Jump to content

Fader Automation


Zam
 Share

Recommended Posts

Master Thorsten !!!

tks :)

I confirm the bug correction, 1.033 pre6 work fine here

Now I need to lower the jitter :)

Best

Zam

Link to comment
Share on other sites

Proto 2

here it is !!!

 

quick spec:

-motor analogue PI driver

-analogue touch detection

-512 steep (9bit) without jitter (can be better with shielded cable from AOUT and to AINSER)

-cross talk digital/motor to audio ... less than my console noise floor !!!

-acoustic noise... did the fader move ?

-response time better than 120ms for 100%travel 100mm

-response time better than 80ms for 100% travel 33mm

-never over shoot

-two automation pass with complex audio material result in null phase test 50dB below my line level, in static condition of course (when fader reach the target)

-to null audio under my console noise floor with two automation pass, I don't need to trim one more than +/- 0.01 dB !!!

 

I stop talking, this is better than words https://www.dropbox.com/s/o6v29atn209kyxi/proto%20II.mov?dl=0

 

Best

 

Zam

Edited by Zam
Link to comment
Share on other sites

  • 4 weeks later...
  • 1 month later...

Thank you zam,

 

i'll have a look, and send the distributor an email... but 110e for one fader may be a bit too much for my purpose. i am actually using the ALPS one from reichelt, but it sounds like an old printer from the 90s... 

Link to comment
Share on other sites

If you don't need fader with audio track forget the TKD, or look at other series with only servo and touch.

The "old priner sound" is not due to fader nor the fitted motor, it's linked to the PWM and H bridge.

Zam

Link to comment
Share on other sites

  • 3 months later...

Hey all

Long time without update on this topic, sorry for that.

In fact I just take the past few month to evaluate the prototype...daily use in a full album production.

The best possible crash test !

It's a positive experience, even if it need little improvement here and there  (I'm working on it!!!) , the whole system run rock solid :)

Zam

 

Link to comment
Share on other sites

Hi Zam 

are you going to share your config and work somehow? 

Regards, novski

Hi Novski

What do you want to know ?

The config is the same as the begining.

the improvements are more in the interaction with analog side to reduce digital/analog cross talk.

Like shielded wire, frontpanel grounding etc. it's so specific to my console mechanical integration, I don't know what to say ?

Best

Zam

Link to comment
Share on other sites

Hi

I still have digital noise crosstalk, it's "acceptable", but I really like to do better

I think I find the EMI sources, with the CLK at ribbon cable to J1.

I have some result with a grounded copper screen around the ribbon, but I would like to try to reduce the clock noise upstream.

So my question is simple,

I scope a 1.4us square wave (giving me 71.4kHz clock right ?), is that the SR scanning???

I'm little confused because at audio side I'v got 4.6kHz peak noise, with ALL the harmonics coming with

So if someone can confirm me the SR clock so I can calculate and try some filter in the clock line

Zam

Link to comment
Share on other sites

  • 4 weeks later...

Ok, I think I find the solution... I just use my remaining spare time to redesign PCB that keep SR far from fader... 85x250mm with 64Din, 64Dout, 8Ain and 8Aout, fitted with micro-match and angled pin header to match the lack of room ... 2cm height available in the armrest of the console. Motor driver and led/button pcb stay in the fader module.

Mechanical Integration Is a headache, but i'm close :)

Zam

Link to comment
Share on other sites

  • 4 weeks later...

some news

After one week fighting to recover my kicad project in the suddenly dead PC (exactly when I trace the last gerber pcb !!!! unbelievable ...), I finally was able to install stable Kicad to an OSX10.11.

I just ordered the PCBs :)

Best

JS

Edited by Zam
Link to comment
Share on other sites

  • 1 month later...

Ok here it is :)
 
1 fader pack with PID and LED pcb, also keeping the original fader PCB which interconnect passive part of the fader, audio and signaling PFL circuit

large.89MotioN_L.jpg.fe29d77f4755e8d9660 
everything fit in original studer cassette

large.89MotioN_T1.jpg.338dad4c795130926d

 
InOut pcb with 8 ch AD-DA and 64Din 64 Dout

large.89MotioN_IO.jpg.0f92a76254b593555f

 
19' rack with hi Q vero PSU +5 and +/-15, the core32 and the line transmiter

large.89MotioN_rack.jpg.c75c568c40944a36

 
 
And the beauty :)

large.89MotioN_light.jpg.7a510ab467fa2aa


I'm proud of what I done, and really like to thanks Thorsten for the beautiful software and all other for help around this place.
At this point everything working great, but it's not finished... I have to run it with full 8 fader pack to be sure it's a viable system, and I'm not able to fund this in a short range.

Zam

 

Edited by Zam
  • Like 1
Link to comment
Share on other sites

Zam, 

Amazing work! So I am 100% newbie on here and would love to understand this fully from the midibox perspective...Your work is amazing and an inspiration to me to do the same thing. Please help me understand this from all of the hardware perspective. Cant wait to see more. Video??

Thanks!!

Eric

Link to comment
Share on other sites

Hi Eric

Welcome here, but your not really newbie in DIY forum ;)

Nice to see you aroud this forum too.

 

Midibox is used for the soft side with MB_NG version, fully programmable midi interface. Midibox hardware design also used for the core and line transmitter/receiver

My part of work in this project is the analogue motor driver.

Also the analogue/digital I/O board, inspired by midibox MBHP modules to keep the system compatible

this 89InouT board is basically a combination and adaptation/improvement for my use case of: 1 AINSER, 1 AOUT_NG, 8 Din and 8 Dout.

It interconnect at core with standard MBHP connector specification, even if I use micromatch due to restricted room for my design, pinout still 100% compatible. just one side of ribbon cable use micromatch connector.

 

The motor driver is inspired from 90' highend console.

It's an analogue PID controller, connected to two CV value with comparator

one CV is the fader servo track (also send to the AD so the system can read/print position)

other CV come from the DA who give the position that the system have already printed/stored

This pcb also include an analogue touch detection (connected to a digital input) and an optocoupler (connected to a digital output) which in my use case active the original channel mute, with fully decoupling from the digital noisy world.

Best

Zam

 

 

Link to comment
Share on other sites

Most impressive!! 

So if I read this correctly the Pid board and the 89 inout board is your design.. very good work, I know that isn't an easy task! Do you plan on producing them or sharing the files? I am very familiar with PID based control systems due to my capsule cnc machine is using a servo based system based on the same technology! I have considered actually going away from HUI based coms and coding a smpte based system but that would be very time consuming and the automation in daws is already pretty comprehensive! Seems getting a MB_NG board learning is a good place to start. I am also trying to source the TDK faders, seems a bit challenging in the states!

 

Eric

Link to comment
Share on other sites

All three 89MotioN board are done, I made them fitting my need, but in the mean time OEM where not in conflict with my hardware implementation.

Small batch/kit/premounted system are of course possible, and I'm sure there is some interest here and at groupDIY for all those pple who dream about DIY true passive motorized fader.

Usecase can be hudge, from console modification to automated advanced summing box with all those ppl intensively using API500 rack

But, I have to see precisely how it work here as the system is opensource with some rules I don't look in details.

Also I won't involve other pple money or trust in this system before I'm not 100% sure it work with reliability. It's not a cheap project, depending of fader choices and mechanical integration requirement it's a 3k to 4k project for 8 fader.

As I say only one fader finished yet. I wait to fund a 8 fader pack, need about 2500 buck for this, main cost is TKD fader and front panel

Hope I can do this in the next 3 month, then at least 3 or 6 month for real life crash test at the studio.

It's a long term project, i'm on it since one and an half year now :pout:

Best

Zam

Link to comment
Share on other sites

I have considered actually going away from HUI based coms and coding a smpte based system but that would be very time consuming and the automation in daws is already pretty comprehensive! Seems getting a MB_NG board learning is a good place to start. I am also trying to source the TDK faders, seems a bit challenging in the states!

I don't see any reason today to have a separate  interface/computer that need to be sync too, to record/play automation data. As you say, modern DAW/computer can handle all this pretty good with advanced protocole, HUI is ok MCU too (the one I use)

Just be aware that midi box don't handle the HUI dual CC data used for 10bit fader, so PT user might have problem in the actual state

Another route I have to test is the OSC protocole, I have all part here to do an MB_ETH board, just need more time

It might not be faster but hope can open more extended bi-directionality and function than HUI/MCU.,

 

Regarding fader the TKD I have don't have particularities except the high qualities of conductive plastic track, I have a 10v motor version.

I will test again cheap ALPS I have with this "finished" integration, maybe he will react better to digital/motor crosstalk than in my first proto(board)

But for sure sliding noise will not change (we talk about 40dB difference between TKD and ALPS !!!)

I'm sure P&G will work fine too, but honestly I can't buy all possible mid/high-end fader to perform test. there is at least 4 or 5 candidate.

 

Best

Zam

Link to comment
Share on other sites

Zam,

I understand your hesitance in releasing pcb's without testing! Over at groupdiy most just to do it to and let the community help debug :) I think you approach is better :). I most interested in your pid controll :) I have PCB manufacturing contacts and can get small runs done if needed... In terms of fader panals I am pretty good in that regard. My capsule cnc machine engraves pretty well... 

So to get started I need,

MB_NG, midibox MBHP module, 1 AINSER, 1 AOUT_NG, 8 Din and 8 Dout. and a PID motor control.(.I have the schematic from groupdiy) :angelnot:?? I may not need all of this for initial testing and learning?  

I actually have a RFQ out with TKD local distributor to get pricing. Can you share the Alps fader you used initially? I am thinking it would be a good place to start testing but I cant seem to find one with all three tracks. 
Do you think that the HUI protocol will work? What are the drawbacks? I am mostly interested in fader automation not mute or other actions really. I don't mind using a mouse now and then. 

I am not sold on protools long term but for now its what I have. 

This is cool stuff!!

 

 

Link to comment
Share on other sites

Eric

Community debuging is a good thing, when most of DIY stuff is well know clone and in fact involve "only"some small rework or component BOM error. Most of built still relative low investment. Here we talk about few k euros as entry "test" that's why I'm little cold to share right now, If only 5 ppl start buying pcb to me and invest high value part involved it will result in a global 20k lost, I just can't assume that even if it's not my money.

As say I just have to read again how it is when selling midibox based system.

To start you need at least a core, 1DA 1AD 1Din and one Dout, and a motor driver, I'm not sure the schematic is enough if you want it before 2017 :innocent:

 

TKD just release a new series, still conductive plastic if I remember, but spec and tolerance lower than MF914 series. Don't know the price I don't ask yet.

The ALPS for the first proto is RSAON12 audio log+servo lin+touch, 30 bucks, carbon track, 8V motor (PID at +/- 10 or 12v should be enough)

Sliding noise 20-30dB below line level (TKD CP track have little low kick just over console noise floor, 60-70dB below line level !!!)

 

I don't think HUI will work, I better let Thorsten reply, but if I remember it's a "deep" implementation problem, HUI use two CC data to map 9 bit value and it's seem not really midi friendly, I where able to send the two CC data with tricky sysex message, but MB can't read it back from DAW. Everything else work fine in HUI but not the most important, the fader.

The MCU use true 14bit pitchbend, My design implementation work without jitter at 11bit, but 10 is enough (0.1mm step) and I save money at DA converter.

Ok time to sleep in W europe now.

Best

Zam

Link to comment
Share on other sites

Zam,

Wow it looks to work perfectly!!! It's an exciting endeavor for sure!! I am far from starting this project I need to install my new neotek first and get all the bugs worked out on that front... And of coarse pay for it!!

I am wondering thought the 4k of estimated cost?? Is it board production in small runs with the fader cost? It's seems a bit high to me.. Or am I grossly underestimating the BOM costs for your boards?? I have a very close friend who is an engineer for a midi instrument company and has great contacts to get pcb's and components at reasonable costs...

Eric

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

×
×
  • Create New...