Korg Poly800/EX800 Users group photo

Yahoo Groups archive

Korg Poly800/EX800 Users

Archive for korgpolyex.

Index last updated: 2026-03-30 01:10 UTC

Thread

New CPU - idea

New CPU - idea

2014-10-09 by Michael Hawkins

Hello Poly fans,

See the attached PDF. This would be a replacement for the 80C85 CPU in the Poly-800. It will use a STM32F373 chip, an M4 with up to 256k flash. It runs at 72MHz

Before I embark on the complete rewrite of the HAWK software from original 80C85 assembler into C I thought I would get feedback from everyone about going in this direction.

The problem here is that all development work on the existing software for the existing HAWK kit would stop. Anyone that has an existing HAWK would be stuck with what they have today. The Atomahawk would also have to be redesigned a bit.

The effort required to rewrite the software is quite large. If I had some help from some committed Poly fans that also have programming experience that would help a great deal.

I am too busy trying to survive in the world these days that I must seek help before we could make this project happen. But, with a 32 bit 72Mhz CPU instead of 8 bit 6Mhz there will be some amazing possibilities available to us.

Any thoughts from you all will be interesting to read.

/Mike

Re: [korgpolyex] New CPU - idea [1 Attachment]

2014-10-09 by Héctor David Polo Cheva

EX800 considered?
Thanks


Enviado desde mi smartphone BlackBerry 10.
De: Michael Hawkins korgpolyex800@... [korgpolyex]
Enviado: jueves, 9 de octubre de 2014 09:22
Para: korgpolyex@yahoogroups.com
Responder a: korgpolyex@yahoogroups.com
Asunto: [korgpolyex] New CPU - idea [1 Attachment]

 

Hello Poly fans,

See the attached PDF. This would be a replacement for the 80C85 CPU in the Poly-800. It will use a STM32F373 chip, an M4 with up to 256k flash. It runs at 72MHz

Before I embark on the complete rewrite of the HAWK software from original 80C85 assembler into C I thought I would get feedback from everyone about going in this direction.

The problem here is that all development work on the existing software for the existing HAWK kit would stop. Anyone that has an existing HAWK would be stuck with what they have today. The Atomahawk would also have to be redesigned a bit.

The effort required to rewrite the software is quite large. If I had some help from some committed Poly fans that also have programming experience that would help a great deal.

I am too busy trying to survive in the world these days that I must seek help before we could make this project happen. But, with a 32 bit 72Mhz CPU instead of 8 bit 6Mhz there will be some amazing possibilities available to us.

Any thoughts from you all will be interesting to read.

/Mike


Re: [korgpolyex] New CPU - idea

2014-10-09 by Michael Hawkins

This would work in Poly, EX-800 and MK2.

/Mike


On Thursday, October 9, 2014 10:27 AM, "Héctor David Polo Cheva hdpolo@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
EX800 considered?
Thanks


Enviado desde mi smartphone BlackBerry 10.
De: Michael Hawkins korgpolyex800@... [korgpolyex]
Enviado: jueves, 9 de octubre de 2014 09:22
Para: korgpolyex@yahoogroups.com
Responder a: korgpolyex@yahoogroups.com
Asunto: [korgpolyex] New CPU - idea [1 Attachment]

 
Hello Poly fans,

See the attached PDF. This would be a replacement for the 80C85 CPU in the Poly-800. It will use a STM32F373 chip, an M4 with up to 256k flash. It runs at 72MHz

Before I embark on the complete rewrite of the HAWK software from original 80C85 assembler into C I thought I would get feedback from everyone about going in this direction.

The problem here is that all development work on the existing software for the existing HAWK kit would stop. Anyone that has an existing HAWK would be stuck with what they have today. The Atomahawk would also have to be redesigned a bit.

The effort required to rewrite the software is quite large. If I had some help from some committed Poly fans that also have programming experience that would help a great deal.

I am too busy trying to survive in the world these days that I must seek help before we could make this project happen. But, with a 32 bit 72Mhz CPU instead of 8 bit 6Mhz there will be some amazing possibilities available to us.

Any thoughts from you all will be interesting to read.

/Mike



Re: [korgpolyex] New CPU - idea

2014-10-09 by Héctor David Polo Cheva

Great!
Thanks!


Enviado desde mi smartphone BlackBerry 10.
De: Michael Hawkins korgpolyex800@... [korgpolyex]
Enviado: jueves, 9 de octubre de 2014 09:29
Para: korgpolyex@yahoogroups.com
Responder a: korgpolyex@yahoogroups.com
Asunto: Re: [korgpolyex] New CPU - idea

 

This would work in Poly, EX-800 and MK2.

/Mike


On Thursday, October 9, 2014 10:27 AM, "Héctor David Polo Cheva hdpolo@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
EX800 considered?
Thanks


Enviado desde mi smartphone BlackBerry 10.
De: Michael Hawkins korgpolyex800@... [korgpolyex]
Enviado: jueves, 9 de octubre de 2014 09:22
Para: korgpolyex@yahoogroups.com
Responder a: korgpolyex@yahoogroups.com
Asunto: [korgpolyex] New CPU - idea [1 Attachment]

 
Hello Poly fans,

See the attached PDF. This would be a replacement for the 80C85 CPU in the Poly-800. It will use a STM32F373 chip, an M4 with up to 256k flash. It runs at 72MHz

Before I embark on the complete rewrite of the HAWK software from original 80C85 assembler into C I thought I would get feedback from everyone about going in this direction.

The problem here is that all development work on the existing software for the existing HAWK kit would stop. Anyone that has an existing HAWK would be stuck with what they have today. The Atomahawk would also have to be redesigned a bit.

The effort required to rewrite the software is quite large. If I had some help from some committed Poly fans that also have programming experience that would help a great deal.

I am too busy trying to survive in the world these days that I must seek help before we could make this project happen. But, with a 32 bit 72Mhz CPU instead of 8 bit 6Mhz there will be some amazing possibilities available to us.

Any thoughts from you all will be interesting to read.

/Mike




Re: [korgpolyex] New CPU - idea [1 Attachment]

2014-10-09 by gordon@...

On Thu, Oct 09, 2014 at 07:22:09AM -0700, Michael Hawkins korgpolyex800@... [korgpolyex] wrote:
> Hello Poly fans,
>
> See the attached PDF. This would be a replacement for the 80C85 CPU in the Poly-800. It will use a STM32F373 chip, an M4 with up to 256k flash. It runs at 72MHz
>

Sounds like a beast of an idea. I bet it could be adapted to other 8085-based machines too.

> The problem here is that all development work on the existing software for the existing HAWK kit would stop. Anyone that has an existing HAWK would be stuck with what they have today. The Atomahawk would also have to be redesigned a bit.

I know you've got a lot invested into the HAWK mod, but if you drop support of the original 80C85 version to concentrate on an ARM-based version you could just release the source and toolchain for the old stuff? That way if anyone cares enough about fixing a bug they can either do it themselves, or pay someone to do it. Who knows, maybe someone will take it in a wholly unexpected direction?

>
> The effort required to rewrite the software is quite large. If I had some help from some committed Poly fans that also have programming experience that would help a great deal.

I'd be quite into this.

>
> I am too busy trying to survive in the world these days that I must seek help before we could make this project happen. But, with a 32 bit 72Mhz CPU instead of 8 bit 6Mhz there will be some amazing possibilities available to us.

Likewise. I have people clamouring for Mirage goodies that I just don't have time to work on. Seems they don't want it enough to pay my usual day rate though ;-)

--
Gordonjcp MM0YEQ

Re: New CPU - idea

2014-10-10 by ultragalore@...

ofcourse this new ARM thing is amazing and is the way to go if you want to make things happen like software midi-thru and proper portamento, but since I have my HAWK kit just 4 days I 'll have to pass on another $$ upgrade anytime soon.

Re: [korgpolyex] Re: New CPU - idea

2014-10-10 by Michael Hawkins

Yes, I figured some more recent buyers of the HAWK might be disappointed. But I can't start development on the ARM board until I know enough people will be interested.

But here's some even more interesting news.

If enough people are interested, I am going to develop a replacement for the tone generator first. The MSM5232 is the most awful sound source in the history of the world unless you're an arcade game. It is without doubt the biggest reason that the Poly-800 is limited and criticized. It drives me crazy!

So after talking to my friendly ARM expert, I realized that instead of replacing the 80C85 CPU first, it would be much better to design a replacement for the MSM5232.

This would give us at least 8 voices (probably more) with a full variety of waveforms and truly clock and modulation independence from one another. The emulator would work with both the HAWK and anything new that we come up with after that.

Stay tuned!

/Mike





On Friday, October 10, 2014 6:30 AM, "ultragalore@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
ofcourse this new ARM thing is amazing and is the way to go if you want to make things happen like software midi-thru and proper portamento, but since I have my HAWK kit just 4 days I 'll have to pass on another $$ upgrade anytime soon.


Re: [korgpolyex] Re: New CPU - idea

2014-10-10 by Michael Hawkins

No, picture was to replace the 80C85 CPU - which I think is well worth doing.

But with a little tweaking of that board design, we can emulate and dramatically improve the TG MSM5232.

And I think that is much more worthwhile doing than upgrading the CPU - for now at least.

/Mike


On Friday, October 10, 2014 11:21 AM, "bperkins211@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
so the "chip" you posted a pic of was a replacement for the old TG?


Re: [korgpolyex] Re: New CPU - idea

2014-10-11 by ultragalore@...

I think I mentioned to you that the amount of hacking you did on the Poly is almost like building a new synth. I see no reason for you to take this outside the Poly group and try and get a broader support for this idea - maybe even set it up on Kickstarter or similar. You could make a great video of your work already accompished by means of the Poly-mods.

Only thing which could be a problem - I imagine you want to use the poly's filter IC. When the stock of this part dries up you're out of business. I know a lot of SID-station project which are now pretty much at a dead end.

Re: [korgpolyex] Re: New CPU - idea

2014-10-11 by gordon@...

On Sat, Oct 11, 2014 at 02:19:20AM -0700, ultragalore@... [korgpolyex] wrote:
>
> Only thing which could be a problem - I imagine you want to use the poly's filter IC. When the stock of this part dries up you're out of business. I know a lot of SID-station project which are now pretty much at a dead end.

It's a pretty simple 24dB/octave filter. It wouldn't be hard to build a replacement for it.

Of course, by then you're close to just sticking new guts into a Poly 800 case ;-)

--
Gordonjcp MM0YEQ

Re: New CPU - idea

2014-10-11 by tim.tashpulatov@...

Mike,

This is going to be really exciting. More performance = more possibilities, just look at PreenFM project based also on STM32 chip able to do some 4+ operators FM.

I have been working with Atmel and ST microcontrollers for a while, so would be happy to be of any help. (And a bit of FPGA, too, if you decide to replace oscillator chips with anything more up to date :))

BTW there could be at least one concern, with new CPU replacement being too fast to drive old peripherals (I doubt the rest of Poly 800 could be clocked higher than 2MHz; the caveat I have discovered once trying to mate 72MHz STM32 micro and some retro IC that would not run faster than 4MHz)

Regards
Tim

Re: New CPU - idea

2014-10-12 by bperkins211@...

I like the idea of re-voicing the poly.  But how much would this cost?

I wish we had 8 VCF's to use like a DW8000, that bottleneck of the Poly800 usually causes me to treat it as a monosynth. 

If you eliminate the voice chip, couldnt you just use the CPU replacement to also do the work of a new TG and patch it's audio out into the VCA's and VCF on the filter chip?

Re: [korgpolyex] Re: New CPU - idea

2014-10-13 by Michael Hawkins

Yes, I think a CPU replacement would also be good as a TG replacement.

The VCF is a problem. It would be good to figure out a solution.

/Mike


On Sunday, October 12, 2014 11:22 AM, "bperkins211@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
I like the idea of re-voicing the poly.  But how much would this cost?

I wish we had 8 VCF's to use like a DW8000, that bottleneck of the Poly800 usually causes me to treat it as a monosynth. 

If you eliminate the voice chip, couldnt you just use the CPU replacement to also do the work of a new TG and patch it's audio out into the VCA's and VCF on the filter chip?


Re: [korgpolyex] Re: New CPU - idea

2014-10-13 by gordon@...

On Sun, Oct 12, 2014 at 06:22:55PM -0700, Michael Hawkins korgpolyex800@... [korgpolyex] wrote:
> Yes, I think a CPU replacement would also be good as a TG replacement.
>
> The VCF is a problem. It would be good to figure out a solution.

There's that guy in Germany who does replacement Juno VCA/VCF modules. IIRC they're about 60EUR each. Not especially cheap, but they look like a solid product and they'd be an easy fit.

You might also look at the Shruthi filter boards.

--
Gordonjcp MM0YEQ

Re: [korgpolyex] New CPU - idea

2014-10-13 by John David Duncan

Hi,

I've missed a little of the discussion but I think I've seen three different ideas:

- replace the 8085 CPU
- replace the tone generator
- add a tone generator based on a new processor chip

Of these three ideas, I like adding a tone generator best.

It maintains the value of all the software and hardware work Mike's done so far.
It can be done in a way so that one 8085 code base works both for people who have the new tone generator and people who don't.
It preserves the Poly 800. By this, I don't mean every facet of the Poly 800 is worth preserving. I'm glad to see tape support go out the window. But I would prefer to keep the ability to play the original patches (via some upgrade path) and make the original sounds, rather than simply transform a low-end 1980s synth into a low-end 2010s synth.

JD

Re: [korgpolyex] New CPU - idea

2014-10-13 by Michael Hawkins

I think most of us would agree with everything you said.

The problem is that the easiest way to upgrade the tone generator would be to remove the existing MSM5232 and simply replace it outright with the new one.

I am reasonably sure that we can duplicate the exact behavior of the MSM5232 by the replacement 40 pin module EXCEPT that the envelope generator behavior will be 100% digital as opposed to the existing half digital, half analog control.

The way it works now is the 80C85 does all of the 8 bit envelope generator math but then outputs the EG level into an 8 bit digital to analog converter because the MSM5232 has VCA inputs for each voice level.

In the digital replacement, we're going to have the 80C85 send the EG level to the drop in replacement TG via digital write messages. And the new TG will then do its own 32 bit digital math to produce the appropriate voice output levels.

I don't think there's a big distinction to be made with the above. They're essentially the same. But I look forward to hearing other opinions.

The new TG will have to emulate the existing 16', 8', 4', 2' feet outputs for the 8 DCO's. When it switches to enhanced mode those outputs will EACH be a voice. The existing MSM5232 does mixing of the digitial voice outputs internally and then there are two blocks of resistor ladders that determine square and triangle waveforms. In the new device, we will duplicate that behavior but in advanced mode the outputs will be analog waveforms for each voice.

And an interesting side feature will be that the analog inputs that were used for each DCO envelope generator level might be able to be retooled to provide modulation control of the oscillators.

It really opens up many new possibilities because by using a 32 bit processor for tone generation we can probably offload a ton of things away from the 80C85 and let the new TG do it. This will dramatically alter the capabilities of the synth.

As I've said before already, I am going to need help! Volunteers please.

/Mike






On Monday, October 13, 2014 10:55 AM, "John David Duncan john.david.duncan@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
Hi,

I've missed a little of the discussion but I think I've seen three different ideas:

- replace the 8085 CPU
- replace the tone generator
- add a tone generator based on a new processor chip

Of these three ideas, I like adding a tone generator best.

It maintains the value of all the software and hardware work Mike's done so far.
It can be done in a way so that one 8085 code base works both for people who have the new tone generator and people who don't.
It preserves the Poly 800. By this, I don't mean every facet of the Poly 800 is worth preserving. I'm glad to see tape support go out the window. But I would prefer to keep the ability to play the original patches (via some upgrade path) and make the original sounds, rather than simply transform a low-end 1980s synth into a low-end 2010s synth.

JD



Re: New CPU - idea

2014-10-14 by bperkins211@...

It sounds like a fun project to see come to life, problem is the popularity of any TG running all voices into the one VCF. 
I know I sound like a glitched sample loop saying that.  It's just that the project may not have enough interest to run a set of boards.
A 6+ Osc stacked SuperSaw mono patch would be awesome though.. all of them slightly detuned.  just like a JP8000...

I take it this would all be SMD?  Just order one or two large boards and cut the modules out?
Any idea how much this new TG would cost?

What do you need help with in coding?  Taking assy. code and converting it to C?
Mostly changing format, matching declarations to functions and what not?
I might be able to help with that, I've dabbled in LUA scripting and arduinos a bit.
I assume most of the math/formulas are in the assy. code, it just needs rewritten in C format and match the new declarations you'd probably come up with.

I'd be interested in helping to learn and get some experience with C

~Blaine

Re: [korgpolyex] Re: New CPU - idea

2014-10-14 by Michael Hawkins

Well, I've thought about the VCF issue and all I can say is that I *think* the STM32 would be able to do its own digital filters on the voices. I don't know this for certain and I am not a math whiz with digital filters (actually I've never created a digital audio filter in code and wouldn't even know where to start). If anyone has a link or ten to help me write an audio digital filter in C, I am all ears. But, even if we keep the VCF as is, we shall be able to produce eight totally independent voices with a long list of waveforms all independently clocked for separate modulation goodness. That can't be a bad thing! My own opinion is that it will be the finishing touch on Poly that will transform it into a "real" synth.

Yes, the module will be all SMC so that we pluck out the MSM5232 and replace it with a 40 pin DIP board that has everything it needs on it.

There will be a need to alter the HAWK code to push off a ton of the functions to the new TG. That's more work to be done on the HAWK assembly.

And there will be a need to develop the tone generation code in the STM32 using C. A key point here is that the TG will generate the waveforms for all eight voices, will need to generate its own envelopes and will need to generate its own oscillator modulation and finally, will need to emulate the MSM5232 perfectly.

There's plenty of work to go around for everyone!

But first, I have to make a decision to pull the trigger on developing the board. Cost to develop the hardware is about $500 USD. I've decided to farm out the work to someone who has done work for me in the past and does a great job of it. I make ask for a donation drive at some point. But I am still mulling over the decision to proceed or not. I have a workload in my business that is shocking to alot of people. 65 hours working per week are common for me.

/Mike




On Tuesday, October 14, 2014 2:23 PM, "bperkins211@... [korgpolyex]" <korgpolyex@yahoogroups.com> wrote:


 
It sounds like a fun project to see come to life, problem is the popularity of any TG running all voices into the one VCF. 
I know I sound like a glitched sample loop saying that.  It's just that the project may not have enough interest to run a set of boards.
A 6+ Osc stacked SuperSaw mono patch would be awesome though.. all of them slightly detuned.  just like a JP8000...

I take it this would all be SMD?  Just order one or two large boards and cut the modules out?
Any idea how much this new TG would cost?

What do you need help with in coding?  Taking assy. code and converting it to C?
Mostly changing format, matching declarations to functions and what not?
I might be able to help with that, I've dabbled in LUA scripting and arduinos a bit.
I assume most of the math/formulas are in the assy. code, it just needs rewritten in C format and match the new declarations you'd probably come up with.

I'd be interested in helping to learn and get some experience with C

~Blaine


Re: [korgpolyex] New CPU - idea

2014-10-15 by John David Duncan

> Well, I've thought about the VCF issue and all I can say is that I *think* the STM32 would be able to do its own digital filters on the voices. I don't know this for certain and I am not a math whiz with digital filters (actually I've never created a digital audio filter in code and wouldn't even know where to start). If anyone has a link or ten to help me write an audio digital filter in C, I am all ears. But, even if we keep the VCF as is, we shall be able to produce eight totally independent voices with a long list of waveforms all independently clocked for separate modulation goodness. That can't be a bad thing! My own opinion is that it will be the finishing touch on Poly that will transform it into a "real" synth.


I also don't know a whole lot about DSP, but I'm guessing you can write DSP code the obvious way (with floating point math), or you could write it in a less-obvious optimized way (with integer math). (By "obvious" I only mean obvious to someone who's taken a college course in DSP, which I haven't). And that the performance between these two is pretty large. As an example, all iPad DAWs run on the same hardware, and some can only record 8 tracks, but Auria can record 40 tracks. I assume that Auria has highly optimized integer DSP code and this makes the difference.

So, I have a Nord Micro Modular with a single Motorola DSP56303, 24-bit, 100 Mhz chip. The STM32 is a 32-bit 72 MHz chip, so I can talk about the Nord and we can keep the two clock rates in mind.

As far as the patches you can create, the actual Nord Modular is amazing. But running these patches on the Micro Modular, the polyphony is quite limited. A lot of patches only give you 1 or 2 voices. An oscillator + a multi-stage envelope uses 11.6% CPU, so you can get 8 voices. If you add a static 12DB filter, you get to 13.4%, so you're down to 7 voices. A good 24DB filter with a modulation input gets you to 18.2%, or only 5 voices.

My takeaway from this is, yes, you can improve the Poly 800 by adding an STM32, but it might be a better idea to add *two* STM32s.

And going back to the software issue: I think Nord also has highly optimized integer code and they consider it a competitive asset. So that would be two STM32s with some high-performant software.

JD

Re: [korgpolyex] Re: New CPU - idea

2014-10-15 by gordon@...

On Tue, Oct 14, 2014 at 02:49:10PM -0700, Michael Hawkins korgpolyex800@... [korgpolyex] wrote:
> Well, I've thought about the VCF issue and all I can say is that I *think* the STM32 would be able to do its own digital filters on the voices. I don't know this for certain and I am not a math whiz with digital filters (actually I've never created a digital audio filter in code and wouldn't even know where to start). If anyone has a link or ten to help me write an audio digital filter in C, I am all ears. But, even if we keep the VCF as is, we shall be able to produce eight totally independent voices with a long list of waveforms all independently clocked for separate modulation goodness. That can't be a bad thing! My own opinion is that it will be the finishing touch on Poly that will transform it into a "real" synth.
>

The STM32 should handle eight voices with filters. It's got a fairly decent ALU which can cope with signed multiply and divide. I've implemented a single 12dB/octave biquad on an AVR8 which is trickier since you only have unsigned multiply and no divide (obviously you've got signed multiply and division in C, but it's slow compared to unsigned mult which compiles to a single instruction).

This is my automatic bassline generator (Amen break played back and filtered sawtooth generated in realtime at 32kHz)
http://gjcp.net/~gordonjcp/mp3s/acid6.ogg

The code for it is floating around in here
https://github.com/gordonjcp/gyoza

and the quick and easy guide to biquad filters that you may find helpful is here
http://www.musicdsp.org/files/Audio-EQ-Cookbook.txt

> And there will be a need to develop the tone generation code in the STM32 using C. A key point here is that the TG will generate the waveforms for all eight voices, will need to generate its own envelopes and will need to generate its own oscillator modulation and finally, will need to emulate the MSM5232 perfectly.

MSM5232 emulations already exist in MESS/MAME

Can't help but think it will so radically alter the character of the Poly800 as to render it unrecognisable, though. Wouldn't this almost be better as a separate standalone module?

--
Gordonjcp MM0YEQ