Microx,
You haven't mentioned this so before you get too far into the in-depth testing, have you fully cleaned all of the rotary switches? These old HP's are notorious for flaky connections through the rotary switches and most folks go through them and manually clean the switch decks right from the start. If you haven't performed a clean on the switches in this unit then get some cotton swabs and alcohol and get started. It is not recommended to use spray cleaner because many sprays will leave residue behind or blast garbage off the decks and into an area you do not want it.
The 339 I bought had multiple weird issues and most of those issues disappeared after I performed the switch cleaning. My 339 had some sort of electrical grease lightly applied to the switch wafers and it had turned brown over time. It took a LOT of cleaning to get it all out. I'm talking like 3 - 4 cotton swabs per wafer.
You haven't mentioned this so before you get too far into the in-depth testing, have you fully cleaned all of the rotary switches? These old HP's are notorious for flaky connections through the rotary switches and most folks go through them and manually clean the switch decks right from the start. If you haven't performed a clean on the switches in this unit then get some cotton swabs and alcohol and get started. It is not recommended to use spray cleaner because many sprays will leave residue behind or blast garbage off the decks and into an area you do not want it.
The 339 I bought had multiple weird issues and most of those issues disappeared after I performed the switch cleaning. My 339 had some sort of electrical grease lightly applied to the switch wafers and it had turned brown over time. It took a LOT of cleaning to get it all out. I'm talking like 3 - 4 cotton swabs per wafer.
OK, wave form at monitor output pretty clean, TP1 7.34v U3C pin 8 is7.8v. I have read of people repairing the opto devices, is there not a Vactrol type that would do
Chamberman, I bought this unit on ebay and the first thing I did was clean the wafers with a non-residue spray and they were dirty. I will clean them again anyway as you suggest just to be sure, thanks again.
scope on TP4 clean at 5.7v. U3 pin 6 11.5v ac, pin2 0v ac
Chamberman, I bought this unit on ebay and the first thing I did was clean the wafers with a non-residue spray and they were dirty. I will clean them again anyway as you suggest just to be sure, thanks again.
scope on TP4 clean at 5.7v. U3 pin 6 11.5v ac, pin2 0v ac
I think the opto-coupler is the likely culprit. I'd assumed saturation of TP1 closer to the 9V spec on CR5, but on closer scrutiny, it's driven by less than 1mA via R21. So 7.8V clamp sounds believable.
Perhaps other forum members know of suitable part numbers/vendors for these devices. I'd advise searching out a replacement for the E1 device as a precaution as well. I'm betting this not a new dilemma.
Perhaps other forum members know of suitable part numbers/vendors for these devices. I'd advise searching out a replacement for the E1 device as a precaution as well. I'm betting this not a new dilemma.
I haven't run across anyone that had been successful in finding a modern replacement for the 339A opto-coupler's. I know others have brought up this issue in the past. You might be able to find something close that with some resistance tweaks might get it working. Its more probable though that you'd have better luck replacing the LED in the existing coupler.
Good point KC.
I recall someone suggesting that now...replacing just the LED in the coupler. That information was covered in one of the threads, which I don't recall, but it was covered. I don't know if whomever was doing it was successful or not, perhaps I want to say. Without find the posts, that will have to wait for now.
Cheers,
I recall someone suggesting that now...replacing just the LED in the coupler. That information was covered in one of the threads, which I don't recall, but it was covered. I don't know if whomever was doing it was successful or not, perhaps I want to say. Without find the posts, that will have to wait for now.
Cheers,
Hi Microx.
I confess didn't pay close attention to your amplitude measurements of TP4 and U3. Now that I consider more carefully, I'm unsure of my pronouncment of photo-coupler death.
You report that voltage at U3 output is about twice the amplitude at TP4. That suggests opamp feedback resistance is twice the resistance of R48, or about 30K. But R49 is only 3.48K. That leads me to suspect R49 may be open. With power off (no photocell illumination), use an ohmmeter to measure R49 circuit resistance, perhaps probing near the photocell leads. I suggest this in order to avoid mechanical stress to the body of R49 on the chance it or a nearby connection might be intermittent. Resistance should be less than 3.48K since other paths are conducting in parallel.
I'm sorry I didn't catch this earlier.
I confess didn't pay close attention to your amplitude measurements of TP4 and U3. Now that I consider more carefully, I'm unsure of my pronouncment of photo-coupler death.
You report that voltage at U3 output is about twice the amplitude at TP4. That suggests opamp feedback resistance is twice the resistance of R48, or about 30K. But R49 is only 3.48K. That leads me to suspect R49 may be open. With power off (no photocell illumination), use an ohmmeter to measure R49 circuit resistance, perhaps probing near the photocell leads. I suggest this in order to avoid mechanical stress to the body of R49 on the chance it or a nearby connection might be intermittent. Resistance should be less than 3.48K since other paths are conducting in parallel.
I'm sorry I didn't catch this earlier.
checked that out and R49 is not open and is not 3.48K. It is 6.8K mounted on two stand off pins with a small ceramic cap across it, maybe a selected value in production. With it out of circuit the Cds cell reads 1.2m, about right I would say. Scoped the input to U3, normal sine wave but on pin6 it is modulated with garbage, a roughly sawtooth shape at about 2.5mHz. I guess it is oscillating.
Further checking shows the input is also dirty with the same garbage
Further checking shows the input is also dirty with the same garbage
This sounds very strange. Do you see evidence that this unit has been hacked on by a previous owner?
I suggest fixing the oscillation first. It will cause only grief if not dealt with. You mention it's present at the input. Is the oscillator clean? If oscillator is ok, I suggest shorting the analyzer input. It will be easier to hunt for oscillation without an input interfering with scope sensitivity and sync. See if you can home in on the source, where the amplitude is largest.
BTW, I assume power supplies are clean, very little noise when checked with AC coupling, high scope sensitivity?
I suggest fixing the oscillation first. It will cause only grief if not dealt with. You mention it's present at the input. Is the oscillator clean? If oscillator is ok, I suggest shorting the analyzer input. It will be easier to hunt for oscillation without an input interfering with scope sensitivity and sync. See if you can home in on the source, where the amplitude is largest.
BTW, I assume power supplies are clean, very little noise when checked with AC coupling, high scope sensitivity?
I retract my earlier comment about strange. I found mention of component changes in the "Manual Changes" section of the service manual. Among the revisions are change of R48 to 27.4K and R49 to 6.81K with 20pF in parallel. These changes were made "to eliminate ??5MHz oscillations in A3U3." ?? because those characters are obscured by a hole punch black spot in my PDF of the manual. Could be 2.5, but I can't tell. Perhaps these revisions were not entirely successful? 
Can you determine if determine if this oscillation is locally generated at U3? It's possible that they might appear to be present at the input but appearances might be false. You could try experimentally disconnecting from Buffer Amplifer source and grounding to see if oscillation persists. These phenomena can be very vexing.
Good luck.
Can you determine if determine if this oscillation is locally generated at U3? It's possible that they might appear to be present at the input but appearances might be false. You could try experimentally disconnecting from Buffer Amplifer source and grounding to see if oscillation persists. These phenomena can be very vexing.
Good luck.
Going back to board A4 error detector I have a clean sine wave on TP 3,4 and 5 what should I expect on TP1 and 2. On TP1 I have a jagged unstable waveform and on TP2 no waveform just DC. I don't think this unit has been hacked, work has been done but done well. When I short the input to distortion amp the garbage vanishes. I am still curious that the frequency leds NEVER light at all . Yes power supplies are clean, I have changed the four smoothing caps for 4,700 uf Panasonic 105 deg
Last edited:
Hi. Just back from errands and lunch. I'm glad we're not faced with a hacked init. I'll study schematics a few minutes and get back to you.
Not sure about frequency LEDs. If you drove analyzer from a separate generator so that you could apply a frequency other than analyzer setting, you might see arrow activity.
Not sure about frequency LEDs. If you drove analyzer from a separate generator so that you could apply a frequency other than analyzer setting, you might see arrow activity.
If everything were working properly, you would see DC voltages on TP1 and 2--- somewhere between 0 and the clamp zener clamps CR5 and CR11. I think typical voltages would be 2 or 3 volts--- instinctive guesses. In fact, you can surmise from the bias voltages on U6A and U6C that the allowed range for TP2 is 0.7 to 8.5V TP1 and 2 are "integrator" outputs. If you were to measure U3A or U5A outputs, you'd find they been driven to 0. Non-0 voltage would mean that control path was unable to balance.
Given oscillation on A3U3, I'd expect to see related signals on A4TP4. But you say TP4 was clean sine wave. 2.5MHz might get attenuated before it it arrives on TP4, but this discrepancy should be explored. Of course, TP4 would be the nulled distortion imperfections if the nulling circuit were working.
Oscillation still need to be cured and then nulling problems tackled, I think.
Given oscillation on A3U3, I'd expect to see related signals on A4TP4. But you say TP4 was clean sine wave. 2.5MHz might get attenuated before it it arrives on TP4, but this discrepancy should be explored. Of course, TP4 would be the nulled distortion imperfections if the nulling circuit were working.
Oscillation still need to be cured and then nulling problems tackled, I think.
OK, I will work on the oscillation. I have driven the unit with an external generator and the arrow leds do NOT even blink. Regarding the repair of the opto devices by replacing the led, the colour of the led would be critical. Cds cells are overly sensitive to red light so a red led would work in a very different way to a green or blue etc.
Thanks again for the help
Thanks again for the help
Good luck with the oscillation battle. I'll offer some thoughts--- not sure what they're worth.
As best I can tell, U3 and U4 are LM318. I do not understand the compensation method HP seems to favor with these opamps--- namely C22 and especially C27. I don't find these component connections suggested in any manufacturer's data sheet. The following link seems to offer useful advice: http://www.analog.com/media/en/technical-documentation/data-sheets/lt0118.pdf
I know that capacitance loading on an opamp output can lead to oscillation. I would try removing C22 and C27 and see if that helps. Same story with C25, C26 and maybe other sites as well.
Sometimes it's difficult to know for certain which device is generating an oscillation because the signal seems to appear everywhere. Often it's possible to disable nearby opamps by biasing them into saturation to rule them out.
If you get stuck, I'll try to help you brain storm.
Again, best of luck.
As best I can tell, U3 and U4 are LM318. I do not understand the compensation method HP seems to favor with these opamps--- namely C22 and especially C27. I don't find these component connections suggested in any manufacturer's data sheet. The following link seems to offer useful advice: http://www.analog.com/media/en/technical-documentation/data-sheets/lt0118.pdf
I know that capacitance loading on an opamp output can lead to oscillation. I would try removing C22 and C27 and see if that helps. Same story with C25, C26 and maybe other sites as well.
Sometimes it's difficult to know for certain which device is generating an oscillation because the signal seems to appear everywhere. Often it's possible to disable nearby opamps by biasing them into saturation to rule them out.
If you get stuck, I'll try to help you brain storm.
Again, best of luck.
That's goos news. I hope the good behavior continues.
I'll try to offer some guideposts about the notch circuit. Twin-T networks do not generate deep nulls. That's what underlies the comment about -16dB rejection in manual section 8-31. Consequently, to yield a null, the resistors in the "Feed Forward Signal" path in Figure 8-3 must have a ratio of -16dB. Your unit has R48 = 27.4K, so net feedback resistance at null will be about 4.34K. Since your unit has R49 = 6.81K, nominal photocell resistance at null will be about will be about 12.5K.
Further, the signal amplitude at U3-pin3 should be about 16db below TP4 amplitude. From your post #22 indicating pin 2 was 0, I'd infer pin 3 was also 0, not 16dB below TP4. Should probably recheck that experiment. Some of those results still don't make sense.
It's unusually cool today, so I need to take advantage to do some repair in my attic. But when I emerge, I'll check back.
Good luck.
I'll try to offer some guideposts about the notch circuit. Twin-T networks do not generate deep nulls. That's what underlies the comment about -16dB rejection in manual section 8-31. Consequently, to yield a null, the resistors in the "Feed Forward Signal" path in Figure 8-3 must have a ratio of -16dB. Your unit has R48 = 27.4K, so net feedback resistance at null will be about 4.34K. Since your unit has R49 = 6.81K, nominal photocell resistance at null will be about will be about 12.5K.
Further, the signal amplitude at U3-pin3 should be about 16db below TP4 amplitude. From your post #22 indicating pin 2 was 0, I'd infer pin 3 was also 0, not 16dB below TP4. Should probably recheck that experiment. Some of those results still don't make sense.
It's unusually cool today, so I need to take advantage to do some repair in my attic. But when I emerge, I'll check back.
Good luck.
Wow! That's a great clue!
That's down about down only about 4.4 dB re TP4. Sounds like there'e a problem in the Twin-T network. At this point it may be helpful to use an external generator so that you can explore the frequency response of the T-T with analyzer setting held constant. Poke around with scope. A decade below and above analyzer test frequency there should be little attenuation (see Figure 8-4). You might do same test in various frequency ranges (x10, x 100, etc.). Behavior might vary range by range.
Try tracing signal from TP3 along path test through S1D Front and S1C Front. Switch continuity might be a problem. You can probe on either sides of switch and look for discrepancies.
I think you may be closer.
That's down about down only about 4.4 dB re TP4. Sounds like there'e a problem in the Twin-T network. At this point it may be helpful to use an external generator so that you can explore the frequency response of the T-T with analyzer setting held constant. Poke around with scope. A decade below and above analyzer test frequency there should be little attenuation (see Figure 8-4). You might do same test in various frequency ranges (x10, x 100, etc.). Behavior might vary range by range.
Try tracing signal from TP3 along path test through S1D Front and S1C Front. Switch continuity might be a problem. You can probe on either sides of switch and look for discrepancies.
I think you may be closer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.