they are not transistors. I could be wrong. Over the weekend I intend to get more in-depth, checking the manual, and checking the transformer connector for voltages. This is where I stand now. Regards.

Kean · « **Reply #20 on:** February 25, 2022, 04:29:31 am »

Components labelled A-D are linear voltage regulators
A = HS1003/U1003 LM337T for -5.2V (VEE) supply
B = HS1002/U1002 LM317T for +18V supply
C = HS1001/U1001 LM337T for -18V supply
D = HS1004/U1004 LM2940CT-5.0 for +5.0V (VCC) supply

You can find these in the above linked service guide on page 140 (schematic) and page 142 (assembly drawing). These linear regulators are all supplied by regulated and filtered raw voltages coming from transformer taps (also shown on page 140, top right).

If the voltages are very low, then I believe your problem is that the voltage selection switch may be set incorrectly.
If that checks out OK, then see if you can measure the raw rectified voltages on the inputs to those regulators (labelled RAW1IN thru RAW4IN on the schematic).

Kean · « **Reply #21 on:** February 25, 2022, 04:50:07 am »

Forgot to mention, measuring from chassis could be giving you bad readings.

Only one of the power rails is earth referenced, all the others are floating which includes all the ones you are measuring (as mentioned on page 98).
Try to locate PGND and use that for the floating supply voltage measurements. The metal tab of U1004 (on heatsink HS1004) may be the easiest point to use for PGND.

There is a separate +5V0 power rail that is earth referenced and regulated by U1051 (another LM2940T-5.0) and supplied from a separate transformer tap. U1051 is over the other side of the PCB hidden under the mains fuses and voltage selection part of the mains input.

All the floating circuit grounds (GND, PGND, AGND, A2GND) appear to be connected at a star point JM1006, there is a 1M resistor and 43V bi-direction TVS between circuit grounds and chassis earth (EGND).

nfmax · « **Reply #22 on:** February 25, 2022, 09:48:35 pm »

A thought. The error codes all have a common factor of the ground referenced I/O processor, and the failure after 5 or 10 minutes operation suggests something is getting too hot. The fan is powered from the unregulated ground referenced supply. If this supply is low, or has excess ripple, that might cause both problems. Is the fan running? What does the fan supply measure?

Henry Finley · « **Reply #23 on:** February 25, 2022, 10:44:27 pm »

Quote from: nfmax on February 25, 2022, 09:48:35 pm

A thought. The error codes all have a common factor of the ground referenced I/O processor, and the failure after 5 or 10 minutes operation suggests something is getting too hot. The fan is powered from the unregulated ground referenced supply. If this supply is low, or has excess ripple, that might cause both problems. Is the fan running? What does the fan supply measure?

I will have to hunker down this weekend and do my measuring so I can report back on these questions. Yes, the fan runs. and I agree the I/O is problematic, insofar as I think that is the circuitry that can link to a computer. And thirdly there most certainly is a heat problem. One of the heat-sinked voltage regulators gets HOT within a few minutes and the one next to it is not far behind. Wouldn't it be nice to just ignorantly buy new ones and put them in, and the problem solved? But we all know things can't be that easy. If something gets hot, the problem is likely elsewhere. I might mention after the reboot, I can hear the click of a solenoid , which I don't hear at first when it boots up from cold and works correctly. I don't think that is helpful though. And lastly, the machine is set to the correct line voltage. I tink this thing has received a zap or lightning struck. Just my gut feeling. Until I get busy on these meaurements, I'm just wasting the forum's time. I would not want to put you folks through that for naught.

nctnico · « **Reply #24 on:** February 26, 2022, 12:45:46 am »

No. Don't go swapping components at random. If you measure the right voltages around the regulators, then these are fine.

Henry Finley · « **Reply #25 on:** February 26, 2022, 02:54:17 am »

Thank you to the board. Please bear with me for the weekend till I can get these figures done. If I falter, it may take an additional weekend. During the workweek i have to put this pursuit out of my mind. This project is not for me. It is a labor of radio fellowship, and intent to make a searchable thread for others with this machine who need help. This particular generator belongs to K4LRH. I expect nothing in return. Just because we do something for a particular person, doesn't mean THAT EXACT person will repay you, nor do we expect or require it. But what goes around, comes around. From where, we can't know. That's radio. It's funny. Someday I may glom on to some jewel of vintage gear, and think to myself, "must be some seed I sewed somewhere". Radio is the best fellowship there is. Regards, KN4SMF

Henry Finley · « **Reply #26 on:** March 05, 2022, 04:54:29 am »

Quote from: Henry Finley on February 25, 2022, 10:44:27 pm

Quote from: nfmax on February 25, 2022, 09:48:35 pm
A thought. The error codes all have a common factor of the ground referenced I/O processor, and the failure after 5 or 10 minutes operation suggests something is getting too hot. The fan is powered from the unregulated ground referenced supply. If this supply is low, or has excess ripple, that might cause both problems. Is the fan running? What does the fan supply measure?
I will have to hunker down this weekend and do my measuring so I can report back on these questions. Yes, the fan runs. and I agree the I/O is problematic, insofar as I think that is the circuitry that can link to a computer. And thirdly there most certainly is a heat problem. One of the heat-sinked voltage regulators gets HOT within a few minutes and the one next to it is not far behind. Wouldn't it be nice to just ignorantly buy new ones and put them in, and the problem solved? But we all know things can't be that easy. If something gets hot, the problem is likely elsewhere. I might mention after the reboot, I can hear the click of a solenoid , which I don't hear at first when it boots up from cold and works correctly. I don't think that is helpful though. And lastly, the machine is set to the correct line voltage. I tink this thing has received a zap or lightning struck. Just my gut feeling. Until I get busy on these meaurements, I'm just wasting the forum's time. I would not want to put you folks through that for naught.

It is now 2 weeks later and i have reached my conclusions, with due credit to you fellows for having fed my original conclusions. I follewd it as best i could. In the end i have my own years of instincts to heed also. The machine works perfectly, but throws an error and continues working perfectly. Filter caps. They leak some ripple, and throw crazy nonsensical error codes. All of which mean trying to hook it to a computer will no longer work. So what? Who uses windows 95 anymore anyway? The machine runs for hours perfectly and does a stellar job on what it's supposed to do. I say the same thing now as when I tried to diagnose it a year ago and ended up with the came conclusion, before you fellows were available. In the end i have made my final decision: filter caps. It's not my machine. I can't risk damaging it. I'd raaher return it in defeat and shame, working with a stupid pointless error, than risk ruining it completely. This is not a machine for repair procedures on the kitchen table with cheap soldering irons and shadetree mechanical tecnique This is a machine that can be fouled up permanently up in a hearbeat. Thank you, KN4SMF


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: HP 33120A Problem (Read 3777 times)

Share me