Any interest in remote controlled preamps?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

nielsk

Well-known member
Joined
Nov 12, 2004
Messages
1,734
Location
Megapopulas, Florida
Is there anyone interested in this?
I am considering undertaking a remote controlled preamp design to work with the Yamaha AD8HD control protocol. This could work with the DM series consoles, and there are cheap software packages available that will controll it from a PC.
On the preamp end, it should not be too difficult to either controll a motorized pot, or a set of relays (or both).
I have not been able to find the protocol published anywhere as of yet, does anyone have experience with data acquisition from 422? This should be a fairly simple data stream, a device ID, a channel ID, a level and a few switch positions......
then I think a programmable device and some drive circuitry
I would be happy to hire someone to help with the digital end, and make the final board (s?) available to the forum
 
I thought of doing a complete digital control in the front but everything else analogue in the background, such as motorized potentiometers. The only reason of this is that I like robotics. ;D

Anyhow, the digital control side would be pretty straight forward with a microcontroller. I have been working with PICs but I have a very good engineer friend who does the serious programming for me.

In terms of controlling the potentiometers, the servo mechanism does not need to be over the top. Easiest is to use standard R/C servomotors that are normally used in radio controlled planes, boats etc. They can be as cheap as few dollars in quantity. Same here in the UK, only few quids. I get a good discount from the importer because I order in hundreds.

These servos take pwm positional command  and move the output shaft accordingly. Although they are normally designed to move a maximum of 90 degrees for RC applications, you can actually modify the feedback mechanism to extend the rotation to 270 degrees which matches the potentiometer's mechanical move. If the microcontroller operates in 8 bits, that gives you about 1.05 degrees resolution on the potentiometers. I do not think you will need more than that for audio control.

Of course there will be the issue of mechanical coupling of the servo to the potentiometer. Now, I do not have a problem with that. I have a good in-house machine shop and a master engineer who does all my milling and turning. But if you go down the route of using already motorized potentiometers like Alps then you can design your own R/C servo control. Apart from Hitec and Futaba most other manufacturers used to use Mitsubishi chip M51660L. I have designed loads of control with it and it is very easy to use. I have about 500 in stock.

Phase reversal etc. functions could be controlled with digital outputs as usual.
 
Thanks for the responses :)
I like the idea of a device that could be used with several different preamp designs. A servo motor could be easily shaft-coupled to any pot, and does sound less expensive than a motorized pot, I really like that.
I was thinking that a LM39** display driver could be used to fire relays at increasing voltage levels (& could probably fire the relays directly) for the preamps that use switched gain levels instead of pots.
Any suggestions about undertaking the decoding of the control protocol? I have  DM 1K....and was checking this out:
http://www.windmill.co.uk/serial.html
As I understand it, the  controled  device (servo motor?) does need to supply information to the controling device (microcontroller) to tell it that actually it is where it needs to be, but at the control position we are looking at audio levels.....
 
I can design the entire hardware no problem and get my friend to do the firmware but I would have to pay him and I can't do that until we have interest. Unless of course, somebody else here would be interested. One member is a dsp guy and this would be a piece of cake for him. It is a matter of having the time to write it.

I have already got a controller and the control software for windows that I supply with my robot arms. Although the GUI is designed to control a robot arm it would do this job comfortably. At least for development stage anyhow. After that a new GUI can be written according to the adopted command protocol.

Servomotor does not have to talk to the controller, that is complication. Good side of the R/C servo is that it operates open loop externally (which means there is no feedback between the servo and the controller) but closed loop internally. It takes the pulse coming from the controller, compares it with the  feedback potentiometer and reacts accordingly. Obviously you are at the mercy of the servomotor's internal feedback in terms of repeatability but these things are repetable. I have been using them on driving robot arms. For driving potentiometers they will last no problem.


 
I've been currently wondering if using the Analog ADG408 mux'es to switch between feedback divider ratios would work well in a remote controlled mic pre.

2 x 408s would provide 16 steps of gain and could be used to switch two amps in cascase like soem of the old British preamps or the JLM99V etc.

Dan K uses them to select caps and inductors in his EQ, how bad could they be for a mic pre?

-Tom
 
You can certainly use it, just another way of skinning a fish. However, the idea here is to have the gain control by a potentiometer. Yes, there is the cost of servo but hey, it's a diy. Something different.

I tell you the other advantage of mechanical control of potentiometer by a servomotor. You do not have to bust your gut trying to find a reverse log pot. Use a normal log pot and reverse drive it by the servo.
 
Yeah that is very true. Without trying to derail your efforts from the motorised pot thing, is there any thing untoward to watch out for with these MUXs when running analogue signals through them apart from on and off resistances and noise?

Motorised pots seem nice, ALPS one are pricey though.

-Tom
 
I have not passed audio signal in them before therefore I can not specifically say anything but those two are the major factors that you consider. I am sure bcarso, JR or PRR would make detail comments.

 
TomWaterman said:
Yeah that is very true. Without trying to derail your efforts from the motorised pot thing, is there any thing untoward to watch out for with these MUXs when running analogue signals through them apart from on and off resistances and noise?
-Tom
These analog switches are excellent. There is no noise issue with them except for charge injection from the control voltage, so you have to make sure that you switch signals of low enough impedance, but not so low as to induce distortion. Typically 5k to 20k guarantees pristine performance. Note: slowing the control lines with RC's will not help, since the charge injection happens from the internal logic line. Anyway I would not recommend using them for any application where there is continuous switching, like a fader or pan-pot, because of that (and the inevitable zipper-noise). You have to make sure that the signals won't be off limits, since it can turn on an actually off gate or vice-versa. I use them in particular to change the frequency range in state-variable biquads, nobody ever complained about it. Check out the MAXIM MAX313/314, amazing specs (less than 10 ohms Ron). I have ordered samples that I expect to play with in the next coming days.
 
Nielsk,

A pulse (square) width between 0.5mS to 2.5mS will move the servo to a maximum of 180 degrees. I have been using Hitec servos and this is the case, but for different makes it might be slightly smaller angle than that. However, this is not really a problem as 180 degrees or smaller mechanical move of the servo can be stepped up by using a proper gear ratio to 270 degrees to meet the potentiometers electrical/mechanical move. I have been modifying the feedback mechanisms for my own applications but this is not necesary here.

1.5mS centers the servo (127 in 8 bits). So any pulse between 1.5mS - 0.5mS moves the servo in one direction and 1.5ms - 2.5mS moves it in the opposite direction. Keep the off time between each pulse around/below 20mS.

 
I think you are on to something, here.  Consoles have rapidly devolved into stacks of black boxes.  I can't help but think that the future of the console is an open architecture of digitally controlled analog devices. This allows the knobs to be put back in front of the engineer, but still allows for the choice of diverse boutique eq, comps, and preamps.  A robotic knob might be a good first step.

I've always thought that mixing outside the box would be a lot better if the a/d devices operated at full volume, and their outputs were attenuated after the fact.  This is what summing boxes do, but they have no automation.  If a robotic attenuator could be controlled via midi or perhaps a hui protocol, you could automate in the box and mix outside the box.  Ideally, you could hijack the rubber band volume control for each track and have it control an attenuator.

One thought would be to get a head start by adding a real potentiometer into the moving fader control of a device like the Behringer B-Control Fader.  I dunno if there's room or if the motors would take the extra load, but a lot of work would have already been done, including software for the DAW.

I'm just thinkin'...

 
That is a great hack idea. Machanically coupling additional faders to the existing may not be possible in terms of space but if I can study what sort of feedback system  the existing faders have, then perhaps we could grab that feedback and control servo/rotary potentiometers that are placed somewhere in the case. So the rotary potentiometers would work in tandem with the faders.
 
Sahib, can you suggest any particular servos to research? There seems to be many on the market.
From what you describe, it would be easy to adjust a pot with a servo in real time, but how would this work when recalling a stored setting? It seems there has to be a position sensor signal to the MP to be able to store & recall settings from the control device, like the extra trace on a moving fader (I may be missing something here..)
 
Oh how I second this stuff. For a long time it has been a dream I've been working on, and it is quite a tricky one but not off the line. This was btw the last breath Neumann drew before they went broke and sold out to Sennheiser's pax romana - their entire consoles were designed as analog signal processing, digital control. Faders that run, EQs based on multiplexers...the entire wash. This was in the 80s. And some of those consoles are still running today in the big old houses of Germany, maybe sometimes it's lucky not all the places get the money to kick all the old stuff out these days.

Lord only knows what would have come if they had been able to continue, but maybe that's a page of history that we can help write.
 
Nielsk,

You will certainly need a memory to store all the commands and positional data. The system works as follows.

You have two interfaces. First one is the interface between the operator and the system. Second one is the interface between the first one and the servomotor.

All the control faders, rotary potentiometers, buttons, switches etc are connected to the operator interface which takes the instructions from the operator and sends commands to the servo controller interface accordingly. It also stores all the commands and positional data for future re-call. In reality you can do that with a single microcontroller to a degree but considering how many recording sessions one can and will do, this interface has to have quite a large memory. Instead of designing a custom hardware for this, using PC or industrial computer boards make it much more practical and cost effective. If we are thinking of a GUI it definitely makes sense.

All operator interface does is to send positional commands to the servo controller interface, say in 8 bits binary.

We said that the RC servo would move to a maximum of 180 degrees according to the positional pulse applied to it between 0.5mS and 2.5mS.

Imagine that you moved one of the control faders to its maximum position. The operator interface read it, stored its position and then sent binary 256 to the servomotor interface.

Servomotor interface (micro) took 256 binary, converted it into 2.5mS pulse output and sent it to the servomotor.

Servomotor received 2.5mS and moved to 180 degrees.

The potentiometer mechanically coupled to the servo also moved to its maximum position of 270 degrees (with the aid of step up drive gear ratio).

To get familiar with this sytem feel free to download my control software (in windows) from the below link. Between the two robot images where it says download software. When you launch the software it will give you an error message saying "the robot is not found". Click o.k. and continue. You can move the sliders, turn the outputs on etc. without actually being connected to the conttroller. You can even write programs and store them. Although there is the help menu if you pm me your e-mail address I can also e-mail you the user and programming manual.

http://www.robotica.co.uk/robotica/ramc/products/robotic_arms/edubot250.htm

You can use any RC servo you wish. You do not need to buy expensive stuff like Futaba. Having said that even their standard servo is very cheap. Check Tower Hobbies since you are in US.

If you wish to experiment I would happily supply a bear board with BOM and a programmed chip at cost. You can buy the rest of the components and populate the board. Stick the micro off you go. No special calibration required.

This offer applies to entire forum.
 

Latest posts

Back
Top