NorthernLightX Posted September 2, 2012 Report Share Posted September 2, 2012 Hello, I am still in the planning phase for the internals of my build. I have run into an unforseen issue: I have 4 SSM2044 boards and an "original" AOUT board. I made a slight miscalculation, as the AOUT can only supply 8 CV outputs, and 4 stereo filters need 16 CV inputs (both Cutoff and Resonance). What would be the best solution in this case? 1. I could splice the AOUT outputs to have the Left and Right filter channels respond to the same CV output. I have read that this can produce weird issues with polyphony though, where MB6582 will not retrigger the filter envelope after 3 notes; it plays the 3rd note on the other SID so it expects that there is a dedicated AOUT channel for this SID. Are there options to circumvent this? 2. I could build another AOUT board, but I do not know if daisy-chaning is supported in the firmware. It is most probably not within my skillset to implement it myself. 3. I could settle for 4 SIDs with external filters and the other 4 SIDs without. Currently I am leaning towerd option 3, and implementing a patch matrix on the frontpanel to have some added flexibility with routing which SIDs get external filters. If you know of a better solution, or have anything to add otherwise, please chime in. Quote Link to comment Share on other sites More sharing options...
rosch Posted September 2, 2012 Report Share Posted September 2, 2012 i haven't done it yet. but from what i've read with Aout_NG's you'd need one for every sid pair. it's been explained to me couple of times, but some time ago and i have to read back again to understand it -again. Quote Link to comment Share on other sites More sharing options...
ytsestef Posted September 3, 2012 Report Share Posted September 3, 2012 I'm currently working on altering the firmware so that it always routes AOUT to channels 1/2 anyway, regardless of the SID that is triggered by the dynamic voice allocator (ie even if the right SID is playing, the curoff/resonance values are forwarded to the left AOUT pair). I can't really guarantee that I will succeed, but you should hear from me during the next week, I am currently trying. :) Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted September 3, 2012 Author Report Share Posted September 3, 2012 That sounds very useful! I hope you succeed, keep me posted please :thumbsup: Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted September 3, 2012 Author Report Share Posted September 3, 2012 i haven't done it yet. but from what i've read with Aout_NG's you'd need one for every sid pair. it's been explained to me couple of times, but some time ago and i have to read back again to understand it -again. That is what I was hoping, but I have not found any useful docs to verify this. If you can find the info you were given and post it here I would be very grateful! Quote Link to comment Share on other sites More sharing options...
Schrabikus Posted September 3, 2012 Report Share Posted September 3, 2012 with Aout_NG's you'd need one for every sid pair. Oh...that's sad. I was hoping to control 2xSSM2044 + 2xCEM3379 LPF's from one AOUT_NG :hmm: Quote Link to comment Share on other sites More sharing options...
rosch Posted September 3, 2012 Report Share Posted September 3, 2012 i think i've bookmarked the threads where it's explained. i think someone (maybe jojjelito) took the time and explained to me again in private message. i'll look for it later Quote Link to comment Share on other sites More sharing options...
Schrabikus Posted September 3, 2012 Report Share Posted September 3, 2012 found that The SID doesn't control the VCFs, that's done by the 4 core modules. Env and VCF data is local to each core (no, it's not shared from a "master" CPU via CAN) so you'll need one AOUT_NG per core if you want the full custom shine. AFAIK the AOUT_NG is connected using I2C SPI to the core, I don't think that's set up as a multi master bus as is. Perhaps TK, Wilba or one of the guru's can tell us if that will change later down the road? The connection is: Core -> SID, Core -> Aout, SID Audio -> VCF/VCA, AOUT CV -> VCF/VCA. Plus control bits as necessary using the available port headers at each core. Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted September 3, 2012 Author Report Share Posted September 3, 2012 with Aout_NG's you'd need one for every sid pair. Whoops, I misread that completely. I read 1 SSM2044 per SID pair, and one AOUT total. It does make sense though, that 1 AOUT per Core is needed. I have a few spare MAX525 chips, but not enough for 4 AOUT boards, let alone the footprint of 4 of those (and the other parts needed; precision trimmer hell!). I don't expect TK (or anyone else for that matter) to modify the MBSID firmware so that Filter info is sent to the master PIC Core. Maybe in a future version where for example all PIC cores act as slaves to a single LPC Core (if slave PIC cores are even necessary in an LPC solution), the LPC Core could generate all the Filter info and the total solution could function with only one AOUT or AOUT_NG, but if that ever happens I expect it will be many months before it's released. A daring option would be to do a trace cut on the CS, SI and SC lines on the AOUT board, and feed the left side of the board from Core1 through J1 and the Right side from Core2 through J2. This way no CV channels are wasted. It would still imply I need a second (also trace-cut) AOUT board for SID pairs 3 and 4, but that still sounds better than 4 AOUTs with 6 unused channels per board. The trace cut on the original AOUT board is feasible (I don't think it is feasible on AOUT_NG though). If it is "allowed" to run an AOUT board with only one set of MAX525 and TL074 stuffed, I think this might work :drool: Quote Link to comment Share on other sites More sharing options...
TK. Posted September 9, 2012 Report Share Posted September 9, 2012 A daring option would be to do a trace cut on the CS, SI and SC lines on the AOUT board, and feed the left side of the board from Core1 through J1 and the Right side from Core2 through J2. This way no CV channels are wasted. It would still imply I need a second (also trace-cut) AOUT board for SID pairs 3 and 4, but that still sounds better than 4 AOUTs with 6 unused channels per board. The trace cut on the original AOUT board is feasible (I don't think it is feasible on AOUT_NG though). If it is "allowed" to run an AOUT board with only one set of MAX525 and TL074 stuffed, I think this might work :drool: yes, thats possible! :) Best Regards, Thorsten. Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted September 9, 2012 Author Report Share Posted September 9, 2012 Great! As soon as I continue with the MB6582 project I'll document my progress with this. I'll need the "enclosure" for that, hopefully that can be finished next week. Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted September 14, 2012 Author Report Share Posted September 14, 2012 I'm still in doubt how I will wire everything up. I like to have it as flexible as possible, so I'm inclined to integrate patchpoints. This is the current list of in- and outputs that I have made up so far: - MIDI IN and OUT (obviously). I will use an IIC MIDI module to attach the internal keyboard to the MB6582 baseboard. - 8x audio OUT. I want to use mono plugs - 16x CV OUT - 4x CV per Core module, with the AOUT boards wired like described in post #9 optional: - 8x Gate OUT - 2x Gate per Core module. I'm not sure if I need these, are they useful? - 8x CV IN - 2x CV per Core module. Not sure about these as well, why would I want to control the SIDs with CV if I have the powerful MBSID engine at hand? I will then later add the SSM filters as expansion modules with their own frontpanel, with audio IN, CV IN, and audio OUT per channel. I did not solder the relays so don't need any gate signals for these then, or do I? Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted October 6, 2012 Author Report Share Posted October 6, 2012 OK, stumbled upon another piece of confusing info while planning the CV outputs: Seppoman describes how the AOUT_NG only outputs + / - 2,8V without modification, and that the Original AOUT probably suffers from the same "bug". Since I'm building this thing with compatibility with other external modules in mind; do I need to change resistor values on the Original AOUT as well to get + /- 5V? I'd also like some more info on the difference and necessity of unipolar and bipolar CV outputs. I'll try to Google some more info tonight but if anyone has some good recommendations please post them! :thumbsup: Quote Link to comment Share on other sites More sharing options...
smokestacksproductions Posted October 6, 2012 Report Share Posted October 6, 2012 you'll need to use bipolar for your ssm filters Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted October 6, 2012 Author Report Share Posted October 6, 2012 (edited) you'll need to use bipolar for your ssm filters Yes, I know that. I was asking for literature about WHY I need it. [edit] For example: all I find about CV voltages is that it's between 0v and 5v (because some guy called Moog thought it up that way). Now, why do I need a negative voltage then? Only for calibration of the 0v - 5v range perhaps? Edited October 6, 2012 by NorthernLightX Quote Link to comment Share on other sites More sharing options...
Shuriken Posted October 6, 2012 Report Share Posted October 6, 2012 I am pretty sure that Seppoman mentioned his SSM2044 VCF design matches the 2,8V. So basically you cannot mix his designs and common modular VCF designs. Well you can but you will have less resolution ofc. Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted October 6, 2012 Author Report Share Posted October 6, 2012 For the AOUT_NG a fix was suggested that only required to change a few resistor values. I was still under the presumption that a similar fix would be possible on the filter (and VCA) modules. Maybe I'm totally wrong? Quote Link to comment Share on other sites More sharing options...
Shuriken Posted October 6, 2012 Report Share Posted October 6, 2012 (edited) He actually mentions this in the post you linked: [...] BTW to all SSM2044 builders: The module is designed with this (wrong) bipolar range, so you don't need to change anything on your AOUT_NGs. I guess for proper +/- 5V operation I'll have to try out a few other resistor values on the SSM module, but this will only apply if you if you're using a different CV source than an AOUT_NG. So basically you would have to change both the AOUT and the VCF. Edited October 6, 2012 by Shuriken Quote Link to comment Share on other sites More sharing options...
NorthernLightX Posted October 6, 2012 Author Report Share Posted October 6, 2012 So basically you would have to change both the AOUT and the VCF. I'm okay with that :thumbsup: Now I only need to know why the frequency control need a bipolar supply and the resonance doesn't, and why a bipolar supply is needed at all for proper v/oct response (which to my knowledge is still only 0v - 5v). Quote Link to comment Share on other sites More sharing options...
Shuriken Posted October 6, 2012 Report Share Posted October 6, 2012 I am afraid i can't help you with that. Need to fix my AOUT and VCF as well :wink: Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.