Amek Supertrue & Virtual Dynamics

[krabbencutter]

Hey,

I'm working with a friend of mine on potentially creating a custom & open replacement board for Amek Virtual Dynamics. Fortunately enough he still has a working Amek desk with Supertrue & Virtual Dynamics. But before I think about putting probes anywhere in there I wanted to ask if anyone here has any deeper insight into those systems. Right now the most interesting part would be the serial communication between Supertrue and the 8051 based MCUs on the Virtual Dynamics boards (80C32). I tried running Supertrue in Dosbox, Dosbox-X as well as on FreeDOS but had no success. I can run the console configuration setup just fine, but when I start Supertrue I only get a grey screen :/

 

[sahib]

Firstly, thank you for the hex for the EPROM that you posted in the library.

Strangely enough I was also planning to work on the same. I have a full set of boards.

However I am in the process of moving office and I will not be able to start before mid-end of May. If you can wait we can run both projects in parallel.

 

[krabbencutter]

Awesome! This project is definitely going to take a while and I'm happy to collaborate
Is the PCB on the left the one that is connected to your PC? If so, can you maybe upload another image of it?
What's on the upside-down PCBs? I have a spare 1147 VD board from an Angela 2 console for analysis and it doesn't feature those upside down PCBs.
And why aren't there stickers on your eproms? The opening in the middle of the chips is used to erase the data with UV light.

 

[sahib]

I will have answers to your questions once I put the patient on the operating table. But indeed the EPROMS have no version label on them and I am wondering if they are actually blank. I'll ask Graham about that. The picture is just some of the stuff in the box that was shelved since I got them. The upside down boards should be the the VCA cards. However, these boards belong to the stand-alone version of the VD that was proposed at some point. I have all the documentation too but I have not gone through them yet. I will do as soon as I settle in the new office.

Where did you get the hex file?

 

[krabbencutter]

Exciting stuff!
I read out the eproms myself with an arduino. The arduino is connected to a CD4040 and will step through all the address registers of the eprom and read the data. I then use putty to save the serial binary data to a file. Last but not least I use VScode to trim the data and convert it to hex with this extension: https://marketplace.visualstudio.com/items?itemName=attilathedud.data-converter
Even though it's a makeshift solution it works very reliable. I should also be able to rewire & rewrite it to burn data to the eprom.

 

[sahib]

Welldone with reading the EPROM.

The micro on the boards I have is Intel 80C321 which I would imagine would be read protected. But since this is a development board with a bit of luck they may be unprotected.

Can you post a photograph of the board you have?

 

[krabbencutter]

Sure, I'll post an image later. But the board is mostly identical to yours, just without the "upside down" VCAs.
The 80C32 does not have internal ROM and all the ROM-Data is in the 27C64.
Edit: There's a 9098i service manual in the Amek thread and the Virtual Dynamics circuitry on page 8 is (afaik) consistent across all Amek desks: https://groupdiy.com/threads/amek.44846/
Edit: image attached

 

[sahib]

Indeed. 541 has the internal ROM.

I have the 9098i manual. I also have the full schematic independently.

Graham said the board was fully working at the time. He also said there have been a few tweaks made to the EPROM firmware. So, I am guessing that the whole thing was shelved and they forgot to seal the EPROMs. I am not sure how many times the box was opened. Hope not too many, particularly in bright daylight. I have opened it only once as I had other boards. So, there is the unfortunate possibility that the bit pattern may have been corrupted if not wholly erased. But we'll know once I have it laid on the table. I will of course open the box under low light and seal the ROMs immediately

 

[sahib]

I have dug out some pictures to try to see what's in the box.

Now, the picture I posted above has the processor board AM1147B which is what you have, less the additional VCA boards. I have the schematic for this board.

The processor board in 9098i manual is HM787E. I have a blank earlier version HM787D. The QC label signs it off as passed. So, I would imagine when populated it should work.



The small board on my first picture connected to AM1147B processor board is the Serial Interface and Solo card TK 1433B. I also have the schematic for this board.

 

[krabbencutter]

Indeed. 541 has the internal ROM.

I have the 9098i manual. I also have the full schematic independently.

Graham said the board was fully working at the time. He also said there have been a few tweaks made to the EPROM firmware. So, I am guessing that the whole thing was shelved and they forgot to seal the EPROMs. I am not sure how many times the box was opened. Hope not too many, particularly in bright daylight. I have opened it only once as I had other boards. So, there is the unfortunate possibility that the bit pattern may have been corrupted if not wholly erased. But we'll know once I have it laid on the table. I will of course open the box under low light and seal the ROMs immediately

If you read out all four eproms we should be able to piece the data together. Visual Studio Code has a very useful comparison feature that will automatically highlight differences between two files. And even though I'm a bit paranoid about UV erasable eproms, you need strong amounts of UV for extended periods of time. It's also unlikely that all four eproms would lose data in the same registers.

Thank you for digging out the additional images I haven't compared the schematics in detail but the AM1147 & HM787 boards should be mostly identical when it comes to the VD control circuitry. I also found schematics for the 1433 and the serial interface doesn't look too complicated. I will try to source the ICs so that I can build one myself.

 

[sahib]

Thank you too. I'll be on this as soon as my bum hits the seat in the new office.

 

[Matt Syson]

Handy hint: Connecting plus 18 Volts to the 5 Volt logic supply terminal makes the logic chips catch fire!
VD interfaces were 1040 and 1060 boards as well as the later 1433 which has the extra VCA so that PFL/AFL will 'follow' channels that are soloed or PFL/AFL. VD also worked with an Atari so the data sent was a TTL level version of midi. I THINK the 787 boards were the same EPROM version for ATARI and 'IBM' automation, or at least I don't remember changimg VD eproms ever ewhen converting ATARI Supertrue desks to IBM (Mozart, Hendrix and Big desks).

 

[krabbencutter]

Firstly, thank you for the hex for the EPROM that you posted in the library.

Strangely enough I was also planning to work on the same. I have a full set of boards.

However I am in the process of moving office and I will not be able to start before mid-end of May. If you can wait we can run both projects in parallel.

View attachment 127464

How's the new office coming along? I've passed my final university exam last week and am ready to dig into the virtual dynamics

 

[sahib]

Almost complete but I need at least another week to become operational.
Congratulations on your exams.

 

[Tubetec]

They had an Angela II in a place I worked but we never used the virtual dynamics ,
it had the VCA group thumbwheel in each channel ,
You easily A-B the inputs between monitor and main mix path with the flip switch ,
suffice to say VCA path sounded inferior in every respect .
In the end we got a bunch of non VCA mic input modules from an Angela mk I which had a lot of extra edge .

Its an exciting development integrating the virtual dynamics concept into the modern DAW environment ,
maybe the balistics of vintage compression gear could be virtualised .

 

[krabbencutter]

They had an Angela II in a place I worked but we never used the virtual dynamics ,
it had the VCA group thumbwheel in each channel ,
You easily A-B the inputs between monitor and main mix path with the flip switch ,
suffice to say VCA path sounded inferior in every respect .
In the end we got a bunch of non VCA mic input modules from an Angela mk I which had a lot of extra edge .

Its an exciting development integrating the virtual dynamics concept into the modern DAW environment ,
maybe the balistics of vintage compression gear could be virtualised .

I've read multiple times that the VCA circuitry degraded the sound on the older Amek desks. Maybe Matt has some insight what the issue was?
And don't get your hopes up about any kind of DAW integration I expect this project to be quite complex and resource intensive.

 

[Matt Syson]

Since the many different AMEK desks effectively used the same circuit 'blocks' it is largely down to a vociferous number of AMEK haters that happily vent their perceived dissatisfaction. (failing musicians trying to 'blame' someone/something)? Angela the originals are nothing like Angela 2 which is effectively an Einstein 'built properly'.

 

[Tubetec]

It was an Angela 1 we had actually , but as I said it had VCA's throughout until we noticed the ALT path with sounded better , ie the monitor mix section when tracking .
You can can take my word for it or not , but the non VCA mic input modules sounded better .
I cant comment on the dynamics as we never used it with a computer.
The fader grouping and mute always worked reliably .

 

[gyraf]

yeah, I hear ya: We got a Angela-1 on vca's in our B-studio for years, and there was definitely something wrong with its basic sound. I didn't think to suspect the DBX2150 VCA's at the time, as our M2500 on AlisonResearch TA-101 VCA's sounded just fine, and that was much older technology..

 

[Matt Syson]

There are many ways to 'mess up' the performance of a VCA as the control port is the same sensitivity as the audio port and reading the datasheets shows the pitfalls which many fail to get to grips with. Modulation of the control port from 'other' sources will be significant factor so the exact placement of circuit traces will make some difference. The distortion 'trim' needs to have stable reference which may or may not happen, and then they are temperature sensitive so that SHOULD be compensated. Calrec used the tA101 'blocks and put work into optimising them as did many others. You also have to quantify 'sounding better' as well as finding out WHY. The presense or absense of the vCA may not be the root cause of difference'