PMC - Poor Man's Compressor

[antonis]

IMHO, you shouldn't try 4 transistors 'cause 3 BJTs current mirror delimits Poor - Wealthy man margins..

[rankot]

Listen to it here:

Nice sound!

[amz-fx]

Nice project! Thanks for sharing!

Best regards, Jack

[Eb7+9]

I found that with a regular 2 transistor current mirror the output impedance wasn't high enough, so the bias point moved around with input signal level, causing control breakthrough and audible thumping. When I upgraded to a 3 transistor Wilson mirror it fixed this issue. I didn't try the 4 transistor mirror. Maybe I should have. 

your ota is driving 5k(ac) to ground at the output - it's not a Zout issue // a plain-jane (non-cascoded) discrete ota will more than do here ... i'm not asking for nothing as it's obvious why your bias (down stream) would drift wildly ... it's a similar situation to the other engineer'ed limiter in this forum that includes an ota ... hint: where luck plays a big, sometimes disastrous, role

[Rob Strand]

If you think about the offset I am seeing, the bias point is higher than expected, adding 100 ohms in series with the base of TR6 would actually make it slightly worse in this case since it will make TR6 turn off a tiny bit more. It would also increase the noise floor slightly - this topology is very sensitive to base resistance, I find even 100 ohms is barely acceptable.
I think it's more likely small mismatch in VBE of the LTP transistors. In any case it doesn't matter much

Yes, the offset goes the wrong way for adding a resistor to TR6 - I only eyeballed the asymmetry.  It could be imbalance in the mirror since that also needs to match Vbe's.  A common balancing act is to add emitter resistors to the mirror  (provided they match better than the Vbe's).   Like you say the imbalance doesn't seem to matter, sounds fine.

I always like your designs because you aren't afraid to keep the impedances low to keep the noise down.

[jonny.reckless]

your ota is driving 5k(ac) to ground at the output - it's not a Zout issue // a plain-jane (non-cascoded) discrete ota will more than do here ... i'm not asking for nothing as it's obvious why your bias (down stream) would drift wildly ... it's a similar situation to the other engineer'ed limiter in this forum that includes an ota ... hint: where luck plays a big, sometimes disastrous, role

I don't know what you mean, can you explain? The bias doesn't drift wildly, it does move a little with input signal level but it's not audible. The Wilson mirror is there for a reason: the performance was inferior without it. You need fairly good static and dynamic current mirroring in this topology to avoid thumps and audible artifacts.

[rankot]

your ota is driving 5k(ac) to ground at the output - it's not a Zout issue // a plain-jane (non-cascoded) discrete ota will more than do here ... i'm not asking for nothing as it's obvious why your bias (down stream) would drift wildly ... it's a similar situation to the other engineer'ed limiter in this forum that includes an ota ... hint: where luck plays a big, sometimes disastrous, role

Which one?

[11-90-an]

Sorry if this is a very idiotic queston but...

Are Vdd and +9v the same in this circuit? If so, then why are there 4 decoupling caps, 2 100nf and 2 100uf? Or can they just be replaced with 1 100nf and 1 100uf to lower part count?   

[antonis]

Are Vdd and +9v the same in this circuit?

No they don't..!!
(you can realize it either by measurents taken or by V_BE drop on TR3/TR4 Emitter follower (Sziklai pair)

[11-90-an]

No they don't..!!
(you can realize it either by measurents taken or by V_BE drop on TR3/TR4 Emitter follower (Sziklai pair)

So do I connect anything to the VDD...? 
Or is it just a label that has no significance on breadboard? (Sorry for my stupidity)

[antonis]

It's never a stupidity matter ..

You probably don't undestand well overal circuit function..

[jonny.reckless]

So do I connect anything to the VDD...? 
Or is it just a label that has no significance on breadboard? (Sorry for my stupidity)

There are no stupid questions, only stupid answers 
9V is the supply rail that comes into the pedal from your PSU. This supply powers the side chain transistors TR12 and TR13 and their associated components. Then there is a filter / smoother comprising TR3 and TR4 which generates VDD. VDD is about 0.7V lower than 9V, but has much less noise and ripple on it. VDD is used to power the rest of the circuit which is much more sensitive to noise. When you are playing clean it's not usually a problem but if you put a compressor in front of a drive pedal for example, the noise starts to become really obvious, and I wanted to keep it as quiet as possible, in terms of both hiss and hum. You could certainly build the unit without the extra VDD power supply and it would work but it will be a bit noisier depending on how high quality your power supply is. If you are running from batteries this will likely be fine. I use a Voodoo labs pedal power and that is also pretty clean but some of the cheaper 9V adapters generate a lot of high frequency switching ripple and 120Hz hum on their outputs, sometimes 10 - 20mV or even more which will adversely affect a design like this without extra filtering. Feel free to experiment!

[rankot]

There are no stupid questions, only stupid answers 
9V is the supply rail that comes into the pedal from your PSU. This supply powers the side chain transistors TR12 and TR13 and their associated components. Then there is a filter / smoother comprising TR3 and TR4 which generates VDD.

Sorry to bother you Jonny, but how exactly this transistor pair filters the noise? I'm really curious to learn this!

[Eb7+9]

You need fairly good static and dynamic current mirroring in this topology to avoid thumps and audible artifacts.

exactly ...


I've got your "Reckless Pair" idea down to 9 transistors
similar to yours, it's exhibiting no static or dynamic level shifting at the output

I'm also playing with scale-able input gain - for gtr/bass headroom

---

note:

using op-amp buffers gives the signal path a much higher cmrr
(with psu ripple way less of a an issue) and there's less noise overall ...

the ota is just a standard four-device structure ...
achieving close-to-ideal one sided ota performance

---


btw, where's the half-wave side chain from ? ... I've seen it before

[PRR]

> how exactly this transistor pair filters the noise?

Emitter follower with an R-C filter to the base. Instead of raw supply crap you have filtered supply crap.

An elaboration: the Sziklai pair to allow smaller capacitor (or just because he uses the Sziklai everywhere else).

Also because any stray impedance in the filter will let Vdd bounce with the beat which may add thump. In this form the output impedance is very low.

[11-90-an]

okk... BB'd the PMC and it works nicely... 

(anyone has any chicken pickin riffs?  )

[niektb]

There are no stupid questions, only stupid answers 
9V is the supply rail that comes into the pedal from your PSU. This supply powers the side chain transistors TR12 and TR13 and their associated components. Then there is a filter / smoother comprising TR3 and TR4 which generates VDD.

Sorry to bother you Jonny, but how exactly this transistor pair filters the noise? I'm really curious to learn this!

In essence it's a capacitance multiplier! (The capacitance located at the base of the transistor)

[11-90-an]

Redrew the schematic for people (Like me) who get dizzy with all the intersecting-lines-that-aren't-connected-to each-other... 

[jonny.reckless]

Redrew the schematic for people (Like me) who get dizzy with all the intersecting-lines-that-aren't-connected-to each-other... 

Yes, that looks correct to me

[jonny.reckless]

okk... BB'd the PMC and it works nicely... is there a way to make it 'tighter'? 
(anyone has any chicken pickin riffs?  )

I am not sure what you mean by "tighter" sorry, can you explain?