Creamery hi expansion issues

[comradehoser]

Well, I'm almost done with building a rack of pedals, and out of (13? I think?) 2 are giving me weird behavior.

One is the okko diablo I am building for a friend, and it will get its own thread.

This one: creamery compressor because I wanted to see what squishy compression could do for me.

Subbed rc4558 for the 4 xx4558 op amps, THAT 2181c for the 2159, 2n4124 for mrs4124, 51p mlcc for 50p. Did not pull any resistors according to build reporting that it wasn't necessary.

The issue: compressor compresses nicely, I haven't noticed any major noise. However, if I turn up the hi-ex knob at all, I get a harsh distortion on the attack of a note. The knob doesn't seem to much change the sound either, although the subtlety of compressors eludes me.

Is this SOP for the pedal, or should I pull that resistor out, or does this sound like something else?

 

[comradehoser]

Let me provide a link to video.

My video won't upload, so here you go

Ibanez ar4something on bridge humbucking mode into little orange cube amp.

It's only when hi-ex is turned up at all; sounds like the level is momentarily clipping during the initial attack before the compression has a chance to bite. Generally is very consistent when you are hitting a note from lower volume or silence, not there if compression is already engaged. Then if you turn up attack CW, it turns into some ugly distortion

Is this just what this compressor does? Am I suffering from "I think this pedal should sound different but this is actually what it sounds like" syndrome again?

 

[comradehoser]

Is it just that the input signal is too hot?

[edit: edit:] I thought I had switched 47k and 470k resistors, but not so after all.

So, still stuck on the source of the frapping noise

 

[comradehoser]

All parts correct and correctly installed, but without 2181c compensation.

 

[comradehoser]

Removed r31 per 2159 to 2181c conversion, put in a jumper, and it actually made the problem worse.

Kinda lost as compressors are not my jam

[edit: failed to notice that I had pushed Q2 out of the socket when I was removing r31. Issue still remains, but not as bad as without Q2]

 

[Feral Feline]

Sorry, I'm of no help, but I've been following this thread since you launched it.

Good on you for carrying on without any forum input so far!

Hang in there, keep at it.

 

[comradehoser]

Merci monsieur felin.

Guess I'll just do the usual, do a voltage reading and then do an audio probe. I just wish somebody could tell me if I actually have an issue or not.

 

[comradehoser]

Looks surprisingly consistent, except IC 4.

Measured with the setting that I last had on it, Level: 12:00; Attack 12:00; Comp full; Hi-Ex full.

Forgot to measure transistors, I'll add in those readings later.

In: 9.39v

IC1 (RC4558):
1: 4.487
2: 4.579
3: 4.574
4: 0
5: 4.574
6: 4.574
7: 4.574
8: 9.12

IC2 (THAT 2181C)
1: 4.576
2: 4.806
3: 4.571
4: 4.802
5: 1.743
6: 4.573
7: 9.13
8: 4.570

IC3 (RC4558)
1: 4.506
2: 4.571
3: 4.575
4: 0
5: 4.576
6: 4.572
7: 4.563
8: 9.13

IC4 (RC4558)
1: 8.03
2: 8.03
3: 8.02
4: 0
5: 4.577
6: 4.574
7: 4.559
8: 9.13

IC5 (THAT 2181C)
1: 4.575
2: 4.572
3: 4.566
4: 4.574
5: 1.715
6: 4.577
7: 9.13
8: 4.574

IC6: (RC4558)
1: 4.889
2: 4.574
3: 4.576
4: 0
5: 4.576
6: 4.576
7: 4.572
8: 9.13

Q1 (2N4124):
e: 0
b: 0.02
c: 8.41

Q2 (2N4124):
e: 0
b: 0.01
c: 8.42

Q3 (Interfet through-hole J201):
d: 8.36
g: 9.13
s: 8.41

 

[jwin615]

Do you have a signal generator?
Maybe try feeding a 2k test tone through it and probing? Can also use YouTube, worst case.

 

[comradehoser]

Is the one side of ic4 normal that it reads 8v when everything else is at 4.5?

Jwin, yeah, audio probe is next, I guess. I usually just patch some music or a loop.

Would a constant tone generator reveal much? The issue seems highly dependent on attack, esp. in the lower frequencies. Turning the hi-ex up any increment with attack up any increment seems like it is boosting or allowing in a too-hot raw signal; it sounds like it is exceeding the headroom of the pedal before the compression clamps down. Once the compression is on, though, the signal stays compressed and it's fine, so you can slide between chords or whatever. The issue happens when the signal drops and the compressor has to ramp up again in response to a new attack--I don't know if I should be aiming to have a clean-ish signal with attack fully CW?

Attack all the way CW (delayed compression) is consistent with a too-hot raw signal, it results in pretty much a fully clipped/distorted signal. Not in a pretty way, either.

Could it be my j201 is giving too much juice?

I don't know, I might just have bad luck/bad choice with compressors:
Aion polyphemus/philosopher's tone: basically, a distortion pedal.
Delegate boneyard edition: very very very subtle bordering on meh.
And then this one.

 

[airbud]

Is this just what this compressor does? Am I suffering from "I think this pedal should sound different but this is actually what it sounds like" syndrome again?

I just wish somebody could tell me if I actually have an issue or not.

Man, I really feel these vibes.

I'll check out the clip when I get home. I've only built two compressors but have spent a ton of time with studio compressor plugins, for whatever it's worth.

 

[airbud]

This may or may not be helpful, but for the sake of linking threads together that may possibly be helpful:

Post in thread 'Obsolete parts… did I just waste $12?' https://forum.pedalpcb.com/threads/obsolete-parts…-did-i-just-waste-12.18016/post-237375

 

[comradehoser]

Thanks Airbud!

I did check that out (of course). Thought I'd posted, but not until now.

How's your heterodyne adventures going?

I was really being sincere when I said I felt your pain, because I'm feeling it right now, as you can see.

At this point, I think I might just go to build madbean's trash compactor pedal and call it a day.

[edit: maaaan, sold out.]

 

[comradehoser]

Update:

Audio probed things and determined that the level clipping was happening in the loops around IC6.1 and IC6.2 and the level pot: detected at IC audio output pins 1 and 7, and c12, c13, c14; r25, r26, r27, r28. I do have c12 and c14 as 51pf as opposed to 50pf--surely 1pf can't make that much difference?

Figured it might just be a bum IC, so I subbed in what I had, TL072 and JRC4558, both of which resulted in a reduction in the clipping. Subbed JRC4558 into all of the ICs; the clipping is much less but still present, especially once compression, hi expression, and attack are around or past noon. Ditto with the harsh distortion once attack is past 3:00--still there, but less.

Frankly stumped. The only component I would maybe look askance at would be the level pot since it seems to be in the middle of things? Or I just need to find a chipset that has a high enough headroom to absorb the gain/level spikes?

It looks like the THAT chips are not involved in signal at all, but instead seem to regulate voltage to the amplifier ICs, so is it possible that IC5 immediately before the IC6 has something odd with it--that it's pushing too much signal or maybe not enough into IC6?

Do I even have an issue?

 

[MacGuffin_]

Update:

Audio probed things and determined that the level clipping was happening in the loops around IC6.1 and IC6.2 and the level pot: detected at IC audio output pins 1 and 7, and c12, c13, c14; r25, r26, r27, r28. I do have c12 and c14 as 51pf as opposed to 50pf--surely 1pf can't make that much difference?

Figured it might just be a bum IC, so I subbed in what I had, TL072 and JRC4558, both of which resulted in a reduction in the clipping. Subbed JRC4558 into all of the ICs; the clipping is much less but still present, especially once compression, hi expression, and attack are around or past noon. Ditto with the harsh distortion once attack is past 3:00--still there, but less.

Frankly stumped. The only component I would maybe look askance at would be the level pot since it seems to be in the middle of things? Or I just need to find a chipset that has a high enough headroom to absorb the gain/level spikes?

It looks like the THAT chips are not involved in signal at all, but instead seem to regulate voltage to the amplifier ICs, so is it possible that IC5 immediately before the IC6 has something odd with it--that it's pushing too much signal or maybe not enough into IC6?

Do I even have an issue?

Yo, having the same issue (to a different degree in 2 different creamery builds). So far I have found that if level is high, the clipping problem goes away, or at least lessens (even if you lower volume elsewhere). Could be something with the jfet?

As for my ICs, I am using NJM4558 (or NLM as it reads from the bag Tayda sent, but they are NJMs), and 2N4124 NPN as subs

 

[comradehoser]

So strange! Maybe if the level is up the pedal is more likely to already be in compress mode, since the pop seems to happen on the initial attack--like the pedal lets in a smidge of over-boosted signal before it clamps the signal.

Well thank you so much . I guess you're confirming that the pedal is operating as designed and I should quit obsessing over it.

Cheers, bud, I appreciate that for real.

 

[MacGuffin_]

Quick question for you, does yours have a pretty bad pop when you turn it on? (especially at higher lvl). This is the first pedal I have had that does this to me. May just be DC leakage (in which case I believe I can use a 1M resistor from the input/output of the PCB (from the switch) to ground to fix it? I've never had to do pull-down before, so if anyone can confirm.

 

[benny_profane]

Quick question for you, does yours have a pretty bad pop when you turn it on? (especially at higher lvl). This is the first pedal I have had that does this to me.

That's not uncommon for true bypass compressors. With the level high in bypass, the input signal is essentially negative infinity, the change of state is then amplified (whether that's AC, DC, or mechanical interference).

I haven't looked at the schem to confirm that the impedance checks out, but you could wire it so that the input is always connected (i.e., just switch the output for active/bypass). The compression circuit would always "see" the guitar signal and the transition from bypass to active would not be jarring to the circuit input stage.

 

[MacGuffin_]

That's not uncommon for true bypass compressors. With the level high in bypass, the input signal is essentially negative infinity, the change of state is then amplified (whether that's AC, DC, or mechanical interference).

I haven't looked at the schem to confirm that the impedance checks out, but you could wire it so that the input is always connected (i.e., just switch the output for active/bypass). The compression circuit would always "see" the guitar signal and the transition from bypass to active would not be jarring to the circuit input stage.

Tried this with a quick jumper held in place with clips, didn't do much to stop the clipping (first image). Would 1M from both pins in the 2nd pic do the pull down thing?

 

[benny_profane]

Tried this with a quick jumper held in place with clips, didn't do much to stop the clipping (first image). Would 1M from both pins in the 2nd pic do the pull down thing?

Neither of those schemes will work. Remove the wire from the breakout IN pad (the one going to the main circuit) and tie it to the far left pad (i.e., the input jack connection).