Madbean Total Recall 2024 - Can't calibrate

[Lorok]

Hey there

The problem​

Having been thoroughly enjoying Dan's DM-3 project, I've decided to try to tackle the legendary DMM in the form of MadBean's Total Recall. But even though I can get all the knobs to work as intended (there is delay with a mix from none to full wet, the repeats go from 1 to oscillation, the modulation warbles along), it just so noisy after hours of trying to calibrate.

I have tried both the scope (I have a cheap mini from Aliexpress) and audio probe methods to no avail: there is always noise along the signal path that I can't trim out.

Schematic​

Page 18 of @bean's pdf.

Voltages​

Here are the voltages with reference values from Madbean's build doc linked above in brackets. I notice that starting from the V571D, I have some oscillation that is being picked up by my multimeter in the VDC setting (I note it like "min ... max"), which the reference doc doesn't show:

Knobs @ 0

IC1 TL072CP
1 -7.58 (-7.66)
2 -7.58 (-7.66)
3 -7.55 (-7.62)
4 -15.02 (-15.12)
5 -7.5 (-7.57)
6 -7.57 (-7.61)
7 -7.57 (-7.66)
8 0 (0)

IC2 V571D
1 -13.83 (-13.88)
2 -13.10 … -13.21 (-13.27)
3 -13.10 … -13.21 (-13.27)
4 -15.02 (-15.12)
5 -13.18 (-13.28)
6 -13.15 (-13.26)
7 -5.77 .. -6.02 (-4.87)
8 -13.18 (-13.28)
9 -13.18 (-13.28)
10 -6.58 (-7)
11 -8.22 (-8.72)
12 -13.18 (-13.28)
13 0 (0)
14 -13.17 (-13.27)
15 -13.16 (-13.27)
16 -13.96 (-14.18)

IC3 JRC4558D
1 -7.86 .. -8.21 (-6.65)
2 -3.93 .. -4.11 (-3.32)
3 -3.93 .. -4.11 (-3.31)
4 -15.02 (-15.12)
5 -7.54 .. -7.59 (-7.65)
6 -7.56 .. -7.62 (-7.68)
7 -7.54 .. -7.65 (-7.68)
8 0 (0)

IC4 JRC4558D
1 -7.56 .. -7.62 (-7.68)
2 -7.59 (-7.6)
3 -7.55 (-7.62)
4 -15.02 (-15.12)
5 -7.72 (-7.68)
6 -7.8 (-7.74)
7 -7.71 .. -7.98 (-7.74)
8 0 (0)

IC5 JRC4558D
1 -7.52 .. -7.67 (-7.68)
2 -7.54 .. -7.65 (-7.67)
3 -7.51 .. 7.63 (-7.65)
4 -15.02 (-15.12)
5 -6.49 (-6.91)
6 -6.55 (-6.99)
7 -6.56 (-6.98)
8 0 (0)

IC6 JRC4558D
1 varies (varies)
2 -7.14 .. -7.23 (-7.23)
3 approx -6 .. -8 (-7.3)
4 -15.21 (-15.28)
5 -7.14 .. -7.23 (-7.23)
6 -7.15 .. -7.25 (-7.25)
7 varies (varies)
8 0 (0)

IC7 CD4047BE
1 -6.98 (-7.17)
2 -8.31 (-8.14)
3 -8.25 (-8)
4 0 (0)
5 0 (0)
6 0 (0)
7 0 (0)
8 -15.21 (-15.28)
9 -15.21 (-15.28)
10 -7.84 (-7.51)
11 -7.84 (-7.5)
12 -15.21 (-15.28)
13 -6.95 (-7.11)
14 0 (0)

IC8 MN3005
1 0 (0)
2 -7.84 (-7.5)
3 -10.15 .. -10.46 (-11.56)
4 -10.15 .. -10.46 (-11.56)
5 -15.21 (-15.28)
6 -7.84 (-7.5)
7 -6.81 .. -7.12 (-8.55)
8 -14.28 (-14.31)

IC9 MN3005
1 0 (0)
2 -7.83 (-7.5)
3 -11.83 .. -12.09 (-10.9)
4 -11.81 .. -12.05 (-10.89)
5 -15.21 (-15.28)
6 -7.84 (-7.51)
7 -7.71 .. -7.98 (-7.74)
8 -14.24 (-14.29)

REG L7915CV
I -15.2 (-15.28)
O -17.77 (-17.93)
G 0 (0)

Q1 2N5087
C -13.42 (-13.67)
B 0 (0)
E 0 (0)

18.05vDC isolated supply (18.24vDC isolated supply)
Clock Frequency Range: 75.76kHz to 8.94kHz (76.97kHz to 8.35kHz)

Audio samples​

I have recorded the following:

• input 250Hz sine wave
• Test Point 1 (sounds good)
• Test Point 2 (with high-pitched noise)
• Test Point 3 (with more high-pitched noise)
• example output (with shhh-shhh noise)
• adjusting BIAS 1 while monitoring TP2 (high-pitched noise does not go away).

Frequency analysis​

The high-pitched noise in the Test Points is in line with the 8.94 kHz clock frequency:

Modifications​

• Thinking that my 240pF cap is off by too much, I removed it, added socket terminals and auditioned other 240pF caps. The square wave period at pins 10 and 11 of the 4047 was always in the 110us .. 112us range. This is lower than referenced on David Morrin's calibration page, but I guess this is because my Delay pot is a bit shy of 100k.
• The closest 100KB pot I had was 94k, which is what I used for the Delay pot.
• In some places I used physically bigger caps than what was called for, opting to spend my existing parts inventory instead of spending money on smaller parts
• The relay bypass switching is completely based on pedalpcb's project, but with a smaller relay that was available to me
• Lead dress is not my strong suit, but the wires to and from the in/out jacks are screened with a ground connection at one side

Photos​

In conclusion​

Please let me know if you have any ideas on what I can check and try to get this thing running cleanly, it really would mean so much to me.

Thanks!

 

[bean]

@bean did you have a chance to look into this?

Sorry, I had a lot going on this past week. I will get on this tomorrow if I can. Apologies.

 

[bean]

So good news/bad news. Good news is that (after replacing the chips I took off this build and doing a quick recalibrate) is that I've got no detectable clock noise at the delay output (after the compander) as was demonstrated in the sound clip earlier in the thread. So, this may come down to a build issue.

Bad news I'm getting a lot of gunky noise when the delay pot is turned (like DC across the pot sound) and causing some white noise on the delays as some have described. So, next up I'm going to replace that pot and see what happens since this was not the case when I built the prototype. Could just be the pot is dirty/bad after a year. I'll solve my problem here first then we'll see about solving your guys' problem next.

 

[vires]

In the absence of another solution I’d like to start pulling up resistors to bypass parts of the audio path on the pcb via a wire. The idea behind this is that the white noise issue could be some sort of crosstalk. Does anyone have any ideas on where I might start with this? From my testing the white noise seems to be entering the audio path somewhere between test points 2 and 4.

 

[vires]

Today I attempted to further troubleshoot:

•Pulled R23 and wired it to IC 4

The idea was to bypass the pcb traces that crossed w many others in case there was some sort of crosstalk. Unfortunately the same issue persists so I took a quick video.

I play a bit at the end just to show that it's a beautiful sounding pedal other than this one issue. A shame if we can't make some progress to a solution :/

@Lorok is this consistent with your experience?
@bean any next steps?

 

[rurEEE]

First post here, so hopefully I'm not adding confusion to this topic, but I recently built two of these pedals and have a similar issue on both. It's a similar 'white noise' which only occurs at certain delay times, although it can't be the delay pot itself as the noise comes in and out as the modulation changes the delay time.

Interestingly after playing around with the trim pots (assuming I was doing something wrong with the calibration/gain staging causing the compander to increase noise or something) I noticed that the noise dissapeared when setting the gain 2 trimpot close to minimum (around 9 oclock or below).

@vires @Lorok @bean I'd be interested to know if your builds behave the same way? Perhaps it points to an issue around IC4A's gain stage?

 

[Lorok]

@rurEEE thank you so much for chiming in with your experience.

Ok, going through the Calibration by ear procedure from madbean's build doc I get the following.

1. Audio probe TP1 for signal. This verifies that audio is passing to the input of the first BBD (IC8).

When playing a note I hear it loud and clear. Once the note starts dying down, white noise starts coming in louder and louder. If I deaden the strings with my hand, then the white noise starts oscillating: woosh-woosh-woosh-...

2. Now probe TP2. Adjust BIAS1 until you get the cleanest delayed output possible. It’s usually around the middle of the trimmer.

Without playing anything, I get a high-pitched noise at TP2. As I referenced in my first post, this high-pitched whine coincides with the 8.94 kHz clock frequency.

When playing, I hear a delayed signal through TP2. The whine does not go anywhere. By adjusting BIAS1 I can get the whine to be softer, louder, or disappear completely. I adjust BIAS1 a little off-center CW, since that is where the notes ring out the clearest. The whine is louder than the notes.

3. Go to TP3. Adjust GAIN1 so that audio level is about the same the level that you had at TP1.

The whine is heard here too, louder than the guitar. I leave GAIN1 in the middle as it sounds similarly loud to TP1, but with the whine.

4. Power off and add IC9. Power back up

Skipping this, as IC9 is already installed.

5. Go to TP4. Adjust BIAS2 until you get the cleanest delayed output possible from IC9. You may need to make a slight adjustment on GAIN1 to achieve this. Leave BAL and GAIN2 in the middle for the time being.

The whine is here, but it's softer -- about the same level as the guitar. I adjust BIAS2 a little CW where the guitar is the loudest.

6. At this point, you can check all the controls for function. Turn the FDBK knob up for multiple repeats and the BLEND to 1/2 to hear
a mix of clean and delay signal. Check the DELAY knob range, check the MOD knob for modulation.

There is no whine at the output. With the calibrated settings (FDBK and MOD at 0, DELAY and BLEND at max) I get the expected 1 feedback. There is no whine, no hiss, no white noise, no woosh-woosh.

I can turn the BLEND knob freely: everything is OK. I leave it in the middle position.

When I turn the DELAY knob, I hear some quick noise while turning. It goes away once the turn is complete. It's not a hiss like a dirty pot, it's the sound an analog delay makes when adjusting the delay time, but instead of a note ringing out, it's like the white noise is going up (or down) in pitch. There are positions closer to the middle of the DELAY knob that when I stop there I can hear a noise oscillating woosh-woosh. When I play in these positions, the oscillation is gone. Once I dampen the strings, the woosh-woosh is back. The repeats are NOT clear, they are noisy. As the DELAY knob gets turned down closer to 0 more and more, the delays get noisier and noisier Until finally at 0 it is clear. I check DELAY at max -- still clear, as that is the calibrated position, I guess.

Turning the MOD I get these soft oscillating squeaks of I noise The guitar repeats are indeed modulated, but with this oscillating squeaky noise heard pretty prominently.

Trying different settings of FDBK, DELAY and MOD, I get some that are very good and clear, but most are plagued by noise of different levels.

Interestingly after playing around with the trim pots (assuming I was doing something wrong with the calibration/gain staging causing the compander to increase noise or something) I noticed that the noise dissapeared when setting the gain 2 trimpot close to minimum (around 9 oclock or below).

After setting GAIN2 to 9 o'clock (so a quarter turn CW from the middle) the pedal indeed started being more clear! A lot of the noise went away. Even the repeats seem a lot clearer. MOD is actually usable, since there is no longer any sqeaking oscillation without a guitar signal.

So I went back to check TP2 -- there is no whine in this position of GAIN2!

I don't know if you are getting us closer to the solution, but I certainly confirm that setting GAIN2 to 9:00 made my pedal a lot closer to usable and to what I assume is closer to nominal operation.

 

[rurEEE]

After setting GAIN2 to 9 o'clock (so a quarter turn CW from the middle) the pedal indeed started being more clear! A lot of the noise went away. Even the repeats seem a lot clearer. MOD is actually usable, since there is no longer any sqeaking oscillation without a guitar signal.

So I went back to check TP2 -- there is no whine in this position of GAIN2!

I don't know if you are getting us closer to the solution, but I certainly confirm that setting GAIN2 to 9:00 made my pedal a lot closer to usable and to what I assume is closer to nominal operation.

@Lorok That's great to hear! And definitely sounds like we're onto something... I didn't notice the whine you've talked about at TP2 (or that gain 2 affected it) but it's possible my builds are the same. I wasn't too concerned about clock noise as long as it wasn't audible at the output, but I should have some time this weekend to try this out and get back to you.

I've also just ordered some different opamps (TL072) to try dropping those in and see if that has any affect on the noise, as I figured it could be a quirk of how the specific 4558s I have were behaving in this circuit, so will report back on that too!

 

[vires]

@rurEEE thanks for your input on this. I've brought this up on the groupdiy forum in hope that we can get more suggestions.
Discussing the white noise:
- - - -

- - - -
I don't have an oscilloscope - but if this seems worth checking out and someone can measure next time they are in front of it, I can report back to and from the other forum. I'll also check out turning gain 2 down to 9:00 and share my results...

 

[vires]

On turning down the gain 2 pot to around 9:00 or so:

The white noise can indeed be made to disappear. I also notice at this level subtle but audible clock noise on the repeats. Pedal will not self oscillate with gain 2 in this position (even w feedback knob maxed) is the other thing I made note of. So clearly not properly calibrated like this, but it presents some good (and new) information for sure!

It is encouraging that we may be getting closer to a solution. I look forward to helping in any way I can

 

[Lorok]

I don't have an oscilloscope - but if this seems worth checking out and someone can measure next time they are in front of it, I can report back to and from the other forum. I'll also check out turning gain 2 down to 9:00 and share my results...

Does this help? Probing pins 10-11 of IC7. First with GAIN2 at middle, then at 9:00. The audio is from the pedal output.

 

[rurEEE]

I haven't had a chance to crack out the full calibration/testing rig, but the TL072s I mentioned previously arrived so I thought I'd try swapping one of those in for IC4 and see whether that affected the issue. Long story short, the problem appears to be completely gone, but I recorded a video to show how one of my pedals behaves with the 4558, then with the TL072.

This still suggests something strange is going on, so I'll still have a go with the audio probe when I can to try and see exactly where the noise is coming from with the 4558, but hopefully this is a useful clue for someone who knows more about the circuit than I do!

 

[Lorok]

I swapped my IC4 from a 4558D to a TL072 and it just made the noise at the output louder

IC1 (TL072) Pin 7: No noise
IC2 (V571D) Pin 6: No noise
IC2 (V571D) Pin 7: Woosh-woosh noise
IC2 (V571D) Pin 14,15: Loud squeaky noise

By turning the GAIN2 knob down, the noise at IC2 pins 7, 14, 15 gradually disappears.

 

[rurEEE]

I swapped my IC4 from a 4558D to a TL072 and it just made the noise at the output louder

IC1 (TL072) Pin 7: No noise
IC2 (V571D) Pin 6: No noise
IC2 (V571D) Pin 7: Woosh-woosh noise
IC2 (V571D) Pin 14,15: Loud squeaky noise

By turning the GAIN2 knob down, the noise at IC2 pins 7, 14, 15 gradually disappears.

Strange! I'm having a play around now probing different points in my pedal and I think we may have separate/different issues. From your video probing the IC7 pins 10/11 the noise is coming in and out as if the modulation is turned up, and it still seems to be there after you turn gain 2 down. If my modulation is set to 0 then the noise is pretty constant, without of the 'wooshing' effect that you're getting. I'd be interested to know if the 'wooshing' changes frequency with the C/V switch?

Checking the different test points on my pedal (all with blend at 100%, level around halfway, feedback at minimum, delay at max, and mod at minimum):

TP1 has no clock noise, as expected. If I feed a signal it sounds clear with no noise. If I turn off the signal some white noise will slowly creep up and stay at a constant level. I assume this is the compander bringing up the level gently as TP1 is after the first half of the compander chip. If I turn the level control down this seems to reduce the level of noise, which would make sense if it's just the compander amplifying any background hiss on the input. All of this is unaffected by the Gain 2 setting.

TP2 has some clock noise whether I feed signal or not, but again this seems normal as there is no BAL trimmer to cancel it out perfectly. Again, this is completely unaffected by Gain 2's setting. The clock noise does seem to be louder when signal is there vs when muted but that may just be a quirk of the delay chip?

TP3 is the same as TP2, the only difference is that Gain 1 affects the level of this signal as expected.

TP4 is where things get interesting - I have a similar sound to TP3 but with less clock noise as this is cancelled (as best possible) with the BAL pot. However, if I adjust Gain 2 I start to hear and see distortion on the waveform. See below video:

As TP4 is before gain 2's gain stage (IC4A) this suggests to me that for some reason Gain 2's adjustment is messing with the bias/operation of IC9 (second MN3005) as the only real difference between TP3 and TP4 is that TP4 is after the IC9.

To test further I probed the pins of IC9.
Pin 1 (ground) - no noise or signal as expected, gain 2 has no affect.
Pins 2/6 (clock signal) - just pure clock noise, no affect from gain 2. Affected by delay time/mod as expected.
Pins 3/4 (outputs) - Signal with clock noise as expected, however when adjusting gain 2 interference comes in and out (same as above video)
Pins 5/8 (power) - no audio signal as expected
Pin 7 (input) - Signal as expected with clock noise. When signal is muted it is possible to hear some interference as gain2 is adjusted, but at a much, much lower volume.

So now I checked voltages on IC9 with different gain 2 settings, still max delay time and no mod:

Gain 2 at minimum (no noise issues):
1: 0.00
2: -7.62
3: -11.48
4: -11.51
5: -15.20
6: -7.62
7: -7.99
8: -14.23

Gain 2 at around 11:00 (lots of noise on TP4/Output)
1: 0.00
2: -7.62
3: -11.41
4: -12.01
5: -15.20
6: -7.62
7: -7.99
8: -14.23

So voltage at pins 3/4 seems to be getting affected by gain 2 adjustment. Not sure if that's normal or not, so I tried some different delay times and the voltage at pins 3/4 seems to decrease (more negative) as the noise/interference gets worse.

Swapping back in the TL072 in IC4 the noise is not present, and the voltage at pins 3/4 is constant at -11.58 and -11.61 respectively at max delay time, no matter how gain 2 is adjusted. The voltage decreases (more negative) as delay time is shortened down to around -12.34v on both pins at minimum delay time.

At this point I don't know enough about delay chips/circuit design to know if this voltage change is causing the noise or a consequence of it, but hopefully someone who does might be able to chime in. @bean any thoughts?