Aion Electronics Lab Series L5 Build

[jstone]

Turn the compressor switch off, so it will remove it from the circus and will isolate the distortion circuit

Thanks, but I've tried it with the compressor switch both off and on and it makes no difference either way. The range I'm reading on pin 6 of IC10 is ~58mV with the trimpot all the way counterclockwise and ~16mV with the trimpot full clockwise. Changing the trim value does have an audible effect on the Master Volume level, so I know it's doing something.

[bsas]

I noticed something interesting: If I apply 1kHz 50mV sine wave on the circuit input, with my current configurations, I can measure 450mV across C38 if the compressor switch is OFF. If I turn the compressor switch ON and set the LIMIT where it can kick in constantly, the voltage across C38 drops to 17mV!!!

[jstone]

Bumping this up. I tried swapping the two CA3080s on the board and got the same results, so it looks like my problem is not with the ICs themselves. Does anyone how has also built this project have any ideas about what might be happening? Much appreciated in advance!

[bsas]

Bumping this up also because I already lost my hope to find the compressor issue by myself
I will basically forget about the pedal waiting for a good soul to give me any tip about what should I debug...

[digi2t]

Bumping this up. I tried swapping the two CA3080s on the board and got the same results, so it looks like my problem is not with the ICs themselves. Does anyone how has also built this project have any ideas about what might be happening? Much appreciated in advance!

I apologize if this has been mentioned already in the thread, but where did you source your 3080's (all your parts for that matter) from? If the y came from a dealer of ill repute, and are both fakes, then it wouldn't really matter which chip you put in which slot, now would it?

Also, go over all your resistors again. Ensure that you don't have a 220R in place where a 220K should be, or something like that. We had a forumite, paulyy, here a while back. He was building the Colorsound Vocalizer, and after several days of wits end drama, he still couldn't get it to work. I finally offered to look at it, so he sent it to me. First thing I did was go over all the components with a fine toothed comb. I found one resistor, if memory serves me, a 22K where a 2.2K should have been. The difference between the red and orange on the third band was so subtle, anyone could have mistaken it for one or the other value. It's for this reason that my meter is always on when I'm loading resistors, and each one gets checked before I load it onto the board. Yes... even if the package is properly labelled, and I've just pulled 3 good ones out of it. I myself got played with this when I first started building, due to incorrect package labelling. The resistor pack was labelled as one range, when in fact it was another. Trust me, when you pull out enough hair, you don't want to go through that twice. I don't care if I've pulled 45 resistors out of a pack, and every one has been the right value as per the package label, I'll still pass them though the DMM. 3 seconds spent here, is hours of troubleshooting saved later.

As for this project, mine fired right up, and the tuning was as per the instructions. The only changes I made were IC types, settling on OPA134's in IC1, IC2, and IC13 positions, and TL072 for the rest of the opamps. But really, it worked fine right out of the gate with the originally spec'd IC's, so I don't see this as a problem. The IC swaps were to please my own ears.

[jstone]

Thanks digi2t. I bought my parts from Mouser and Small Bear. The 3080s came from Small Bear and I've actually ordered four more just to be sure. They should arrive today. I also ordered more trimpots from Mouser.

I'm going to go over all the resistors as you suggest and will report back.

[digi2t]

Thanks digi2t. I bought my parts from Mouser and Small Bear. The 3080s came from Small Bear and I've actually ordered four more just to be sure. They should arrive today. I also ordered more trimpots from Mouser.

I'm going to go over all the resistors as you suggest and will report back.

OK. Good sources on both fronts.

Don't be afraid to take a break from it as well. Sometimes, the harder you stare at something, the less you see. I find that sometimes taking a break from the problem, and doing something completely different, helps reset the brain. Take a few days off, and come back to it fresh. You might be surprised at how quickly the problem might reveal itself, and you'll wonder how you never saw it the first time around. (DOAH!!  )

[jstone]

I went back and checked all the resistor values. Wow! I counted nearly two dozen resistors that are significantly off-spec. R2 is supposed to be a 1M resistor, but I measure it at 220K! I guess I need to order some replacements. From now on I will measure everything before soldering. 

[yuka42]

1) Children boards LEDs behavior:

First I tried to wire the LEDs in the standard "on/off" mode where you "link" both small boards by the "CH" lug. The LEDs never turned off. The weird thing is that both 3PDT were working fine and doing what there were suppose to do but the bypass LED and the channels LED were always on.

Then I removed the link and I wired the CH from the channel select board directly to ground (the "always on channel LEDs option) and all LEDs work fine now, with a small exception that the bypass LED is never fully "off". It is like very weak light while off and very strong while on.

I had the very same issue.  Could only get it to work in the "channel always on" configuration by grounding in to the board.  I don't have the low light bypass issue though....the bypass shuts off.  I like that it shows the channel when bypassed anyway.

If you need to compare voltages I'd be happy to now that mine's up and running.  It has PLENTY of gain.  My Master pot hasn't exceeded noon....

I calibrated the comp trim pot sort of in-between the recommended voltage and all the way off (as kevin states in his build doc....he left it at zero to get more useable control). 

My voltage regulators get quite warm as well.....

I'll have more time to help tomorrow.

[yuka42]

As a side note, I downloaded the build doc when he first posted it, which originally had C51 and C52 at 100uf.  They are now 270uf.  I installed the 270uf. 

[yuka42]

I went back and checked all the resistor values. Wow! I counted nearly two dozen resistors that are significantly off-spec. R2 is supposed to be a 1M resistor, but I measure it at 220K! I guess I need to order some replacements. From now on I will measure everything before soldering.

The measurements of resistor values will not be to spec when installed in a circuit.  It will take into account the surrounding components in the circuit.  I would download a resister colour code app and use that. 

[digi2t]

I went back and checked all the resistor values. Wow! I counted nearly two dozen resistors that are significantly off-spec. R2 is supposed to be a 1M resistor, but I measure it at 220K! I guess I need to order some replacements. From now on I will measure everything before soldering.

The measurements of resistor values will not be to spec when installed in a circuit.  It will take into account the surrounding components in the circuit.  I would download a resister colour code app and use that.

+1. Measuring resistors on the board will not give you accurate measurement. Like yuka says, you're probably reading other resistances in parallel or series with the component you're reading. Get the color chart out, and a good magnifying glass.

[jstone]

+1. Measuring resistors on the board will not give you accurate measurement. Like yuka says, you're probably reading other resistances in parallel or series with the component you're reading. Get the color chart out, and a good magnifying glass.

Good points, thanks. I'm still new to this, but I figured if I just measured from one resistor lead to the other, I would be measuring just the resistance of that single component. I see it doesn't work that way once everything is hooked up.

[bsas]

If you need to compare voltages I'd be happy to now that mine's up and running.  It has PLENTY of gain.  My Master pot hasn't exceeded noon....

Hi yuka42, thank you for helping me!
Here is my detailed and boring voltages:

1) Knobs for ALL measurements:
- All knobs are at noon, except both Volumes which are MAX, Master is MAX, and Limit is around 2:00 (according to calibration instructions);
- Bright switches OFF, Input at HI

2) What I did to calibrate:
Following instructions AS IS, I pumped 1kHZ 30mVA on IC10:PIN2 and I was able to calibrate:
- IC10:PIN6: 4.41 VA (4.337 VA with Comp ON)
- PIN_OUT to 1.171 VA with Comp ON

3) I already measured the D4 and D5 diodes (from Mouser) and they both check -0.6V drop.

4) The Comp LED blinks on the expected moments, when I make a hard attack on my notes, and it fades out as expected too, while the note fades out. The problem is the sound. I am expecting my note to squick (compress) but instead if literally mutes the note completely on the attack and if "fades in" if I let the note ring. Almost sounds like if the moment the Comp LED blinks, my guitar volume drops to zero, and then slowly comes back like I am sweeping the volume knob back to 10 (tricky to explain).

5) Measurements:

- All are made with 4 possible scenarios:
A - 1kHZ 30 mVA on IC10:PIN2 / Comp ON (this is where the problem happens, and the Comp LED is always ON).
B - 1kHz 30 mVA on IC10:PIN2 / Comp OFF
C - 0 mVA on IC10:PIN2 / Comp ON
D - 0 mVA on IC10:PIN2 / Comp OFF

- If I use (...) I mean voltage AC, If I don't, means DC:

IC10:
  | A       | B       | C       | D       |
1 | (0.09)  | (0.09)  | (0.152) | (0.152) |
2 | (0.03)  | (0.03)  | (0)     | (0)     |
3 | 0       | 0       | 0       | 0       |
4 | -15     | -15     | -15     | -15     |
5 | -14.32  | -14.32  | -14.32  | -14.32  |
6 | (4.33)  | (4.41)  | (0)     | (0)     |
7 | 14.96   | 14.96   | 14.96   | 14.95   |
8 | (0.130) | (0.130) | (0.130) | (0.130) |

IC9:
  | A      | B      | C       | D       |
1 | (0.23) | (0.69) | (0)     | (0)     |
2 | 0      | 0      | 0       | 0       |
3 | 0      | 0      | 0       | 0       |
4 | -15    | -15    | -15     | -15     |
5 | 0      | 0      | 0       | 0       |
6 | 0      | 0      | 0       | 0       |
7 | 0      | 0      | 0       | 0       |
8 | 14.96  | 14.96  | 14.96   | 14.96   |

IC11:
  | A       | B       | C       | D       |
1 | (0.117) | (0.117) | (0.117) | (0.117) |
2 | (0.22)  | (0.05)  | (0)     | (0)     |
3 | 0       | 0       | 0       | 0       |
4 | -15     | -15     | -15     | -15     |
5 | -14.32  | -14.32  | -14.32  | -14.32  |
6 | (1.25)  | (3.9)   | (0)     | (0)     |
7 | 14.96   | 14.96   | 14.96   | 14.95   |
8 | (0.130) | (0.130) | (0.130) | (0.130) |

IC12:
  | A       | B       | C       | D       |
1 | (0.74)  | (0)     | (0)     | (0)     |
2 | (0.067) | (0)     | (0)     | (0)     |
3 | (0.067) | (0)     | (0)     | (0)     |
4 | -15     | -15     | -15     | -15     |
5 | 0       | 0       | 0       | 0       |
6 | 0       | 0       | 0       | 0       |
7 | (0.74)  | (0)     | (0)     | (0)     |
8 | 14.96   | 14.96   | 14.96   | 14.96   |

Q1 - 2N5457 (this guy I wasn't sure to measure AC or DC, so I did both):
  | A              | B            | C           | D           |
D | 0              | 0            | 0           | 0           |
S | (1.17) / -0.17 | (3.7) / -3.9 | (0) / -0.42 | (0) / -0.42 |
G | (1.27) / 0.32  | (4.1) / -3.4 | (0) / 0.03  | (0) / 0.03  |

Q2 - 2N3906:
  | A      | B      | C      | D      |
C | 0      | 0      | 0      | 0      |
B | -0.58  | 0      | 0      | 0      |
E | -12.3  | -14.4  | -14.4  | -14.4  |

Q3 - MPSA13:
  | A      | B      | C      | D      |
E | -15    | -15    | -15    | -15    |
B | -13.76 | -14.3  | -14.3  | -14.3  |
C | -14.32 | -0.385 | -0.385 | -0.385 |

My voltage regulators get quite warm as well.....

I measure 100F - 120F :O. I hope that is expected...

[digi2t]

The Comp LED blinks on the expected moments, when I make a hard attack on my notes, and it fades out as expected too, while the note fades out. The problem is the sound. I am expecting my note to squick (compress) but instead if literally mutes the note completely on the attack and if "fades in" if I let the note ring. Almost sounds like if the moment the Comp LED blinks, my guitar volume drops to zero, and then slowly comes back like I am sweeping the volume knob back to 10 (tricky to explain).

You mean like a limiter. Description;

A limiter is a device designed specifically to prevent audio levels from going higher than a specified point. In practice a limiter is basically a compressor with a very high compression ratio (20:1 to infiniti:1). The idea is that once the level reaches a certain point, the limiter will not let it go beyond that.

In practice, that's what this compressor does. The lower you set the pot (counter clockwise), the more it will clamp down on louder signals. When I do use it, I set it right on the edge of my hardest attack, and then a tiny bit more. It doesn't mute my hard attacks, but it does keep them from spiking the rest of the passage.

[jstone]

+1. Measuring resistors on the board will not give you accurate measurement. Like yuka says, you're probably reading other resistances in parallel or series with the component you're reading. Get the color chart out, and a good magnifying glass.

Good points, thanks. I'm still new to this, but I figured if I just measured from one resistor lead to the other, I would be measuring just the resistance of that single component. I see it doesn't work that way once everything is hooked up.

I went back through and checked the color codes of all the resistors I thought were not measuring correctly, and the color codes are all correct. Back to square one. Any other suggestions before I pull the board out and check for solder issues? As I mentioned in an earlier post, the pedal seems to be working, although I'm clearly not getting the correct voltages out of IC10.

[bsas]

+1. Measuring resistors on the board will not give you accurate measurement. Like yuka says, you're probably reading other resistances in parallel or series with the component you're reading. Get the color chart out, and a good magnifying glass.

Good points, thanks. I'm still new to this, but I figured if I just measured from one resistor lead to the other, I would be measuring just the resistance of that single component. I see it doesn't work that way once everything is hooked up.

I went back through and checked the color codes of all the resistors I thought were not measuring correctly, and the color codes are all correct. Back to square one. Any other suggestions before I pull the board out and check for solder issues? As I mentioned in an earlier post, the pedal seems to be working, although I'm clearly not getting the correct voltages out of IC10.

Did you tried to put all knobs at noon? I notice that when some knobs are too high or too low, the IC10:6 never reaches 4.4V in my case. I made sure that all knobs where noon except all volumes and master, which were MAX.

[bsas]

The Comp LED blinks on the expected moments, when I make a hard attack on my notes, and it fades out as expected too, while the note fades out. The problem is the sound. I am expecting my note to squick (compress) but instead if literally mutes the note completely on the attack and if "fades in" if I let the note ring. Almost sounds like if the moment the Comp LED blinks, my guitar volume drops to zero, and then slowly comes back like I am sweeping the volume knob back to 10 (tricky to explain).

You mean like a limiter. Description;

A limiter is a device designed specifically to prevent audio levels from going higher than a specified point. In practice a limiter is basically a compressor with a very high compression ratio (20:1 to infiniti:1). The idea is that once the level reaches a certain point, the limiter will not let it go beyond that.

In practice, that's what this compressor does. The lower you set the pot (counter clockwise), the more it will clamp down on louder signals. When I do use it, I set it right on the edge of my hardest attack, and then a tiny bit more. It doesn't mute my hard attacks, but it does keep them from spiking the rest of the passage.

Well, I know what a limiter should do I already built a couple.

Unless the limiter circuit in this pedal is the most weird I ever heard (and unfortunately I never heard the original amp to compare), I strongly think something is wrong, because if I for example play the note to the edge of the limiter to kick in, the note drops to zero, believe me. I can record if you guys want.

It is like this:

Uncompressed sound wave:
/\        /\       /\
/  \      /  \     /  \
------------------------
     \  /      \  /     \
      \/        \/       \

Expected limiter result:
__       __      __
/  \      /  \     /  \
------------------------
     \  /      \  /     \
      --        --       -

What I am getting:
  _                                /\           /
/  |             /\     /\       /   \        /
------------------------------------------
    (1)  (2)      \/ (3) \/         \   /
                                           \/

(1) = sudden drop and big silence
(2) = note fades in slowly
(3) = eventually comes back partially full (if it is not on the limiter threshold)

Like I said, almost sounds like if I move the volume of my guitar to 0 the moment the limiter grabs the attack, and I slowly fade-in back from 0-10 in a matter of some seconds (between 1-2 seconds).
You can actually hear a "POP" the moment the circuit grabs the attack, almost like if I step in the 3PDT bypass switch.
It is weird!

[teemuk]

Your "expected limiter result" is incorrect.

Basically, it is not a "hard clipping", not even "soft clipping", -type limiter, so there isn't supposed to be distortion at waveform peaks, instead transients and notes with high enough amplitude and sustained envelope should gradually decrease signal gain in order to prevent clipping in stages that follow. Refer to graph titled "limiter with moderate attack and release" for example. It is common that fast attack characteristics may introduce some waveform distortion for very brief moments, like in the graph above but basically the operation of the limiter should be "inaudible" distortion wise.


So if the compressor entirely mutes your signal, pretty much immediately on note attack, and then gradually starts to recover gain when note's amplitude fades it is a sign of the compressor acting oversensitively, or at least introducing way too much attenuation: Instead of -proper amount- of attenuation during attack it entirely mutes your signal. ...and keeps it muted for a long time until below compressor's threshold the gain starts to eventually recover.
Yes, it is common to hear "POP" sound when signal rapidly goes muted. It's the sound of waveform distortion and "edge" in the waveform at the point where it quickly dives from some arbitrary voltage level to zero volts (muted).

The compressor circuit itself is very simplistic: You have feedback loop around JFET gain stage and an operational transconductance amp in that loop. Gain of the aforementioned amp can be controlled with its Iabc input (Pin #5) and since the stage is -inverting- its output signal is subtracted from (input) signal at JFET's gate, which in turn means attenuation of signal. With no output from OTA there is no attenuation (no subtraction) and with OTA output equal to input signal's amplitude you get full attenuation and the signal is effectively muted. These are the circuit elements that actually control gain, everything else is for controlling those elements and establishing the proper operating point.
The input to compressor circuit is a basic passive attenuator consisting of potentiometer and trimmer for fine tuning. Following the attenuator is a gain stage with gain of around 11x and a full wave rectifier consisting of pair of diodes and an inverting unity gain amplifier. Full wave rectified (and filtered) signal controls transistor Q102's base, and that transistor controls Iabc input of the OTA and drives another transistor, which just drives that LED indicator. R183 controls Iabc current and basically sets the operation point of the OTA to match voltage levels of the circuit. Too much current and the OTA is toast, too little and it will not attenuate.
This is a very simple and extremely basic compressor circuit since its attack and decay characteristics are entirely controlled by that single filter cap C139 and associated RC time constants. Not "high fidelity" even by a longshot but likely sufficient enough for a guitar amp.


So, I would check that the trimmer is adjusted correctly (as per instructions in the schematics) and that all component values are correct. OTA, for example, does not accept high amplitude input signals so make sure it actually has that huge amount of attenuation in its inverting input. Do note that in comparison to generic opamp gain stages this one is -open loop- and has tons of voltage gain. You only need few millivolts of signal to generate an output signal with magnitude of couple of volts. Proper input attenuation is therefore extremely important! Also make sure that you got that current limiting resistor in Iabc input right.

[bsas]

Wow!!! Thank you so much for this very detailed explanation teemuk!

You just made me understand at least 5X more of how this circuit supposed to work

Well, I have setup the trim pot according to the schematic (1.17V with the pot set to 2:00). I will double check again the things you pointed out and try to find where the Iabc got so much cancellation.

Thanks!!!