Hi everyone,
I THINK this is the appropriate area, mods feel free to move this thread.
Anyway,
I have an old old guitar effects processor that I've recently pulled out of storage and started playing with again.
One of the things that always bugged me about it is that as the input signal fades and the signal indicator LED switches off, there is a noticeable and objectionable fuzz/distortion sound for a moment.
It's just at the threshold of the LED flickering and going dark.
So I managed to get my hands on the schematic and pulled the unit apart to have a look.
I've posted the relevant (small) part of the schematic to get some opinions on what the cause is.
From what I can tell from the schematic, when the input signal goes above around 2.4 volts, the signal LED will light.
What I can't work out is how is it feeding a signal back through the input!
If I pull the op-amp, no noise, but no LED of course. I've tried chucking a cap across various points to see if if makes a difference, nope. Tried a couple of different op-amps, nope.
I'm not good enough at reading schematics to come up with much else.
I've also attached a shot of the scope at and just below the point of the noise. You can quite clearly see the distortion.
HELP! The noise drives me nuts!
Any suggestions or possible solutions will be gladly accepted.
Cheers
Mark
I THINK this is the appropriate area, mods feel free to move this thread.
Anyway,
I have an old old guitar effects processor that I've recently pulled out of storage and started playing with again.
One of the things that always bugged me about it is that as the input signal fades and the signal indicator LED switches off, there is a noticeable and objectionable fuzz/distortion sound for a moment.
It's just at the threshold of the LED flickering and going dark.
So I managed to get my hands on the schematic and pulled the unit apart to have a look.
I've posted the relevant (small) part of the schematic to get some opinions on what the cause is.
From what I can tell from the schematic, when the input signal goes above around 2.4 volts, the signal LED will light.
What I can't work out is how is it feeding a signal back through the input!
If I pull the op-amp, no noise, but no LED of course. I've tried chucking a cap across various points to see if if makes a difference, nope. Tried a couple of different op-amps, nope.
I'm not good enough at reading schematics to come up with much else.
I've also attached a shot of the scope at and just below the point of the noise. You can quite clearly see the distortion.
HELP! The noise drives me nuts!
Any suggestions or possible solutions will be gladly accepted.
Cheers
Mark
Attachments
Only quick thing I can think of is that the opamp bursts into oscillation as the signal crosses the threshold at which the opamp changes state.
Try a small cap across pins 6 and 7 (something like 100pf at a quick guess) and/or a high value resistor from pin 7 to pin 5 (1 to 10meg range) to add a little hysteresis so it switches cleanly.
Try a small cap across pins 6 and 7 (something like 100pf at a quick guess) and/or a high value resistor from pin 7 to pin 5 (1 to 10meg range) to add a little hysteresis so it switches cleanly.
Could the glitch be being transmitted via the power supply rails? Are the audio bits running on the same supply rails as this LF353? If so then decoupling the supply to this chip with 22ohm series resistors (in +/-12V) and caps to gnd from pins 4&8 might fix it up.
Another possibility is the glitches are getting on to the gnd - there's no reason for the LED to be connected to gnd, it could go to the -12V rail and double the 4k7 feeding it to 10k. Be sure to move C41 too as that was giving you problems even without the LED in circuit.
Another possibility is the glitches are getting on to the gnd - there's no reason for the LED to be connected to gnd, it could go to the -12V rail and double the 4k7 feeding it to 10k. Be sure to move C41 too as that was giving you problems even without the LED in circuit.
The fundamental problem is that the op-amp used as a comparator amplifies the signal with its full open-loop gain. When the signal is hovering around the threshold, it outputs a full amplitude square wave. This easily gets into everything through stray capacitance and through the power rails.
There is no simple cure, the easiest way (IMO) is to disable the circuit altogether by removing the opamp. Some of the suggestions given above might well help though.
There is no simple cure, the easiest way (IMO) is to disable the circuit altogether by removing the opamp. Some of the suggestions given above might well help though.
Scopeboy,
Interestingly, the other half of the dual op-amp (driving the clip LED) doesn't seem to have any issues noise/distortion wise.
Maybe because it is buffered with the transistor?
I guess I could always duplicate the clip LED output circuit for the signal. I'm hoping not to have to do that
Mark
Interestingly, the other half of the dual op-amp (driving the clip LED) doesn't seem to have any issues noise/distortion wise.
Maybe because it is buffered with the transistor?
I guess I could always duplicate the clip LED output circuit for the signal. I'm hoping not to have to do that
Mark
Hehe, yeah, it's not exactly hifi...
That said, even with a clean setting you can't really hear it clip.
So I tried out a couple of things quickly tonight.
100pf across pin 6 & 7. No change
1M (highest I have here) across pin 5 & 7. No change
Both together, No change. Did have a little whistle though
100nf across 4 & 8. No change
Had a look at the power rails. There was a tiny wiggle when the distortion hit. I could just see it among the noise floor of my scope (which is admittedly a bit crap) and the hash on the rails. Not sure this is causing it, as it is a couple of orders of magnitude smaller than the spike I see on the input waveform.
Oh by the way, in the pics I posted, I had the oscilloscope input inverted by mistake. So the little spike would be on the positive, which makes more sense.
I also had a look at the output of the op-amp.
At the threshold it was giving some nasty triangle waves as it bounced between the 12V rails. I'm guessing triangle because C41 is charging and discharging yes?
Anyway, enough for tonight.
I'll try to get a pic of the op-amp output tomorrow night for your viewing pleasure.
May be a couple of days before I can perform deeper surgery.
At the end of the day, my final solution is just to lift the input (pin 5). I'll still have the clip indicator. It just bugs me that I have a signal light that works, but makes it sound worse. Did they not listen to this thing before they released it? Haha!
Thanks for the help and suggestions guys, still have a few more to try so I'll let you know how I get on.
Cheers,
Mark
That said, even with a clean setting you can't really hear it clip.
So I tried out a couple of things quickly tonight.
100pf across pin 6 & 7. No change
1M (highest I have here) across pin 5 & 7. No change
Both together, No change. Did have a little whistle though
100nf across 4 & 8. No change
Had a look at the power rails. There was a tiny wiggle when the distortion hit. I could just see it among the noise floor of my scope (which is admittedly a bit crap) and the hash on the rails. Not sure this is causing it, as it is a couple of orders of magnitude smaller than the spike I see on the input waveform.
Oh by the way, in the pics I posted, I had the oscilloscope input inverted by mistake. So the little spike would be on the positive, which makes more sense.
I also had a look at the output of the op-amp.
At the threshold it was giving some nasty triangle waves as it bounced between the 12V rails. I'm guessing triangle because C41 is charging and discharging yes?
Anyway, enough for tonight.
I'll try to get a pic of the op-amp output tomorrow night for your viewing pleasure.
May be a couple of days before I can perform deeper surgery.
At the end of the day, my final solution is just to lift the input (pin 5). I'll still have the clip indicator. It just bugs me that I have a signal light that works, but makes it sound worse. Did they not listen to this thing before they released it? Haha!
Thanks for the help and suggestions guys, still have a few more to try so I'll let you know how I get on.
Cheers,
Mark
I'll have a think too. Although I still suspect some oscillation problem it shouldn't really be affecting the audio so much unless there is a wiring issue somewhere or the rails are being wildly modulated by whatever is happening.
As you have a scope you could try rigging a preset up to give a variable voltage to the clip input and set the preset so that the LED is just coming into conduction and see if that produces interference. At least then you could make it do it permanently and then see where its contaminating the audio etc.
As you have a scope
you could try rigging a preset up to give a variable voltage to the clip input and set the preset so that the LED is just coming into conduction and see if that produces interference. At least then you could make it do it permanently and then see where its contaminating the audio etc.
I think the "feedback" (1M res + 100pf) is not working due to the 0 ohm resistor, inverting input is at ground so you can't add negative feedback.
Things i would try ?
-desolder op-amp and put a 8 DIL socket ... easier to swap different ICs. Maybe try LM358 or a comparator as suggested.
Also i would try increasing R32 from actual 10k to 100k and see if the problem is reduced...just a try
Another test, remove the jumper and put a resistor at R36, i will try 1K and re-try the feedback trick.
If the 1K resistor increase the Clip and Signal indicator threshold too much, you can try lower the 1K value...
Things i would try ?
-desolder op-amp and put a 8 DIL socket ... easier to swap different ICs. Maybe try LM358 or a comparator as suggested.
Also i would try increasing R32 from actual 10k to 100k and see if the problem is reduced...just a try
Another test, remove the jumper and put a resistor at R36, i will try 1K and re-try the feedback trick.
If the 1K resistor increase the Clip and Signal indicator threshold too much, you can try lower the 1K value...
I'd start with decoupling the power supply at the op-amps. Then I'd add some hysteresis to the comparator to turn it into a schmitt trigger. That will absolutely prevent it from oscillating when the signal level is hovering around (or crossing) the threshold. I'd start with a 1 MΩ from pin 7 to 5 (output to +input) and work up or down from there if there is too little or too much hysteresis.
I don't know about unstable oscillation,
The op-amp is already in a socket which makes life easier.
I tried NE5532, LM358, LM833 (because that's what I had lying around)
and I couldn't tell any difference. I would have thought that they might at least distort DIFFERENTLY if it was instability.
I'm at work atm, hopefully I'll have a chance to try some more stuff tonight re: feedback/decoupling.
The unit is and OLD -as in '95 vintage- Digitech GSP-5. Cheap and nasty, as I was a poor student at the time
Cheers,
Mark
The op-amp is already in a socket which makes life easier.
I tried NE5532, LM358, LM833 (because that's what I had lying around)
and I couldn't tell any difference. I would have thought that they might at least distort DIFFERENTLY if it was instability.
I'm at work atm, hopefully I'll have a chance to try some more stuff tonight re: feedback/decoupling.
The unit is and OLD -as in '95 vintage- Digitech GSP-5. Cheap and nasty, as I was a poor student at the time
Cheers,
Mark
RIGHT!
Ok, got some more time to play round with it.
Abraxalito, were sure were on to something when you said C41 was giving me problems, or thereabouts.
So I'm having a close look at the board to work out exactly what R and C was what on the schematic and doing a bit of head scratching as some of the caps aren't labelled, and pretty much all of the resistors have the label directly under them. Handy right?
So I probe around with continuity on the old dmm and something isn't making sense.
Turns out, C41 is added in on top of other component's pads. The kids in the factory back in the 90's have put one leg in the wrong place!
oops.
So C41 is directly on the op-amp output, before R37, see first pic. Not AFTER R37 like the original schematic, see pic 2.
If I disconnect or connect C41 as it was meant to be, no more noise.
I haven't chucked the scope on it, because really, if I can't hear it, I don't care. (at least not when it comes to lo-fi guitar effects)
Thanks for all your suggestions, it gave me some motivation that the stupid thing could be fixed.
I'm just glad I didn't have to go down the route of hacking extra circuitry into it!
Thanks again,
Mark
Ok, got some more time to play round with it.
Abraxalito, were sure were on to something when you said C41 was giving me problems, or thereabouts.
So I'm having a close look at the board to work out exactly what R and C was what on the schematic and doing a bit of head scratching as some of the caps aren't labelled, and pretty much all of the resistors have the label directly under them. Handy right?
So I probe around with continuity on the old dmm and something isn't making sense.
Turns out, C41 is added in on top of other component's pads. The kids in the factory back in the 90's have put one leg in the wrong place!
oops.
So C41 is directly on the op-amp output, before R37, see first pic. Not AFTER R37 like the original schematic, see pic 2.
If I disconnect or connect C41 as it was meant to be, no more noise.
I haven't chucked the scope on it, because really, if I can't hear it, I don't care. (at least not when it comes to lo-fi guitar effects)
Thanks for all your suggestions, it gave me some motivation that the stupid thing could be fixed.
I'm just glad I didn't have to go down the route of hacking extra circuitry into it!
Thanks again,
Mark
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.