DIY fader automation interest ?

[zamproject]

Hi all

I decide to start this tread as my 89MotioN system is at crucial steep.
It's an automated fader system, full analogue PID motor driver allowing compatibilities and noiseless operation with passive audio fader and low level audio path.
The v1 boards (17pcb for 8 fader pack + SMPS, core and line tx/rx)  work fine, actually interfaced with midibox hardware and MB_NG software using MCU protocol. It still need some correction and crash test
I'm on this since two years, I believe I take one of the best route possible, and the result is good, I can leave and work with the actual implementation/design. But the next steep a "release", to be short, will implies little more work.

The main point here is to take temperature and see potential interest around for a system like this:

here is non exhaustive list of actual spec

-8 analogue PID and analogue touch detection for 8 motor fader (TKD and ALPS tested)

-64 digital in and out for led and button as follow:
  8 din interfaced to touch detection
  8 dout used do drive optocoupler, to allow analogue side remote without digital noise... in my case FET mute
  8 dout used to bypass motor
  6 din/out (per fader) illuminated button for remote, solo/mute/auto-read/auto-write/user/select
  8 spare din (button)

-8 AD and DA converter to read (from servo track) and send (to motor driver) automation data, 10bit<=>1024<=>0.1mm  steps for 100mm fader

-Dynamic spec give around 100ms full scale travel.
-Static spec give 0.01dB repositioning accuracy, by comparing (at step position) 2 audio re-recording pass,  phase inverted of the same automation read. Impressive but it's done in a short time range, same temperature etc...

-core STM32F4
-Vero SMPS (you need a high quality supply, with sens)
-Line TX/RX for computer/psu 19" rack up to 10m

Estimated cost is in the 3k (+/-1k) range, depending of fader choices and mechanical integration, for 8 fader pack.
It's not at all a cheap DIY project, nor an easy on, it involves digital and analogue electronic and some programming knowledge.
But it could be the first OEM DIY moving fader system so...

Let me ear what you guys around think about this.

Best
Zam


------------------------------------------
edit 30/01/16

for HUI protocol and PT user,
group member Zander from Belgium start a discussion about it, have a look.
http://groupdiy.com/index.php?topic=61673.0

 

[miszt]

I would really like to be able to record and set the position of Pots via MIDI in my hardware, I guess the principal is the same? making a note on paper does the job, upto a certain point

I did ask a question about it on another forum, but for some reason got a whole host of useless answers that somehow turned into a flame war lol


any particular reason you are using MCU, and not something with better resolution, Eucon for eg? that's the one thing that has always held me back in terms of midi controllers, I always switch back to using a mouse/touchscreen, because the resolution just isn't upto scratch

 

[zamproject]

Hi Miszt
It's not a snapshoot automation, it's a moving fader with automation read/write sync in the DAW,  audio pass through fader.
MCU is the easiest protocol for now, it's what opensource MB_NG have. Eucon is proprietary so no way to use it, OSC may be possible
It work now at 10 bits 1024steps, 0.1mm, with PB data (14bit) I don't see any reason to have better resolution.
I have it tested at 11bit resolution 2048 step 0.05mm for 100mm fader, my hand moving introduce more jitter than the system...
12bit (max ADDA resolution) introduce jitter
Best
Zam

 

[miszt]

Hi Miszt
It's not a snapshoot automation, it's a moving fader with automation read/write sync in the DAW,  audio pass through fader.
MCU is the easiest protocol for now, it's what opensource MB_NG have. Eucon is proprietary so no way to use it, OSC may be possible
It work now at 10 bits 1024steps, 0.1mm, with PB data (14bit) I don't see any reason to have better resolution.
I have it tested at 11bit resolution 2048 step 0.05mm for 100mm fader, my hand moving introduce more jitter than the system...
12bit (max ADDA resolution) introduce jitter
Best
Zam

Yes that is what I meant, automation via a DAW

do you know of any Rotary Switches that can be controlled in this way? most of my stuff is switches rather than variable pots/faders; I've been searching for a while for this kind of solution, so I could record and save settings on hardware, without actually having to mess with the signal chain at all

fair enough if Eucon is proprietary - although I'm sure there must be a high resolution GPL/GNU project out there somewhere! not sure what OSC is, will read up!

1024 sounds pretty good on paper, certainly allot better than midi, I wonder what its like in practice tho? can you get that 0.01db adjustment, or will it be jumping between 0.05 and 0.15, for eg - that would be the test for me personally, and its one of the reasons I now don't bother very much with MIDI in the studio at all (clients don't react well to engineers screaming at the workstation haha)

 

[zamproject]

Hi

I don't know any rotary motor switch... ALPS and other offer motor pot

for audio log fader curve with 1024 steps the resolution is:
0.04dB at top +10db
0.08dB at middle -10dB
0.25dB at bottom -60dB

Keep in mind that the fader need time to react so even steeped target position will be reached fade in (or fade out)

Zam

 

[Rogy]

Hi Zam,

Interested to integrate this into my EMT MM400 mixing desk...

I hope others will chime in as well!

Thanks for willing to share your knowledge and experience about this complex subject,

Best regards,

Rogy

 

[signalflow]

I've been researching doing this for a few months, no attempts to flesh anything out just yet.  My main hold back has been the programming of the midi-box.

What faders are you using?  Are they dual track?  Do you have any pictures of your progress so far?

I'll be following this thread to see how it turns out.

-Casey

 

[zamproject]

Hi

Actual fader is TKD MF914
dual track: servo linear 10K, audio log (10K), and touch
I have a second with math log, but I don't like the curve.
I also try the cheap ALPS, clearly way below the TKD

the progress is here http://groupdiy.com/index.php?topic=56212.msg774490#msg774490

Best
Zam

 

[zamproject]

Interested to integrate this into my EMT MM400 mixing desk...

Hi Rogy
I never come close to this EMT desk, but I belive the design is close to x69 studer series
I'm not sure event the motor fader alone (without any pcb) will fit the modules  :-\
Best
Zam

 

[gato]

I'm definitely interested and already did some preliminary research over at midi box. Your threads there and here have been really helpful. Your work saved me from going to the Mf_ng route! I know a bunch of other people like ruary are interested too. Timing is a little bad for me right now in terms of both time and money, but 6-12 months from now I'd be able to participate.

 

[signalflow]

Hi

Actual fader is TKD MF914
dual track: servo linear 10K, audio log (10K), and touch
I have a second with math log, but I don't like the curve.
I also try the cheap ALPS, clearly way below the TKD

the progress is here http://groupdiy.com/index.php?topic=56212.msg774490#msg774490

Best
Zam

I inquired about their pricing some time back after reading about them on the midibox forum(didn't realize you were the same person).  They have other options a bit more manageable in price than the 914 but they take different voltages. 8V instead of 12V to be exact.  What are your thoughts on dropping the voltage to match these other faders?

http://www.tkd-corp.com/products/fader/motorfader.html

From the rep here in the States

"On the MF914 , We are quoting for the M8V also, J12V is a very expensive Swiss coreless motor but M8V is a very good DC motor

MF914-A10K-J12V.                              $107.20 ea.
MF914-A10K-M8V.                                  $ 80.56ea.
LMH5-1101-A10K-M8V.                      $ 59.87ea.
LMA5-1101 -A10K-M8V.                    $ 57.65 ea.  "

They never got back to me with the LMH2 series cost.

I am very interested in this automation system.  I just went back to an analog console and would love to have the ability for automation.  Following this closely.

-Casey

 

[zamproject]

I have a 10V version
8V is the mabushi, 12V is the coreless

final peak to peak motor voltage is dependant of PSU rail  and resistor, it's flexible.
I wish I can try the coreless one day...to drop out the little sparkle that remain around console noisefloor
But I don't know if the motor already have the "converter" and can directly take DC.
LMx2 series seem to be end of life, replaced by LMx5, maybe that's why you don't get quote.

LMx5 is the new TKD product, conductive plastic too, so hope performance close to the MF914 regarding sliding noise, manufacturing tolerence is below MF914, so depend of curve matching compromise you think acceptable.
Mechanically they look way smaller than MF914, i'm wondering how the distance between servo, audio, and touch track influence the potential parasitic noise.

-edit: i just check, LMx5 is NOT conductive plastic, they say "direct print", believe it's carbon track...

If someone send me an LMH5 i'm ready to test it  :-X, more seriously I'm in contact since one year with european distributor, I will try to order this fader for test, don't know the availabilities yet.

TKD fader are great, i really like the two MF914 I have here for approval, the only complain I have is why they put the motor at the upper 0dB side, in analogue desk you'll be for sure close to analogue trace here, PAN or bus routing.

As personal note i recommend AP taper (audio log) not A taper (math log) , the fade in (out) feeling is better.

Best
Zam

 

[tskguy]

Zam,

Glad to see the interest! I would be willing to talk with the US TKD distributor to see if a group buy would get the price down...

Eric

 

[zamproject]

Zam,

Glad to see the interest! I would be willing to talk with the US TKD distributor to see if a group buy would get the price down...

Eric

By memory eu distributor offer 20% for 10pces and 50% for 100pces

I just put some clarification about midi implementation (as I have request by mail/MP)
Any midi data used by MCU can be mapped to any button/led, MB_NG offer LOT of programming options.
The new 16bit mapping release offer an incredible power to the system, I test it since two day, it's great. The direct consequence for 89MotioN system is that analogue fader attenuation match with DAW numerical value.
Now the limitation
MCU don't update instantly the signalling data to the control surface (only all data once per second) that mean midi mute automation can't be uses through MCU if you use optocouper for real mute like me.
You will have a lag from 0 to 1 sec when DAW send back "light this mute led". I use another midi port for this with all mute in a separate midi track. So solo/mute is usable when mixing, but NOT when printing the mix by using MCU
Some functions can't be full bidirectional with proper signalling update, like auto mode.
The only DAW data coming back for this is sysex to print the mode at LCD...no CC or midi note to sync led status. DAW won't update remote led, but remote button will always update it's own led and DAW automation status.
I use digital performer, maybe other DAW will react better. I have to test again with Reaper
There is now way to my knowledgre to run HUI properly because of fader dual CC handling (also limited to 9bit 512 step)
So PT user have to think about

Best
Zam

 

[gato]

A group buy would be great!
I'm a PT user, but I would change daws in a heartbeat. For me part of the point of building outboard/console/etc is to get to where I'm not looking at a screen all day. Even if I were to continue editing in pt, I'd just load everything up in another daw for mixing. I bet this sentiment will be common among other PT users who are thinking about building this system.

 

[john12ax7]

Very interesting project, think it should be moved to one of the discussion forum sections, more people would see it.

Some of my thoughts:

1) it's pricey for DIY. Is there a way to make a budget friendly non audio version? So just a fader control surface.

2) HUI mode would be useful for pro tools users.

 

[zamproject]

Very interesting project, think it should be moved to one of the discussion forum sections, more people would see it.

Some of my thoughts:

1) it's pricey for DIY. Is there a way to make a budget friendly non audio version? So just a fader control surface.

2) HUI mode would be useful for pro tools users.

Where do you think this tread below? project discussion mixer ?

1) yes as I say not a cheap DIY project,
using 20bucks/pce servo only ALP motor fader will for sure reduce the budget, but I don't see the point here, it's impossible to beat actual control surface like x-touch or icon or even artist mix.

2)I ear that, we will see if I can trick this

Zam

 

[tskguy]

Zam,

For some reason I think that the price is already less here in the states. Send me the exact TKD part number so I can reach back out and get some real numbers. And yes I am a PT user as well but agree that I would change that in a heart beat if this gets further traction. I also think you may be able to reduce the cost even further if we reduce some of the need for all of the functions at the fader side.. I personally don't mind keeping those features at the computer side. I also would be cool keeping mute automation on the computer side. I assume that would keep the cost down? Maybe not...

Eric

 

[signalflow]

I think keeping the features on board would be nice but if you don't need them don't populate them. I agree with you, the way my saber is set up I wouldn't have enough room for the buttons but could come in handy for others as long as it still functions without them.

-Casey

 

[john12ax7]

Very interesting project, think it should be moved to one of the discussion forum sections, more people would see it.

Some of my thoughts:

1) it's pricey for DIY. Is there a way to make a budget friendly non audio version? So just a fader control surface.

2) HUI mode would be useful for pro tools users.

Where do you think this tread below? project discussion mixer ?

1) yes as I say not a cheap DIY project,
using 20bucks/pce servo only ALP motor fader will for sure reduce the budget, but I don't see the point here, it's impossible to beat actual control surface like x-touch or icon or even artist mix.

2)I ear that, we will see if I can trick this

Zam

I would say move it either to  mixer section or drawing board, since it's still a discussion.

I've just been disappointed in control surfaces, so thinking about rolling my own. Mostly I want mixer faders that line up with my 500 series gear, I've fine with either audio or non audio passing through.