THAT "4315 One Knob Squeezer"

[stagekraft]

Hey Ya'll

While I've built some compressors in the past, they were all FET or OPTO, now I'm trying a VCA based design.

I'm currently working a a compressor project for one of my clients based on the THAT "4315 One Knob Squeezer".

As I'm laying this out I'm curious about the Control Voltage from the RMS to the VCA.
The 4315 datasheet shows that in a simple compressor design the RMS CV (pin 5) out goes to the EC- of the VCA (pin 12).
The official THAT One Knob Squeezer schematic is a bit more elaborate, but accomplishes the same thing.

My question is; what is the typical voltage polarity/range of this CV?
I'm working out a compression led bargraph and want to know what kind of signal I'm working with?

My layout is with a 15V (wal-wart) power supply instead of a 9V battery as shown in the schematic..

I've already bought a couple 4315's through Mouser, but being a QSOP package, it's not easy to breadboard.
The THAT website shows a QSOP to DIP adaptor available (4315QDA16), but Mouser doesn't carry it,
I did find a QSOP to DIP at Digikey but have not ordered any yet.

Has anyone here built this circuit?
Can anyone share any working details about the CV, or performance of this circuit in general?

Regards,
JohnR

 

[Bo Deadly]

Hey Ya'll

While I've built some compressors in the past, they were all FET or OPTO, now I'm trying a VCA based design.

I'm currently working a a compressor project for one of my clients based on the THAT "4315 One Knob Squeezer".

As I'm laying this out I'm curious about the Control Voltage from the RMS to the VCA.
The 4315 datasheet shows that in a simple compressor design the RMS CV (pin 5) out goes to the EC- of the VCA (pin 12).
The official THAT One Knob Squeezer schematic is a bit more elaborate, but accomplishes the same thing.

My question is; what is the typical voltage polarity/range of this CV?

See the datasheet. It indicates mV/dB. Ec+ is positive control where Ec- is inverted.

I'm working out a compression led bargraph and want to know what kind of signal I'm working with?


My layout is with a 15V (wal-wart) power supply instead of a 9V battery as shown in the schematic..

I've already bought a couple 4315's through Mouser, but being a QSOP package, it's not easy to breadboard.
The THAT website shows a QSOP to DIP adaptor available (4315QDA16), but Mouser doesn't carry it,
I did find a QSOP to DIP at Digikey but have not ordered any yet.

A good place for "surfboards" is Ebay. Although they usually come from some far off place that takes a few weeks to get.

Has anyone here built this circuit?
Can anyone share any working details about the CV, or performance of this circuit in general?

If you want the best possible performance, you want the THAT 2181. It uses a special fabrication process that allows for creating the closely matched NPN and PNP necessary for low distortion. THD of 4315 is not that great by comparison. The 4315 is designed as an all-in-on "engine" so it's good for a "one knob squeezer".

Note that noise on the control ports will translate directly to noise on the output. So layout is critical. It is difficult to achieve good VCA performance using a breadboard with jumper wires. It will be necessary to make a very carefully laid out PCB to achieve performance close to what's in the datasheets.

Regards,
JohnR

 

[stagekraft]

See the datasheet. It indicates mV/dB. Ec+ is positive control where Ec- is inverted.

And that's the part I'm trying to figure out...

Is that CV "inverted" (as in starts at VCC and pulls down)
or is it "negative" (to the VRef derived from the voltage divider between VCC and GND on pin 7?

A good place for "surfboards" is Ebay. Although they usually come from some far off place that takes a few weeks to get.

Bargraph is all laid out discrete, just trying to figure out what kind of signal I'm driving it with...

 

[Bo Deadly]

And that's the part I'm trying to figure out...

Is that CV "inverted" (as in starts at VCC and pulls down)
or is it "negative" (to the VRef derived from the voltage divider between VCC and GND on pin 7?

The control port voltage is relative to Vref. So if the voltages are equal, gain is 1. If you put +36mV on the Ec- port, you should see -6dB attenuation (gain = 0.5). So the Ec- port is usually used for compression whereas Ec+ is usually used for expansion. But that is not concrete because your side chain output might be inverted or the ports could be driven differentially. But for a "one knob squeezer" just drive Ec- relative to Vref. If you're not sure how it works, I would recommend sticking to the typical application schematics. It will take a little experimentation and fiddling to see it all work correctly (especially if you're trying to use a breadboard instead of a purpose built PCB).

 

[Ricardus]

I'd be interested in this project.

 

[Tubetec]

A knob squeezer sounds slightly obscene .

 

[Script]

For inspiration only, the Dynamaxx is based on the 'one-knob squeezer' design note, plus differential drive, upward compression etc

https://www.soundonsound.com/reviews/spl-dynamaxx

 

[stagekraft]

ok, so I'm putting a positive voltage (relative to Vref) to the EC- port to get compression.

that makes driving the display easier

thanks,
JohnR

 

[noguffay]

Hey Ya'll

While I've built some compressors in the past, they were all FET or OPTO, now I'm trying a VCA based design.

I'm currently working a a compressor project for one of my clients based on the THAT "4315 One Knob Squeezer".

As I'm laying this out I'm curious about the Control Voltage from the RMS to the VCA.
The 4315 datasheet shows that in a simple compressor design the RMS CV (pin 5) out goes to the EC- of the VCA (pin 12).
The official THAT One Knob Squeezer schematic is a bit more elaborate, but accomplishes the same thing.

My question is; what is the typical voltage polarity/range of this CV?
I'm working out a compression led bargraph and want to know what kind of signal I'm working with?

My layout is with a 15V (wal-wart) power supply instead of a 9V battery as shown in the schematic..

I've already bought a couple 4315's through Mouser, but being a QSOP package, it's not easy to breadboard.
The THAT website shows a QSOP to DIP adaptor available (4315QDA16), but Mouser doesn't carry it,
I did find a QSOP to DIP at Digikey but have not ordered any yet.

Has anyone here built this circuit?
Can anyone share any working details about the CV, or performance of this circuit in general?

Regards,
JohnR

I'm kind of doing the same but with the 4320. I found these adapters on Mouser that should work for the QSOP package.
https://www.mouser.ca/ProductDetail/535-LCQT-TSSOP28
There you go, that should do it, but check it over yourself to make sure.
https://www.mouser.ca/ProductDetail/535-LCQT-TSSOP16Confirmed by THAT, Will work.

"Has anyone here built this circuit?
Can anyone share any working details about the CV, or performance of this circuit in general?"
Have you checked out all THAT's Design Notes? They are chock full of detailed mathematical analysis and information on the behaviour of the 43 series. Check their data sheets also.
http://www.thatcorp.com/Datasheets.shtmlI've hobbled together the circuitry of the 4320 with the Manual mode Variable Attack/Release, from the 4315 notes, and added the Comp LED Indicator
Just a bunch of cut and pasting, but THAT verified it'll work as is.

I also own a Keeley GC-2 which is based on the 4320. I'm pretty sure he used non LED diodes for the "Auto" compression attack/release circuitry. There are no Green LEDs on the pcb other than the Indicator bipolar LED (yellow/Amber) I have not traced the GC-2 to confirm its similarities/dissimilarities to the THAT 4320 design notes. But it's probably approx, 99% the same.

https://www.screencast.com/t/d2wduvBO

 

[noguffay]

See the datasheet. It indicates mV/dB. Ec+ is positive control where Ec- is inverted.


A good place for "surfboards" is Ebay. Although they usually come from some far off place that takes a few weeks to get.


If you want the best possible performance, you want the THAT 2181. It uses a special fabrication process that allows for creating the closely matched NPN and PNP necessary for low distortion. THD of 4315 is not that great by comparison. The 4315 is designed as an all-in-on "engine" so it's good for a "one knob squeezer".

Note that noise on the control ports will translate directly to noise on the output. So layout is critical. It is difficult to achieve good VCA performance using a breadboard with jumper wires. It will be necessary to make a very carefully laid out PCB to achieve performance close to what's in the datasheets.

I'm not sure what he's making, (though he did say it's the One Knob Squeezer, that ain't studio gear circuitry) but do you think it would really matter if it was a Guitar effects pedal? 2181@ 0.0025% THD vs 4315@ 0.15% THD? For the added amount of extra pcb layout in a guitar effects pedal to include the 2181, it is also not worthy of such lowered THD. If it were studio rack gear, sure. Though I would have went with the 4320.
"Note that noise on the control ports will translate directly to noise on the output"
ie use good pots and don't move them around while playing lol. Use good everything as recommended by THAT, for componentry.

 

[stagekraft]

I'm not sure what he's making, (though he did say it's the One Knob Squeezer, that ain't studio gear circuitry) but do you think it would really matter if it was a Guitar effects pedal? 2181@ 0.0025% THD vs 4315@ 0.15% THD?

Guitar Pedal- No..., Studio Grade - again No...
We're talking Live Sound...

My client is a drummer, who plays with lots of different groups, he wants a "quick and dirty plug into the mic insert compressor" to fatten his sound and tame his levels. I don't think the THD difference is really gonna matter here.

Note that noise on the control ports will translate directly to noise on the output. So layout is critical.

My signal path (Input amp to VCA to Output amp) is as short as I can get it on the layout.
The CV related parts are very close also.

The rest of the layout is display related, although this is almost as complicated as the compressor.
I have a switched bargraph display using the tried and true 3915, using bi-color LEDS.
In one switch position it shows the output level in normal bar mode, bottom up, 5-green, 3-yellow, 2-red.
In the other position I switch the 3915 to dot mode, with an inverted signal that holds the 3915 at full scale for no compression,
then steps top down 2-green, 3- yellow, 5-red, to show relative compression.

 

[noguffay]

Guitar Pedal- No..., Studio Grade - again No...
We're talking Live Sound...

My client is a drummer, who plays with lots of different groups, he wants a "quick and dirty plug into the mic insert compressor" to fatten his sound and tame his levels. I don't think the THD difference is really gonna matter here.


My signal path (Input amp to VCA to Output amp) is as short as I can get it on the layout.
The CV related parts are very close also.

The rest of the layout is display related, although this is almost as complicated as the compressor.
I have a switched bargraph display using the tried and true 3915, using bi-color LEDS.
In one switch position it shows the output level in normal bar mode, bottom up, 5-green, 3-yellow, 2-red.
In the other position I switch the 3915 to dot mode, with an inverted signal that holds the 3915 at full scale for no compression,
then steps top down 2-green, 3- yellow, 5-red, to show relative compression.

I'm trying to work with the THAT DN #112, http://www.thatcorp.com/datashts/dn112.pdf , using their design of the LED bar graph, but I want to increase it to a 10 segment led and have it register 50db of compression. Would you be willing to share your bar graph design schematic? I'm using the 4320, the manual comp attack/release circuitry, and this led bar graph. I'm a total nube to design and really I'm throwing darts blind folded, but learning as I go. THAT has been super helpful and quite friendly with design help.
This is a work in progress and I need help: 05.26.2022-12.23.06

Ah I see: https://www.mouser.com/datasheet/2/405/lm3915-443929.pdf That's a good idea. Didn't even know this exists

 

[noguffay]

My signal path (Input amp to VCA to Output amp) is as short as I can get it on the layout.
The CV related parts are very close also.

That's another good reason to use the 4320 all in one Audio Engine. The THD is also a slightly better level than the 4315

 

[iprovlek]

Hi Noguffay!

This led bar graph must be connected like in DN112 datasheet,so to ratio wiper and ratio ground....

I have it in my THAT4301 compressor, project from this forum, and works great....

 

[iprovlek]

https://groupdiy.com/threads/that4301-compressor.52021/

 

[noguffay]

Hi Noguffay!

This led bar graph must be connected like in DN112 datasheet,so to ratio wiper and ratio ground....

I have it in my THAT4301 compressor, project from this forum, and works great....

Thank you so much for the assurance, I needed that. Yeah I was comparing the DN #112 and the R00 4320 Pedal Comp that I've modded to include the manual attack/release circuitry. I was thinking, yeah, it must go after the VR2 05.26.2022-12.23.06
But I'm pretty sure I'll have to use pin 3 of the 4558 as the 4320 has a positive dc voltage going from VR2 wiper to QA2 (4320). So flipping the 4558 pin configuration and have pin two go to Vref from U1E(4320).

How come you do not include the manual attack/release circuitry in this layout? as an optional add-on? Sorry I assumed you were the creator/seller of the pcb kit.

The THAT combined design I'm looking to DIY is based on the 4320, like the Keeley Comp Pro that also has the manual attack/release. I also want to add the bar graph. I'm going to also look at the alternative use of the LM3915 chip for this, as Stagekraft has mentioned here.

 

[iprovlek]

No problem Noguffay...

How come you do not include the manual attack/release circuitry in this layout? as an optional add-on? Sorry I assumed you were the creator/seller of the pcb kit.

Im not creator or seller of this pcb, Diy.tubes is....

For attack and release mod you must use DN115 this works great too...

For bar graph GR you can use DN112 or try something else, but sure that DN112 will work great!

I use even 1mA meter which can be solder direct on position where LED bar graph goes, and work too great!

 

[iprovlek]

Here is example of working DN112 in Pumpernickel guitar pedal which are based on THAT4301...

 

[iprovlek]

Pumpernickel is pedal kit from schalltechnik04 from Germany!