Reckless compressor - an original feed-forward diode based compressor

[jonny.reckless]

I've been experimenting with using diode bridges in a compressor for a while. It's a very old school approach, but can be made to work really well, and it has a warm character to the tone which I really like. This is an original design, with a precision full wave rectifier feed-forward side-chain.

It's quite warm and characterful, but also can be set to be very transparent. It fits into a 1590BB.

Here is the circuit:


Here it is in action:

[GGBB]

Is the dual-rail design necessary? In other words could you do this without the 1044 charge pump inverter as a single rail +9V design?

[pinkjimiphoton]

that sounds great, jonny, really clear and a nice squish to it even with the video encoding.
if you end up getting boards burned, please let me know, i'd love to build one up to add to the arsenal.
i usually am not too crazy about compressors, but this one sounds really good, mate, well done!

[noisette]

Great post, thanks!
I have a lot studied diode bridge compressors/vcas but never finalized a circuit.
I like the sound of diode compression, I think it´s very distinct and well suited for electric guitar.

Is the dual-rail design necessary? In other words could you do this without the 1044 charge pump inverter as a single rail +9V design?

Is that a rhetorical/didactical question?
It could be (has been?) done, and maybe I´ll try it, but always better finished than perfect, and telling from the vid, it sounds
already quite tight as it is!

EDIT you could rig it up with dual 9V batteries also, instead of the 1044 circuit...

[GGBB]

Is that a rhetorical/didactical question?

Not at all. The circuit would be simpler, cheaper, easier to build, and potentially have a smaller footprint without the charge pump (I won't use large boxes for pedalboard space reasons). Just wondering why the choice of the +9V/-7V dual rail design - does the circuit require that for some reason?

[noisette]

Ah, ok. 
Well, it could surely be done with one 9V battery split to dual +4.5V/-4.5V.
Single supply is possible but unnecessarily complicated (for me anyway).

Examples of simple diode vcas are the KORG MS50 vca (and vcf).
http://machines.hyperreal.org/manufacturers/Korg/MS-synths/schematics/ms50filt.gif

Examples of single supply diode compressors (HQ broadcast) check Siemens U273/U274
on this site:
http://audio.kubarth.com/rundfunk/index.cgi#u

[GGBB]

Ah, ok. 
Well, it could surely be done with one 9V battery split to dual +4.5V/-4.5V.
Single supply is possible but unnecessarily complicated (for me anyway).

Thanks. I was actually hoping for an answer from the OP since it's his design. The possibility of doing it obvious - I'd like to know if the designer thinks the circuit might function less well that way, and generally why he went with the extra complexity of the charge pump.

[jonny.reckless]

Yes it's possible to use a single supply and a 4.5V virtual earth. I generally prefer not to. Symmetric circuits are easier to design. The extra headroom of split supplies helps. Mostly I can flood the top layer of the PCB with ground plane. In this case signal ground is real ground. Generally people use an 8 pin op amp for the mid rail so the component count and area are about the same.

[GGBB]

Generally people use an 8 pin op amp for the mid rail so the component count and area are about the same.

Thanks. I guess by "people" you mean pro builders. Most DIY designs I come across don''t bother with that unless there happens to be a spare opamp anyway (an unused channel).

About headroom - are you boosting the circuit enough where this (split rail vs. single rail) makes a difference? Usually guitar signals are far below even what a +9V single rail opamp setup would provide as far as headroom is concerned. It doesn't look like your circuit is boosting a lot but I don't know it well enough.

[jonny.reckless]

I think it would work ok if you wanted to modify the design for single rail operation. The issue I've seen in the past with this approach is that the virtual ground you get at 4.5V is never quite low impedance enough for transients, so it tends to bounce with respect to real ground. Depending on the circuit design this may or may not be a problem. In this circuit there aren't any fast edge transients into ground so I guess you'd get away with it. My first attempt at building the engineer's thumb on vero had a tendency to motorboat due to ground bounce. I got a roughly 1Hz thumping sound in the output until I really carefully strapped everything down with thick short wires and lots of decoupling caps.

I laid out a PCB for this, as seen in the video, which is a 2 layer design with ground plane on the top copper layer. It fits into a 1590BB. Let me know if you're interested in one.

[rankot]

This is really interesting Jonny, going to my "to do list"

[Eb7+9]

nice piece of design work Jonny ...

been looking for a solid diode based limiter design for a while now
congrats and thank-you

[pinkjimiphoton]

got my board today, jonny!! thanks mate!! will let ya know how it goes when i get a chance to get to it!

[domestic-bliss]

is this a follow on from the et redux? or a totaly different unit? i realy want to make a et redux :p you guys inspired me lol

[pinkjimiphoton]

hey jonny, total abject moron here obviously, but what the hell is sk3 and sk 4 for? ribbon wire? some kinda socket?
forgive me, but slightly confused. never seen this nomenclature before. thanks man!

[jonny.reckless]

hey jonny, total abject moron here obviously, but what the hell is sk3 and sk 4 for? ribbon wire? some kinda socket?
forgive me, but slightly confused. never seen this nomenclature before. thanks man!

Take a look at the video. They are for a standard 5x2 way 0.1" box header and 10 way IDC 0.05" pitch ribbon cable to the switch and sockets. You can pick up the parts or ready made cables on eBay.
https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.com%2Fulk%2Fitm%2F282084089830

You can wire it up with anything you like, I just use these because they're pretty cheap and convenient if you want to take the thing apart a lot.

[jonny.reckless]

is this a follow on from the et redux? or a totaly different unit? i realy want to make a et redux :p you guys inspired me lol

This is a new original design. I also did the redux mods to the ET on a different thread.

[jonny.reckless]

got my board today, jonny!! thanks mate!! will let ya know how it goes when i get a chance to get to it!

Did you get it built yet? Anything I can do to help?

[pinkjimiphoton]

not yet, gig schedule has been pretty brutal lately. i will check in when i can bro!

[R.G.]

The diode modulator was a common thing in older RF design. Thomas Organ used a diode bridge modulator in the USA Vox amplifiers. It was a good tremo, very deep. The depth came from buffering the dry signal, inverting the modulated signal, then mixing to get cancellation. That let it go all the way to zero in some part of the sweep where the amplitude cancelled.

The bane of all diode modulators is that you have to work hard at keeping common mode signal out of the signal path, and that it's only a low distortion modulator with signals lower than about 25 to 50mV. Beyond that, the signal voltage swing starts changing the diode's equivalent resistance itself.

You don't have to use a bipolar supply, but it helps.

@ J.R.: since you're driving with a current, you might want to arrange to drive the top of the diode modulator with a voltage equal but opposite to the voltage at the bottom of the bridge. It will take some of the load off the differential amplifier to keep control signal feedthrough. Yeah, I know, diffamps are good, but no sense making the work harder than they have to.