Relay Based Gain for Remote Mic Preamp

[noizetoys]

Forgive me if this sort of thing has already been asked and answer but searching the forums didn't turn up anything.


I need some help understanding what might be involved in building a remote mic pre.

Here's my thought:

I would like to build (16) remotely controlled API-style mic press. Why API? Mainly because the shunt-to-ground approach to gain. I would think that using relays to control which resistor is in the path to ground would be doable and not an audio nightmare (but I could be wrong).

A brief description:

• Each of the (16) gain rotary switches (6-12 positions possibly, not sure on the number gain steps) is connector to a micro controller (MCU).
• Each of the (16) MCU's connects to a single Small Board Computer (SBC), a Raspberry Pi or maybe something simpler like an Arduino.
• Over Ethernet or WiFi, the 'Control' SBC speaks to the SBC in the Preamp Rack.
• The SBC in the Preamp Rack talks to the individual MCU in each mic pre.
• The MCU in the mic pre controls the gain step relays.

Or more simply put:
Gain Switch -> MCU -> SBC ~(Ethernet or WiFi) ~> SBC -> MCU -> Relays

I was thinking of using something like a ATtiny or PIC12/16 with four or more ADCs to read the resistor value of the switch (using a voltage divider) and the controller ID (0-7).
The gain switch would have a resistor value for each position into a voltage divider that could be read by a 10bit ADC to give the value. The ID would require (3) ADCs to allow using (3) jumpers to set the ID, simple Binary value. Eight channels per SBC was a thought, I could bump it up to four jumpers and four ADCs for a total of (16) ID's.

The SBCs would talk to the SBC in the rack housing the mic pre's. Then it would sort of go in reverse with the SBC talking to the individual MCUs in each pre. The MCU would probably talk to a DeMux to trigger the appropriate relay. Or I could use another device to handle the relays. Again, not sure what the best approach would be.

Here are my questions:

• What would be the best (or at least good) approach to getting the gain setting from each switch? I know I could use a voltage divider and read the voltage using a ADC pin or I could connect each switch position to a pin on the micro controller or De-Mux the connections. Would it be better to use transistors for the gain steps on the control side? Looks like there are lots of options and I don't know the pro's and con's of each.
• Do I need an MCU for each gain switch? Would it be possible to use just a single SBC instead of having a MCU for each gain switch? If so, which one?
• What method could I use to have the MCUs talk to the SBC and SBC to MCUs? I2C? PWM? USART? Morse Code? Smoke Signals?
• What would be the most reliable protocol for sending messages between the SBC's? HTTP/TCP/IP? I2C? Yelling Loudly?

I've worked with SBCs to make simple things, but never something this involved. I've never used external MCUs to talk to an SBC so am unfamiliar with how that works and the protocols that make that happen. I can handle the programming, it's what I do for a living.

The big problem is that I don't know what I don't know (and barely know what I do know!).

Thoughts?

Thanks,
James

 

[JohnRoberts]

Yes it has been done but it is expensive... search harder.
===
There are modern preamp chips with digital gain control, much easier to accomplish.

JR

 

[wlinart]

So many controllers, I think this could be done easier.
The inputs, connect them all to 1 raspberry pi (or similar) via something like this: https://thepihut.com/products/16-channel-analog-input-hat-adc-for-raspberry-pi That saves you the MCU's at your input.
6 steps seems like too little for a gain switch, and just connect them like a potentiometer and read them in via 1 analog input.
The connection between it all, i don't have a clue, that you'd have to ask at a programmer.
Talking about the controlling of the preamps, i'd take a look at what is done here with a neve preamp, simple approach, with only 7 relays 20 gain steps. If you want less gain steps, also less relays are needed.

 

[gyraf]

api are probably too mushy, have too much "sound" to wish for 16 of

perhaps focusrite ISA would make more sense, iicr that one is a similar gain-setting scheme

 

[john12ax7]

I think you should better define the needs. What character style pre do you want? What gain range will you need? What gain step size do you need?

There are other options that use a single gain setting resistor. Could be something simple like the instrumentation style chips, that1510, ina103, etc. Or go api if that's the sound you want. But I wouldn't pick it just for the resistor.

 

[wlinart]

If wanting a clean style preamp in the style of that1510, better use a pga2500, that one already includes the circuitry to control it digitally, and is used in many professional digital consoles. But of course, that's quite the opposite from using API or neve.

 

[noizetoys]

Thanks for the responses.

The reasons I'm going with the API-style circuit is that it's simple, easy to build, and I have built units I can compare to. Plus I have all the expensive parts (transformers and DOAs). And the schematics I have make it easy to understand and know where to insert relays for switches. I can also take one of my units and wire in the bits to see how it works. Did I mention that I sort of lazy?

I have an 8 channel, 4 bus (one 's') mixer I build that I want to add preamps to and expand out to 16 channels. Maybe I should have mentioned that earlier...

Clean is good, clean is fine but I like distortion and heavy things so it's API this go around. Although I might look into the THAT1510, sound interesting. I can always solder more steel to it so it's heavier!

I'm going to start with an Arduino UNO R3 to talk to the controls. I already have it and the ATMEGA328P is cheap (relatively). I'll put a few together and try to get that talking to the Raspberry Pi I also have (trying to do this on the cheap, well, just not expensive is probably a better way to put it).

If I can get it working it might be possible to modularize it so that it could be used with other pres that use multiple resistors in their gain stage (e.g. Neve, Sphere, etc.)

Thanks,
James

 

[wlinart]

Thanks for the responses.

The reasons I'm going with the API-style circuit is that it's simple, easy to build, and I have built units I can compare to. Plus I have all the expensive parts (transformers and DOAs). And the schematics I have make it easy to understand and know where to insert relays for switches. I can also take one of my units and wire in the bits to see how it works. Did I mention that I sort of lazy?

I have an 8 channel, 4 bus (one 's') mixer I build that I want to add preamps to and expand out to 16 channels. Maybe I should have mentioned that earlier...

Clean is good, clean is fine but I like distortion and heavy things so it's API this go around. Although I might look into the THAT1510, sound interesting. I can always solder more steel to it so it's heavier!

I'm going to start with an Arduino UNO R3 to talk to the controls. I already have it and the ATMEGA328P is cheap (relatively). I'll put a few together and try to get that talking to the Raspberry Pi I also have (trying to do this on the cheap, well, just not expensive is probably a better way to put it).

If I can get it working it might be possible to modularize it so that it could be used with other pres that use multiple resistors in their gain stage (e.g. Neve, Sphere, etc.)

Thanks,
James

That's probably a good idea, hook up an arduino to one of those cheap relay board with 8 relays and connect that to a preamp. Try to make that work. Then connect a raspberry pi to the arduino and try to make that work. If that works, the rest is probably trivial.

 

[john12ax7]

Digitally controlling a resistance has lots of applications aside from just a mic pre. Keep us posted, am interested to see how the digital side of things works out.

 

[noizetoys]

@john12ax7 Keep you posted? I was hoping this was like ChatGPT and all I had to do was describe the problem! Damn, guess that's not the case....

Again, thanks for the replies.

I'll update as I go along. Hopefully sooner than later! Guess I should order those rotary encoders...

Thanks,
James

 

[JohnRoberts]

I was hoping this was like ChatGPT and all I had to do was describe the problem! Damn, guess that's not the case....

It is a little like chatGPT where you could get incorrect answers,,,,,

JR

 

[john12ax7]

@john12ax7 Keep you posted? I was hoping this was like ChatGPT and all I had to do was describe the problem!

I'm more on the analog side of things, so can help point the direction for that. But for digital and MCUs, chances are you will come up with something much better on your own than I would suggest.

 

[Brian Roth]

@noizetoys Just curious. Is this intended for live (as in a sound reinforcement or broadcasting)? If so, you really need to think about failure modes. Cable to the stage gets disconnected? Power blip/outage? etc

Bri

Edit/PS. With an API-style preamp, you need to also consider the need for a switchable pad ahead of the input transformer as one of the "gain setting" options. Like those times when a LDC mic is close to a snare or kick drum! <g>

 

[Brian Roth]

Further musings that are probably not entirely useful...I'm good at that! lol

MANY years ago I pondered this entire concept, well before "modern" digital control was feasible. In my case, it was to put the preamps on the stage for a PA system and run line level back to the FOH mixer....which could be hundreds of feet away. Shrug...a pipe dream back then!

In more recent times....I became pretty familiar with a Neve Capricorn desk in a recording environment. Obviously that (now antique) desk used an analog front end for things like mics and analog tape machine I/O.

Olde analog desks had an inherent memory of the "last used" panel settings.....mechanical switches. If you walked away from an in-process project (or the mains power dropped) you didn't lose the most recent settings.

With the Capricorn, I had to always remember to save the settings for everything. No biggie....click through some menus and 100% recall IF did it regularly and at the end of a day's work when I was leaving the studio for the day (or lunch...lol). Pissed me off a few times when the mains dropped and the entire thing came up into the "last saved" settings which were far behind where I was at. That desk had a timed auto-save but it wasn't always easy to manuever . With the timed auto-save, there would be hundreds of snapshots saved. So....click through the bazillion saved files menu to get back to the last settings.

Anyway....just rambling now.

Bri

 

[noizetoys]

@Brian Roth Several years ago I had an incredible opportunity to help design and program (as in what happens where with what not the code stuff) an almost $40 million dollar performing arts center here in Kennesaw Ga (it’s now called the Murray Arts Center at Mt Paran School, long story). Any who, in the theatre we decided to go all out and installed an SSL C200 (as a side note, that decision is what finally pushed SSL into the live world. We were told this by a former VP at SSL years later). Anyway, the theater was designed by John Storyk (another long story) and feature a 5.1 Nexo GeoT system. The only analog parts were from the mics to the pres and speakers to amps. Those preamps? 64 channels of Aphex 1788’s. The cable runs from the mic to the pre were routinely 30’ in total with the max at less than 80’. The fun part was that the digital signal was converted to MADI and sent to 3 separate studios where it could be recorded straight into Pro Tools without ever leaving digital. Those recording were better than any live recording you have ever heard. Just ridiculous.

I had the pleasure of doing a live mix of Spyro Gyra, Al Di Meola, and a number of other acts on either the 80 input SSL 9K or the other SSL C200 in the B room. I must admit it was pretty cool to have Al Di Meola sitting behind me listening back to the mix on the board with no EQ or other processing and the only thing he asks is “can you make the snare a little brighter?”

That’s when you really know that remote pres are the way to go.

I could bore you for hours going over the tech specs of that place. It’s always fun to spend a lot of someone else’s money to build something the way you know it should be built. I was very honored to have been able to do that.

Alright, blah blah blah. Go away.

 

[Brian Roth]

Excellent! So what the fuc!k is your question? Just curious

 

[noizetoys]

Well, when you put it like that I guess I don't have a question.

I'll just see myself out....