The No.1 Website for Pro Audio
Behringer Yamaha CS 80 clone (DS 80) Speculation Thread
Old 2 weeks ago | Show parent
  #2431
Gear Maniac
Quote:
Originally Posted by Pablo668 ➡️
Basically, how the hell do you clone this?
Essentially, by starting small. Identify the smallest building blocks of the system that don’t interact with others, and reimplement that behavior with modern components. It isn’t necessarily mandatory to adhere to the original circuit diagram as long as the behavior is emulated correctly. It’s also worth knowing that a lot of the CS-80 is discrete 4000-series CMOS logic to route control signals around, power ICs to drive the lamps on the preset selector switches, that sort of thing. Plus, for something like the voice cards, you only have to clone it once and you have all 16.

Aside from stability concerns, once you have the control multiplexing sorted out, a poly synth is just a monosynth replicated several times over.
Old 2 weeks ago | Show parent
  #2432
Gear Addict
 
Quote:
Originally Posted by Pablo668 ➡️
I guess what I'm trying to say is it opened my eyes as to what a real big task it would be to make a clone of this synth.

So yeah, as much as I want a clone of one of these things, I reckon I can wait for them to do it and get it right.
Making a 1:1 clone would be cost prohibitive nowadays, in terms of manufacturing inefficiency (massive wire harnesses, thru-hole parts, etc.) that can all be consolidated with digital circuitry, SMD, and better PCB organization (back planes, etc.). The two really important things are: sound, and performance controls.

The first requires an accurate recreation of the original circuit and its nonlinear characteristics. This means recreating the charge pump oscillators, SVF filters, VCAs, ring mod and chorus, have them all respond properly to control signals, and have them exhibit the same distortion (wah/wah, etc.) as the original. Not an easy task, especially since the original ICs aren't available.

The second requires recreating the tactile feel of the keyboard with its poly-at, knobs, and the ribbon control. Weight of the instrument is also a factor to some degree.

Anything else in the middle doesn't really matter much. Wire looms can be replaced with backplane PCBs or just miniaturized all into SMD and put it on one PCB, like DD MkII (which is just a CEM3340 based system with similar voice architecture, not a close recreation of the original circuit).
Old 2 weeks ago | Show parent
  #2433
Lives for gear
 
kurzweil's Avatar
 
🎧 5 years
With such a high-profile and prestigious clone such as the CS80, I can't see them replacing the most important discrete circuits with off-the-shelf ICs such as the 3240. I can't see them using digital envelopes / LFOs either. They can get away with it in lower profile / lower cost products such as the Monopoly, but not here.

Otherwise, you just get a polysynth that sort of looks like the original, but sounds nothing like it. That would be damaging to the whole clone project. They have shown in the Model D that you can get very close to the original sound and behaviour by using SMT versions of discrete circuits.

Last edited by kurzweil; 2 weeks ago at 12:18 PM.. Reason: I can't spelle
Old 2 weeks ago | Show parent
  #2434
Deleted 7737321
Guest
Any year now.

Old 2 weeks ago
  #2435
Gear Maniac
 
Pablo668's Avatar
 
Great answers guys, thanks!
Old 2 weeks ago | Show parent
  #2436
Lives for gear
 
SkyWriter's Avatar
 
🎧 10 years
Quote:
Originally Posted by kurzweil ➡️
With such a high-profile and prestigious clone such as the CS80, I can't see them replacing the most important discrete circuits with off-the-shelf ICs such as the 3240. I can't see them using digital envelopes / LFOs either. They can get away with it in lower profile / lower cost products such as the Monopoly, but not here.

Otherwise, you just get a polysynth that sort of looks like the original, but sounds nothing like it. That would be damaging to the whole clone project. They have shown in the Model D that you can get very close to the original sound and behaviour by using SMT versions of discrete circuits.
Here's the block diagram. Let's get more specific.

The voice card section under 'BLOCK DIAGRAM' will be done on a best effort basis; analog for analog. They have replaced VCO chips with basic waveform generation with other chips. I don't believe in the sanctity of triangle, saw, or sine wave generation. One triangle is much like another. So that's fine with me.

Filters, EG's LFO's will be done in the original fashion a much as is reasonably possible within the cost envelope.

The rest is up for replacement one way or another as long as 'look anf feel' a prereserved.

GS won't render the image, it's too big I guess.

http://therogoffs.com/cs80/manuals/C...ck_diagram.jpg

Last edited by SkyWriter; 2 weeks ago at 01:53 PM..
Old 2 weeks ago
  #2437
vlz
Lives for gear
 
vlz's Avatar
Interestingly enough, today we get this from Behr. It's almost as if they were following this conversation.

Behringer Yamaha CS 80 clone (DS 80) Speculation Thread-screenshot-2021-04-23-14.19.55.jpg
Attached Thumbnails
Behringer Yamaha CS 80 clone (DS 80) Speculation Thread-screenshot-2021-04-23-14.19.55.jpg  
Old 2 weeks ago | Show parent
  #2438
Lives for gear
 
They're clearly aware of this thread and keeping and eye on this

Old 2 weeks ago | Show parent
  #2439
Lives for gear
 
SkyWriter's Avatar
 
🎧 10 years
Quote:
Originally Posted by vlz ➡️
Interestingly enough, today we get this from Behr. It's almost as if they were following this conversation.
Idk, it all seems a pretty much 'run of the mill' job to me.

Reverse engineering is as much a standard process as building it from scratch in the first place.

As much as folks want to create mystery, I haven't seen anything I would call 'special' here. Even rate of development is typical.
Old 2 weeks ago | Show parent
  #2440
Lives for gear
 
kurzweil's Avatar
 
🎧 5 years
Quote:
Originally Posted by vlz ➡️
Interestingly enough, today we get this from Behr. It's almost as if they were following this conversation.

Behringer Yamaha CS 80 clone (DS 80) Speculation Thread-screenshot-2021-04-23-14.19.55.jpg
Good to hear that they are doing things properly, as one would expect from such a prestige project such as this. It makes the Model D project seem trivial in comparison, as everything needs to be made to work with a potentially huge number of DACs and CV trails, assuming that they will keep some matrix modulation capability under the hood using the common polysynth firmware.
Old 2 weeks ago | Show parent
  #2441
Lives for gear
 
kurzweil's Avatar
 
🎧 5 years
Quote:
Originally Posted by SkyWriter ➡️
Here's the block diagram. Let's get more specific.

The voice card section under 'BLOCK DIAGRAM' will be done on a best effort basis; analog for analog. They have replaced VCO chips with basic waveform generation with other chips. I don't believe in the sanctity of triangle, saw, or sine wave generation. One triangle is much like another. So that's fine with me.

Filters, EG's LFO's will be done in the original fashion a much as is reasonably possible within the cost envelope.

The rest is up for replacement one way or another as long as 'look anf feel' a prereserved.

GS won't render the image, it's too big I guess.

http://therogoffs.com/cs80/manuals/C...ck_diagram.jpg
With all those integrated circuits in the schematics, it looks like a very time-consuming job for one person (?!) to unpick. I bet there are a few errors in the schematics too, or at least some details that don't sound quite right when they are translated into equivalent (nearest tolerance) surface-mount components.
Old 2 weeks ago | Show parent
  #2442
Lives for gear
 
SkyWriter's Avatar
 
🎧 10 years
Quote:
Originally Posted by kurzweil ➡️
With all those integrated circuits in the schematics, it looks like a very time-consuming job for one person (?!) to unpick. I bet there are a few errors in the schematics too, or at least some details that don't sound quite right when they are translated into equivalent (nearest tolerance) surface-mount components.
This is the voice card - the core sound

http://therogoffs.com/cs80/manuals/C...%20Diagram.jpg

Looks like they'll put their VCO chip in there, and the closest filter circuity, EG generator; whether discrete or coolaudio chip. Waveshaper would be custom depending on what it actually does.

The rest the implementation is not important as long as the UI interface and response characteristics are adhered to.

Cost is king.

(Why don't these jpg's render with img tags?)

Of all the synths, this will be the most controversial based on the schematics. There's nothing to hang your hat on - it's all going to be touched and revamped.
Old 2 weeks ago | Show parent
  #2443
Lives for gear
 
kurzweil's Avatar
 
🎧 5 years
So long as the lights are pretty..

Old 2 weeks ago
  #2444
Gear Maniac
 
Pablo668's Avatar
 
If it sounds close enough, and the workflow is the same, I'll be a happy Pablo
Old 2 weeks ago | Show parent
  #2445
Lives for gear
 
🎧 10 years
Studio Electronics has a CS filter on one of their Boomstars, I wonder what's in that?

Edit: and Old Crow / MOTM have a CS80 filter clone. I bet Behringer will be "taking notes" there.

https://synthtech.com/motm/480/
Old 2 weeks ago | Show parent
  #2446
Gear Addict
We've reached the prestigious breadboard stage, I guess?
Old 2 weeks ago | Show parent
  #2447
DGL
Lives for gear
 
🎧 5 years
Quote:
Originally Posted by SkyWriter ➡️
Looks like they'll put their VCO chip in there, and the closest filter circuity, EG generator; whether discrete or coolaudio chip. Waveshaper would be custom depending on what it actually does.
See this reply to a comment on the FB post,

Quote:
Originally Posted by Behringer
our goal is to totally replicate the behavior of all building blocks. No 3340 will be used but discrete components.
Old 2 weeks ago
  #2448
Lives for gear
 
🎧 10 years
9 years back:

VCO153 1U CS VCO

1U x 5U form factor VCO module based on equivalent discrete circuit derived from the Yamaha IG00153 VCO chip as used in the CS80. Features CS-style sawtooth, sine and Yamaha-normalized 50% to 90% CV variable pulse outputs; switchable exponential/linear attenuated FM input, 1v/octave CV response. Note: mounting screws and power cable not included.

https://www.muffwiggler.com/forum/vi...c55564efd19bc2

https://cs80-com.myshopify.com/produ...o153-1u-cs-vco
Old 2 weeks ago | Show parent
  #2449
Lives for gear
 
SkyWriter's Avatar
 
🎧 10 years
Quote:
Originally Posted by DGL ➡️
See this reply to a comment on the FB post,
Snake eyes!

I've seen a lot of goals go by the wayside.
Old 2 weeks ago | Show parent
  #2450
Lives for gear
 
matt pinchin's Avatar
Quote:
Originally Posted by DGL ➡️
See this reply to a comment on the FB post,
Note the wording. Replicate the behaviour. They are not recreating the og chips
Old 2 weeks ago | Show parent
  #2451
Lives for gear
 
🎧 5 years
Quote:
Originally Posted by matt pinchin ➡️
Note the wording. Replicate the behaviour. They are not recreating the og chips
And that is very good.
Old 2 weeks ago | Show parent
  #2452
Lives for gear
 
matt pinchin's Avatar
Quote:
Originally Posted by Mutantt ➡️
And that is very good.
Better than using those other osc chips but they have to get it right. They did this with the ms1 filter which is why it supposedly has a slightly different tone to the og filter.
Old 2 weeks ago
  #2453
DGL
Lives for gear
 
🎧 5 years
If they wanted to replicate the chips just send them to reverse engineer extraordinaire Ken Shirriff, he can even reverse engineer dies even if there wasn't a clear idea of what the chip does in the first place. The issue would be getting clean/usable de-capped chips, esp. if they are in ceramic packages.
Old 2 weeks ago | Show parent
  #2454
Deleted 7737321
Guest
Quote:
Originally Posted by vlz ➡️
Interestingly enough, today we get this from Behr. It's almost as if they were following this conversation.

Old 2 weeks ago | Show parent
  #2455
Lives for gear
 
SkyWriter's Avatar
 
🎧 10 years
Quote:
Originally Posted by DGL ➡️
If they wanted to replicate the chips just send them to reverse engineer extraordinaire Ken Shirriff, he can even reverse engineer dies even if there wasn't a clear idea of what the chip does in the first place. The issue would be getting clean/usable de-capped chips, esp. if they are in ceramic packages.
They have a whole company for that now; cool audio. It's not unique, it just takes time and experience. I did it for 40 years.

In this case, it hardly matters. The chips are not mass produced with a solid sounds anyone can identify. They're little black boxes they can easily replicate anyway they want. Very few folks will be able to authentically point to differences - there's just not enough machines left to compare with.

Besides these are plastic - messy. Ceramic lids almost fall off by comparision.
📝 Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 138 views: 12504
Avatar for Duchemole
Duchemole 1st July 2016
replies: 752 views: 86123
Avatar for jdkJake
jdkJake 28th October 2020
replies: 524 views: 50139
Avatar for bsp804
bsp804 14th March 2021
replies: 85 views: 7872
Avatar for MikeWestSE
MikeWestSE 18th June 2020
Topic:
Post Reply

Welcome to the Gearspace Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…

Forum Jump
Forum Jump