Transformerless Vari Mu Compressor build thread

[richiyobs]

In my poor experience I often saw tantalum for timing net in old design... But in modern days low Esr or np caps can do the job right...
They where used for size capacitance ratio in that days and the duration over time...
I tried them also long the signal path, but distortion was my first discovery even if someone declared they have a wider freq. Response and fast charge.
Conflict minerals is one of the good reason to not use them nowadays
Best
Richi

 

[living sounds]

I don't know how significantly tantalum caps differ from a perfect capacitor so I can't say if it adds anything audible when using them as timing caps. Maybe tantalum caps will give that sweet sound of conflict minerals.

But C2 is a timing cap, right?

 

[living sounds]

In my poor experience I often saw tantalum for timing net in old design... But in modern days low Esr or np caps can do the job right...
They where used for size capacitance ratio in that days and the duration over time...
I tried them also long the signal path, but distortion was my first discovery even if someone declared they have a wider freq. Response and fast charge.
Conflict minerals is one of the good reason to not use them nowadays
Best
Richi

They have a sound in the signal path, Neve etc. But in timing circuits I tend to prefer the sound. I have used them in my 1176s for that reason, too. Conflict minerals are a problem, I have many caps lying around I can use though.

 

[Heikki]

But C2 is a timing cap, right?

Yes, C2 on sidechain is the timing cap.

 

[Heikki]

Since the DC/DC voltage converter for the B+ seems to be out of stock everywhere and has long lead time, here is a schematic and gerber files for alternative power supply board. Transformer with 180V secondary is needed for B+ and Mean Well RS-25-24 for heaters.

 

[living sounds]

About the NTCs in the PSU: Which values would be acceptable? Can I use a 33 ohm part (with sufficient wattage rating)?

 

[Heikki]

About the NTCs in the PSU: Which values would be acceptable? Can I use a 33 ohm part (with sufficient wattage rating)?

33 ohm NTC should be fine. It isn’t critical just as long there’s something to limit the start up current spike.

 

[manulaudic]

Hi, I'm about ready to finish the wiring on the rotary version I am building. I wanted to ask What would you recommend for grounding the main board. I see from the doc that the ground from led meter should go to chassis, has everyone gone that route? Are you having a star ground for all grounds of overboard including power supply ground?

I am asking cause on the main board Ground and 0V are connected but 0V and ground on the power supply aren't.

Manu

 

[Heikki]

I would recommend doing it just like in the layout pictures. That way the return current paths are like I intended. It shouldn’t matter at what ground/0V point is connected to the chassis. It works just the same even if you don’t connect it to the chassis.
https://ghr.fi/proaudio/varimut/layout/comlayoutrotary.pdfhttps://ghr.fi/proaudio/varimut/layout/comlayout2rotary.pdf
I could have have done better job at naming the gnd/0V pads on the boards. Now actually for example the pad marked gnd on power supply board is not at the same potential as gnd on the led meter board and it’s confusing that they are both marked gnd.

 

[manulaudic]

Ok, then I'll just follow as you intended ! amazing, almost there. Can't wait to power it up

 

[manulaudic]

It's probaply best to explain the modifications here too. The idea is that you have unity gain when the gain rotary is at middle position. From the middle position you can adjust the gain between -5.5 dB and +5.5 dB in 0.5 dB steps. The sidechain steps are changed to 1 dB steps and a switch is added to get to get 10 dB more range from the 12 step switch. Step 12 (fully clockwise) is always off position.

U9 and U10 are changed to INA2137. This is related getting the gain to unity when gain is at middle position. R43, R44, R45, R46 are changed to 330 ohm. RT1 and RT2 with the 1k resistors are added to make make and adjustable U pad for the outputs. These trimmers are used to adjust the gain to unity when the gain rotary is at middle position. This also makes it much easier to balance the gain between channels because there's no need to mess with the bias trims to balance the gains.

R39, R40, R41, R42 are changed to 1 ohm to make sure the +15V and -15V voltage doesn't sag to U7, U8, U9, U10, now under heavier load.

I do not recommend doing these changes for any planning to use compressor in tracking or perhaps in a mix bus. Potentiometer sidechain is more than adequate for plenty of uses. I do personally prefer this coarse stepped attenuator to a gain pot.

Edit: Updated version of the mastering modifications.
https://ghr.fi/proaudio/varimut/layout/stepped.pdf

I have a question regarding this. How would you adjust the bias trims to balance the gains still? Or how would you set them if the balance is done with rt1 and rt2?

Also as I didn't read this post until now I am short of 2 INA2137. I wired the whole board with the resistors for INA2134. I am also considering wiring it without this mastering mode... thoughts?

 

[Heikki]

I have a question regarding this. How would you adjust the bias trims to balance the gains still? Or how would you set them if the balance is done with rt1 and rt2?

Also as I didn't read this post until now I am short of 2 INA2137. I wired the whole board with the resistors for INA2134. I am also considering wiring it without this mastering mode... thoughts?

With RT1 and RT2 set RV1 on sidechain boards according to the instructions here Variable Mutual Contuctance Compressor then use RT1 and RT2 to adjust the channel gains.

If you have the 0.5 dB gain rotary switch then I definitely wouldn't change the resistors R43, R44, R45, R46, R39, R40, R41, R42 and add RT1 and RT2.

edit: I meant to say I definitely would change the resistors!

 

[manulaudic]

With RT1 and RT2 set RV1 on sidechain boards according to the instructions here Variable Mutual Contuctance Compressor then use RT1 and RT2 to adjust the channel gains.

If you have the 0.5 dB gain rotary switch then I wouldn’t definitely change the resistors R43, R44, R45, R46, R39, R40, R41, R42 and add RT1 and RT2.

I have the rotary gain switch yes. So maybe stick to original with ina2134?
would be easiest as i already have it set this way

EDIT: I have the 12 steps rotary gain switches. not the 0.5db gain

 

[Heikki]

I have the rotary gain switch yes. So maybe stick to original with ina2134?
would be easiest as i already have it set this way

Oops, I meant to say I definitely would change the resistors, add RT1 & RT2 and change U9 & U10 to INA2137.

The resistor changes and RT1 & RT2 give a much better way of adjusting the channel gains compared to adjusting RV1 on the sidechain boards. INA2134 can be used as U9 & U10 even if the mastering modification is done but even then the resistors should be changed and RT1 & RT2 added. With INA2134 the gain should be adjusted to unity when the 23 step rotary is at position 1. With INA2137 the gain should be adjusted to unity when the rotary is at position 12.

 

[richiyobs]

Thanks for clarifying Ina 2137 & 2134 differences

 

[manulaudic]

Oops, I meant to say I definitely would change the resistors, add RT1 & RT2 and change U9 & U10 to INA2137.

The resistor changes and RT1 & RT2 give a much better way of adjusting the channel gains compared to adjusting RV1 on the sidechain boards. INA2134 can be used as U9 & U10 even if the mastering modification is done but even then the resistors should be changed and RT1 & RT2 added. With INA2134 the gain should be adjusted to unity when the 23 step rotary is at position 1. With INA2137 the gain should be adjusted to unity when the rotary is at position 12.

Thanks ok I get it. I also wrote wrong. I am using these rotary gain : https://ghr.fi/proaudio/varimut/stepped.jpg 12 steps. 2db steps. Would you still recommend the mods on the output?
Manu

 

[Heikki]

With the 12 step more coarse stepped attenuator I would use INA2134 for U9 & U10. The rest of the output mods make it much easier to adjust the channel gain balance and adjust one of the rotary steps to exactly unity gain. Whether you should do them or not really depends if you need them.

 

[manulaudic]

With the 12 step more coarse stepped attenuator I would use INA2134 for U9 & U10. The rest of the output mods make it much easier to adjust the channel gain balance and adjust one of the rotary steps to exactly unity gain. Whether you should do them or not really depends if you need them.

Ok wonderful, should I still change R39, R40, R41, R42 to 1 ohm to make sure the +15V and -15V voltage doesn't sag to U7, U8, U9, U10, now under heavier load?
the 330ohm with trimmer and 1k will go to fine tune the output and unity gain thanks

 

[Heikki]

It the output trimmers RT1 & RT2 are added then R39, R40, R41, R42 thould be changed to 1 ohm

 

[paolomunarini]

I'm planning on the final steps for my build.
Would it be possible to add external side chain inputs switching the J23-J27 and J24-J28 pairs with the external signal?

And furthermore, would it be possible to add a Feedback/feedforward switch, extracting the SC signal before GR?

Paolo