SSL 502 EQ - HMF band needs repair

[barthman.de]

I have been using an SSL 502 EQ in the studio for years, which is built into a 19" housing. Recently, the HMF tape has been causing problems. In the “wide bell” mode, the Q value is rather narrow and the gain significantly higher than with the other bands. If I use a pink noise signal at + 4dbU the full cut with this band goes on an analyser till -80 dB. If I boost, the maximum value of the EQ of 15dB gain is also significantly exceeded. If I activate the "Hi-Q" mode, there are strong oscillations depending on the frequency. Gain and frequency seem to be mutually dependent. With a lower gain, the oscillations occur from a higher frequency, if I increase the gain, the frequency threshold drops.

I have already exchanged the IC 5534 (referred to in the circuit diagram IC 5/6) for ICs that I had lying around here. The problem persists. I can also not imagine that the built-in filter capacitors are defective. If I get this right, these are usually foil capacitors that don't wear out as quickly. Could one of the potentiometers be the cause? How can I proceed when testing the circuit? I am grateful for any advice. Testing and removing it from the rack are time-consuming, so it would be nice to be able to narrow down the problem beforehand.

Attached is the circuit diagram exclusively to support troubleshooting.

 

[gyraf]

I would primarily suspect the H5 FET-switches (whatever they are) that activates the high-Q mode. Perhaps leakage backwards between switches? Try replacing, if possible?

/Jakob E.

 

[abbey road d enfer]

I have been using an SSL 502 EQ in the studio for years, which is built into a 19" housing. Recently, the HMF tape has been causing problems. In the “wide bell” mode, the Q value is rather narrow and the gain significantly higher than with the other bands. If I use a pink noise signal at + 4dbU the full cut with this band goes on an analyser till -80 dB. If I boost, the maximum value of the EQ of 15dB gain is also significantly exceeded. If I activate the "Hi-Q" mode, there are strong oscillations depending on the frequency. Gain and frequency seem to be mutually dependent. With a lower gain, the oscillations occur from a higher frequency, if I increase the gain, the frequency threshold drops.

Hard to troubleshoot from a distance, but I suggest you check trimmer VR1, that may be misplaced.
After that I would suspect the circuit referenced as 85504; it's a multiple FET switch. I couldn't find any reference to that part, so I guess it could be a thick-film assembly, as was popular at the time.
You could check by just shorting pins 8/10 and 11/13. In low-Q mode, they should be shorted.
Of course, one cannot rule out the possibility of a faulty pot, but it should be easy to check with an ohmmeter.

 

[barthman.de]

Hard to troubleshoot from a distance, but I suggest you check trimmer VR1, that may be misplaced.
After that I would suspect the circuit referenced as 85504; it's a multiple FET switch. I couldn't find any reference to that part, so I guess it could be a thick-film assembly, as was popular at the time.
You could check by just shorting pins 8/10 and 11/13. In low-Q mode, they should be shorted.
Of course, one cannot rule out the possibility of a faulty pot, but it should be easy to check with an ohmmeter.

The Q value can only be adjusted within very narrow limits using the trimmer. That is not the cause of the error.
I've also thought about the 85504. These 85504 are installed on two boards for all bands (H5 / H9). These are the only modules that are soldered directly onto the board. All the others have sockets. If one is defective, I will have to look for a replacement part. It will definitely not be easy. I will swap H5 with H9 first and when I get the chance install sockets at the same time.

I thinkt I can't avoid disassabling the EQ. So I can also test the potentiometers.

 

[barthman.de]

 

[Bo Deadly]

You might try to clean all leads around the gates of those JFETs. I don't suspect that is the problem because the impedance is only 50K. So if a quick cleaning with 99% isopropyl and a lens wipe gripped with plastic tweezers doesn't fix it, that's not the problem. It could very well be that one or more JFET has become leaky like gyraf said in which case you need to remove that board.

With the right technique, I think you can remove that board pretty easily. Assuming you have a hot air station? Use a razor blade to cut out a piece of card stock with a narrow rectangular opening that is very close to the footprint of that daughter board. Then place it over the back with some metal tape, again leaving only a narrow gap for that boards pins. Then use a hot air station to heat the pins and pull that board out. If you're quick about it and lucky, you might leave the board pins intact but you can fix them up later if necessary.

With the JFETs board out, you might be able to test the unit to see if the desired behavior is restored. If confirmed, replace the JFETs and re-install. There are probably many JFETs that would work in that situation. Assuming they're N-channel, I would think something like J112 would be good but it looks like Vgs could, in theory, be as high as 24V+17.5V which might be too high.

 

[moamps]

I don’t think the problem is in FETs at all, but I may be wrong. Excessive attenuation or amplification of the signal indicates that the problem is in the VR4 potentiometer (or its connecting wires) in the gang that is in the feedback network (in series with R32).

 

[JohnRoberts]

Pot failures are generally mechanical. I've seen worn tracks, dirty wipers, and even cracked substrates.

If the audible artifacts track with pot rotation that makes it a suspect, but not the only one.

An always worthwhile suspect is a bad solder joint.

JR

 

[barthman.de]

I thank you for your opinions. I wanted a desoldering station for Christmas and then between Christmas and New Years I will dedicate myself to the EQ.

 

[barthman.de]

The first thing I did was to remove the frequency potentiometer (VR4) and measure the resistances of the three decks. When I measure between the red and black cable, the resistance between full CCW and full CW runs nicely from 0 to 22k. It looks different between the green and blue cable: the potentiometer starts full CCW at 7k, runs up to the middle of the potentiometer up to 22k, then there seems to be a short circuit, before it then goes up again from 0k to 1k8 to full CW. Can't imagine that it has to be like that.

The 100k Deck shows a similar problem. Here the potentiometer starts at 62k5, runs up to 100k and then starts again from 0 to 36k5. The function of the 100k deck is also not clear to me because I cannot find the deck in the circuit diagram. Is it only there to save the settings?

And now the next exciting question. If the potentiometer is actually defective, where can I get a replacement part?

 

[barthman.de]

I have just removed a potentiometer of another frequency band to double-check. This potentiometer works normally. The error of the HMF tape is definitely caused by the potentiometer. It seems as if the sliders are shifted within the potentiometer on two decks. Unfortunately the potentiometers are riveted, so that I'll need a replacement potentiometer.

 

[moamps]

I would take Dremel here without hesitation.
IMO, the best chance to find a replacement potentiometer is to make an inquiry on the SSL forum
http://forum.sslmixed.com/index.php

 

[Cyril Jones]

Read your posts with interest. I have a quad 22kohm potentiometer which you might find useful to change tracks or even remove 2 from the quad pot which is bolted through. I have other pots, not suitable but with 3 tracks also bolted through. so correct bolts there.

Another time it could be useful to short pins 11/13 and 8/10 on H5 c and d to determine a response change. the third track is of no consequence being only useful in a console.

I have mounted many mono and stereo modules in 1U racks taken from a complete console,.
Still have a stereo bar graph meter too.

Cyril UK

 

[john12ax7]

SSL will also sell you replacement pots, at least for parts used in current production.

 

[kags]

Lawrence Malchose (tech at East-West in L.A.) is the guy SSL refers service matters to as of late. He was very helpful to me in sourcing some parts for a 9000J since SSL has discontinued factory support and parts. This is valid as of November 2021

 

[barthman.de]

Today I looked at the defective potentiometer again. Unfortunately, I was unable to remove the rivets using a punch. However, I found out that you can pull out the potentiometer axis. If you do this gradually, you can align the individual levels again so that all levels start at 0 ohms. I will first reinstall the potentiometer and see whether it works permanently.

 

[barthman.de]

I had time to put everything together now. The EQ band works perfectly again and the potentiometer seems to hold if you don't pull the axis too hard. Works for me for now. In the long run I will look for a replacement pot.

 

[zonkers31]

Usually the 100K part is for total recall and not part of the audio path. A dual 22k would suffice if you are not using recall automation.