Thumb Sucker Compressor distorts instead

[matopotato]

At the end of this thread is where I am hoping for some more ideas:
https://www.diystompboxes.com/smfforum/index.php?topic=105610.msg951851#msg951851

[Eb7+9]

FYI, your recent efforts in other threads caught my eye ...

... my observation of the distortion is that turning knobs had no or little impact. Of course zero Ratio made the distortion almost go away, but so did any compression.

I don't want to give you any false hope here but ...
for what it's worth this statement you made happens to take you very close to the heart of the "crackling" problem ...

yep, I duplicated this on the bench and scoped it - it's a real thing folks ...
even my TINA sims manifest the same crackling on wave file playback under certain (normal'ish) conditions ...

and btw, I just luv the attacks on accident victims - hilarious

---

seriously though, trust me, none of you can or will figure out what's going on here,
as it was tough even for me ...

in the end it's really a structural issue and so cannot get fixed via random tweaks and farts to the structure "as-is" ...

---

took me 4 re-designs of THUMBS just to figure out the cause and how fix it

by the time I got rid of the other problems (noise and the overall response class) the design was now completely different and original ...

and sounding darn good, with low noise and distortion, a switch for ALC/comp modes and far ranging controls (2), ... now, a stable design

here's a slightly extreme example of the kind of spanking a circuit in this class can do when configured properly ...

providing a very unique AGC feel, unlike most other comps ...


viva Analog "COMP-ALC" DEMO

nothing to do with original THUMBS structure nor what became of it later

--- ain't no Dyacomp competitor either


sorry to say - but I think your chances of solving this systemic problem are basically next to impossible ...
you might wanna drop out of the rabbit chase here

[R.G.]

and btw, I just luv the attacks on accident victims - hilarious
[...]
seriously though, trust me, none of you can or will figure out what's going on here,
as it was tough even for me ...
[...]
sorry to say - but I think your chances of solving this systemic problem are basically next to impossible ...
you might wanna drop out of the rabbit chase here

Wow. I should have learned by now and not been surprised by this, but it's still a full-on "wow" moment. It >is< a fairly accurate capsule of your "contributions" though.  You dug up a nearly two-year-old necro-thread, used it to say how clever you are and how nobody else is as smart as you.

How's that Dale Carnegie "How To Win Friends and Influence People" course coming?  Or was it just a bad day?

[PRR]

How's that Dale Carnegie How To Win Friends and Influence People course coming? 

4.7 stars on Amazon! 10K+ bought or read in past month.
https://www.amazon.com/How-Win-Friends-Influence-People/dp/0671027034
It helped Donna Reed and Charles Manson!

[duck_arse]

.... the attacks on accident victims ....

who? where? what now?

[Eb7+9]

.... the attacks on accident victims ....

who? where? what now?

some replies (to issue posts) from the op and a couple of others have been unnecessarily severe and condescending imo

---

I have posted a bunch of times about how the original THUMB's design was similar in structure and function to the NE570 comp/alc applications ... that is, with the addition of variation in gain factor and with the OTA replacing JUNG's gain-cell // ... the 570 didn't do variable ratio or threshold very well, THUMB's could ... if it were better configured (that was my first build - a partially corrected THUMB's)

but the op doesn't seem to like the comparison ...

admittedly, I have not seen any analysis on this circuit other than what op presents and then my sims and bench testing ... for the most part I'm recreating his exact results (eg., his 15mV rectifier threshold versus my 13mV) ... and as far as transfer curves go, THUMB's turn out to be the same as 570 results when configured in ALC mode

otherwise,

to understand the "frackling" issue better consider how the SIGNETICS and THAT corp designers do one thing with the rectifier side-chain circuit capacitor that THUMB's isn't careful about ...

[R.G.]

some replies (to issue posts) from the op and a couple of others have been unnecessarily severe and condescending imo

Um - the op and a couple of others were unnecessarily severe and condecsending?
Let's see - you said:

[...]
seriously though, trust me, none of you can or will figure out what's going on here,
as it was tough even for me ...
[...]
sorry to say - but I think your chances of solving this systemic problem are basically next to impossible ...

And you think that OTHERS were unnecessarily severe and condescending?

I get that you think that you're the smartest guy on the planet. Your posts historically have made this conceit very apparent. But declaring that you know the secret answer, only you are smart enough to figure it out, you're not going to tell; and then being offended that OTHERS were not nice to you is the kind of thing I'd expect from an ill-mannered child.

Why don't you either help the guy out -or- just not reply at all? Snootily declaring that only you are smart enough to figure it out and that other people were mean to you is just juvenile.

imo

[matopotato]

some replies (to issue posts) from the op and a couple of others have been unnecessarily severe and condescending imo

Hi,
I do not understand in what way I may have offended you. (As I am the OP of the thread)
Can you please share some quote(s) for my understanding?
I try to be civil and appreciate most replies and help I get, so your description sounds a bit "out of character" in my humble opinion.
Thanks in advance.

As for the build itself, I gave it up a few years ago and used the build as fodder for other projects. I built about 4 other compressors and no longer feel any need for the "thumb" type.
Cheers

[matopotato]

but the op doesn't seem to like the comparison ...

admittedly, I have not seen any analysis on this circuit other than what op presents and then my sims and bench testing ... for the most part I'm recreating his exact results (eg., his 15mV rectifier threshold versus my 13mV) ... and as far as transfer curves go, THUMB's turn out to be the same as 570 results when configured in ALC mode

I am beginning to suspect you mixed me up (original poster) with the circuit's original designer. Which is someone else. I only bought and  built a kit (Thumb sucker I think) based on pedalpcb version of Engineer's Thumb if memory serves me right. Then I got some buzz/distirtion and started this thread for ideas to fix.

So I think you got different people mixed up.

[duck_arse]

matopotato - I'd be more inclined to ask our friend about his reference to "accident victims". I can see absolutely no way that terrm can be applied except as a personal insult.


I wouldn't expect an answer, though.

[matopotato]

matopotato - I'd be more inclined to ask our friend about his reference to "accident victims". I can see absolutely no way that terrm can be applied except as a personal insult.


I wouldn't expect an answer, though.

Ok,
Not fully sure what that phrase would mean though...

EDIT: I thought it was accidental victim. But now I understand better after some googling.
Thanks

[Eb7+9]

Hi,
I do not understand in what way I may have offended you. (As I am the OP of the thread)
Can you please share some quote(s) for my understanding?

sorry for being so un-clear Matopotato,
by OP I meant merlinb - the author of Thumbs

the last time we spoke directly you had left me with a request ...

https://www.diystompboxes.com/smfforum/index.php?topic=105610.msg1249828#msg1249828

by whence I went and rented myself a small shack in the woods for a couple of months and pulled some long solder runs
that's where the Thumbs Rescue unit was first built and evolved, among other things I was developing concurrently

sorry I never came back to spill the beans, but hey ...
at least I'm providing impetus to those who think they might be able to pull off a re-design

besides, it sounds like you may have already found something

---

I'm not upset at anybody saying whatever to anybody else - who cares

after all, I got what I needed from merlin's project ...

an idea ... one I think is important

that is, the idea of using a lowly OTA (in my case full discrete)
instead of a high linearity / high headroom Jung gain cell (570/571) or Blackmer VCA
as in most active current feedback amplifier designs nowadays

Lke most circuit designers would probably be I was originally skeptical at first of the premise of the design ...
mainly because of the obvious promise of suffering signal-to-noise degradation as a result of being forced to throw so much signal away to accommodate
the OTA's low input headroom on the return

of course, that's just one aspect here

---

ok, so by now merlin definitely doesn't seem too keen on discussing the crackle side of his invention that some people have experienced ...
or even discussing his design's broader lines for that matter ...

the aim was to try and get a conversation going // c'est tout !

---

here are some comments I was referring to above ...

https://www.diystompboxes.com/smfforum/index.php?msg=1248203
https://www.diystompboxes.com/smfforum/index.php?msg=1203847
https://www.diystompboxes.com/smfforum/index.php?msg=1203847

btw, to answer his claim in the last one
I went thru a spat of provocative conversations with mictester around 2018/2019 ...
a dude who kinda knows a few things about comps
he's the one that convinced me to stop wasting time blabbing on forums ... :p
and get on with reversing/sim'ing the 570 for myself ... which turned out to be a fun bit of work

... and from there I did Thumbs soon thereafter

again, I don't care who treats who how ... it's just a little too bad we couldn't get into all that a bit more here

again, I'm only interested in tearing apart designs so I can better understand how each bit works
it's called circuit analysis

no emotions // nothing personal

though I feel a little bad for the other members around here who might lack the design chops to follow fully the angle I'm trying to advance here
and more so boring as heck reading all this crap without finding a dang schematic prize at the end

my apologies ...

otherwise, an open discussion would have shown where the side-chain was excavated from
(see Horowitz Fig 3.38 and 4.36) and much more revealed ... um, yeah - an open discussion
as opposed to my long boring monologues

I'm a big fan of the double-rectifier stages in Thumbs
no other compressor design has anything like it ...
a lot of work for half-wave rectification and attack/release controls that aren't effective
(mucking with op-amp feedback in the first rectifier stage)

my $0.02 take on things here

As for the build itself, I gave it up a few years ago and used the build as fodder for other projects. I built about 4 other compressors and no longer feel any need for the "thumb" type.
Cheers

fair enough ...

I obsess all the time about compressor designs - keeps crossing my mind for some reason
in fact, quite luckily and randomly, a second new approach to the side-chain in this style of architecture (thumbs-v1 that is) popped into my head this weekend
ie., producing the same fully variable current-feedback amplifier architecture using a discrete OTA for active current sourcing

this time I'm taking the last part of Thumb's side-chain and turning that into its full-wave equivalent
throwing away the useless first bit and using an alternate approach to charging the averaging cap

another thing of beauty if I may say so .. mainly, the feel is really good - just like my first one

and thus, I must say, Thx for the inspiration Mr. Wizard!

[matopotato]

My search ended with the Pumpernickel
https://schalltechnik04.de/en/instructions/pumpernickel
It is based on THAT4301P20-U and originally intended for bass.
I am very glad I managed to buy one of the last ones available. I think the pcb is no longer made and the schematics were not public when I built it. Instead I got help from the designer when I got stock in one place.

[PRR]

schematics were not public when I built it.

https://www.freestompboxes.org/viewtopic.php?t=31550

Pretty much the basic dBx compressor adapted to instrument levels. This family of dBx was always my 2nd favorite.

[matopotato]

schematics were not public when I built it.

https://www.freestompboxes.org/viewtopic.php?t=31550

Pretty much the basic dBx compressor adapted to instrument levels. This family of dBx was always my 2nd favorite.

Thanks a lot for this, Paul!
I built my unit in October 2022 and was googling abound for schematics but never found them, which is strange since they were posted quite some time before.

So what is your first favorite?

[Rob Strand]

FYI the earlier non-X versions of the  dBx units didn't use VCA chips.  They use transistors with some adjustment pots.   Be interesting to see if a DIY build could work.   (As for it competing with a THAT VCA I doubt it.)   At one point someone posted a transistor version of the RMS converter.

Off hand it's like the dBx 163 and dBx 163x.  Popular with bass players maybe late 70's early 80's.

I didn't catch this thread when you started it.  Be interesting to work out why the Thumb Sucker distorts.

[matopotato]

FYI the earlier non-X versions of the  dBx units didn't use VCA chips.  They use transistors with some adjustment pots.   Be interesting to see if a DIY build could work.   (As for it competing with a THAT VCA I doubt it.)   At one point someone posted a transistor version of the RMS converter.

Off hand it's like the dBx 163 and dBx 163x.  Popular with bass players maybe late 70's early 80's.

I didn't catch this thread when you started it.  Be interesting to work out why the Thumb Sucker distorts.

Thanks Rob,

As for the original cause for this thread I already used some parts for other projects although most is still there. If it is component(s) being faulty, then new projects might just transfer the problem to a new build.
I agree that the reason behind the issue would be nice to figure out. As I understand there are builders of that same kit that got it to work properly. A few might have had the same problem as well.
The plot thickens but the trail is going cold.

[Rob Strand]

As for the original cause for this thread I already used some parts for other projects although most is still there. If it is component(s) being faulty, then new projects might just transfer the problem to a new build.
I agree that the reason behind the issue would be nice to figure out. As I understand there are builders of that same kit that god it to work properly. A few might have had the same problem as well.
The plot thickens but the trail is going cold.

It seems odd only some people complain about it.  Is it some people have the problem and some don't?   Is it all Thumb builds or only that kit?   From Eb7+9's comments you would expect all builds should do it.

I did a quick scan yesterday (or the day before) and some of the fixes involve limiting the pot range.  To me that could simply hide a bigger problem.   It's quite normal to get overloads or distortion with the wrong attack/release settings. 

Another possibility is when you run out of gain reduction.   However, the problem seems more than that.    A region of unexplored bad behaviour is when the gain of the opamp need to be less than unity for the non-inverting opamp.   I did some calculations a while back and I think theoretically it should work but what I didn't check is if the circuit itself can actually operate correctly under those conditions.

The problem can only be solved by building it and weeding through the fine details.