• Please review our updated Terms and Rules here

Announcing New VDM-1 Reproduction

P

physicsrob

New Member
Joined
May 20, 2021
Messages
3
Hi All,


It's with great pleasure that I'm announcing a new S-100 board, a VDM-1 reproduction which I've been referring to as the VDM-2021. Now anyone can build a terminal-style text video card for the S-100 bus.

Details of the board, ordering instructions, and much more can be found here: https://physicsrob.github.io/vdm1/

The short history of this board is that I reached out to Lee Felsenstein to ask him what he thought of me reproducing the VDM-1. He was reluctant for me to reproduce the original PCB layout since the original layout had problems (like the ribbon cable), so he suggested I create a new layout for the original VDM-1 circuit. So that's what I did! Lee provided valuable feedback throughout the process.

Initially I will be selling the bare board for $30, or a partial kit for $90 which will include everything necessary to build a VDM-2021 that is not readily available through Jameco or similar electronics distributors. Specifically the partial kit will include:
- 1 x VDM-2021 PCB
- 1 x Character generator ROM (either MCM6574P or MCM66740)
- 8 x 2102 memory chips (equivalent to the 91L02 or 21L02)
- 1 x DM8131 bus comparator
- 1 x DS8836N bus receiver
- 1 x 93L16 binary counter
- 1 x heat sink
- 1 x BNC PCB connector


All remaining ICs are 74-series TTL or 4000-series CMOS.

This is my first S-100 board, and I'm doing this as a hobby, so a few disclaimers:
1) I haven't put in the order yet for the final production PCBs (I've had prototypes printed and I have assembled them successfully). I will most likely place the order in about two weeks, after gathering the initial orders, and depending on the response.
2) The hardest component to obtain is the character generator ROMs. I have about 50 ROMs available for these kits, but I have no idea yet if that's sufficient or way overkill. I believe I have a reliable source for more ROMs if necessary.
3) This is a hobby project, and I have a full time job, so shipping may be delayed, and while I will attempt to support builders, I can't make any guarantees of support.
4) While this board should be straightforward to assemble, this is not a beginner board. There are 48 ICs, and I highly recommend having a scope on hand to follow the integrated assembly-test procedure outlined in the original VDM-1 manual.

I hope you all find this valuable and enjoy! I've had a blast putting this project together, and it's very satisfying to use BASIC on my Altair displaying the output through a vintage monitor directly from this board. I'm excited to play Trek-80 as well, although I haven't gotten the keyboard working with Trek-80 yet.

Cheers,
Rob
 
Nice!

In my own personal project queue is goal to make my own S-100 video card with similar-ish capabilities, but I was going to do it from scratch with more modern parts. (The target to be to be roughly compatible with the PolyMorphic Systems VTI or SSM VB1B, because they're a little simpler and support those TRS-80-style graphics.) But it's great there's a more authentic board out there now.
 
That ribbon cable is not an issue, if it simply gets replaces with the usual small diameter IC wire wrap style wire. In my VDM-1 I threaded these wires under the long axis of the central 6 IC's on the lower edge of the board near the connector, so they are out of the way, held down to the board surface by the IC bodies, and barely even noticeable. But a board without these would be better. When I replicated the Dazzler boards I did a similar thing as it also had a surface cable.

I tried the VDM-1 board in my Sol -20 to test and repair it and it runs concurrently with the other video circuitry for testing but the scrolling doesn't work (perhaps relating to the problem with the Flip flop you mentioned), but it works well enough to test it, as we know the Sol-20 has pretty well the same circuit on its motherboard.

There is an issue with this circuit design (minor) in the part that generates the vertical sync, the sync pulse is too wide, and on many video monitors, this upsets their sync separator causing a tear in the picture at the top as the bias conditions in the sync separator are altered by charge storage on its input coupling network, and this upsets the separated H sync, causing the tearing. I'm pretty sure this was the case on the SOL and the VDM-1. There is a thread on this on the VCF, I'll try to find it, it is just a matter of altering one resistor value, from 100k to 27k, to fix it...here is the thread:

https://www.vcfed.org/forum/forum/te...e-non-standard

Watch out though, replicating vintage computer pcb's gets to be highly addictive its so much fun !

I just looked at the githib info, great.

Can you explain this remark in more detail, since the original board must have had a functional scroll ? did they all have an issue or just some pcb versions ?
  • The blanking circuit for the scrolling functionality needs to be connected to the inverted output of a flip-flop rather than the non-inverted output.
 
Last edited:
Nice work! Sounds like you solved most of the basic issues with this card! Just submitted my order :)
 
Last edited:
Hi Hugo,

Sorry for the late reply. I need to follow this thread more closely!
Can you explain this remark in more detail, since the original board must have had a functional scroll ? did they all have an issue or just some pcb versions ?
> The blanking circuit for the scrolling functionality needs to be connected to the inverted output of a flip-flop rather than the non-inverted output.
Click to expand...

When replicating the revision E schematic I encountered a few problems, some of them were bugs I introduced, but after ironing out every bug I introduced there were still problems with the board. As best as I can reconstruct these bugs reflected genuine mistakes in the schematic. For instance the xtal oscillator frequency was changed between revisions, but the schematic did not reflect the correct divisor for the new xtal (to be specific -- part of the divisor had changed, but part hadn't). The only logical conclusion is that Processor Tech didn't always do the best job at keeping the schematic and PCB in perfect sync.

I very much doubt that any of the problems I came across were actual problems with the PCB, they were probably just problems with the documentation.

As to your precise question, this is what I said to Lee Felsenstein:
The scroll circuit blanking is inverted. Specifically the lines which control when IC26 and 27 count changed fairly significantly between revisions. There's a flip flop (IC13) which changes state when the first line of text is to be drawn. The blanking from this flip flop was inverted. Easily fixed.
Click to expand...

And this is the relevant portion of his reply:
I had forgotten that the schematic had evolved a little bit - I now recall an emergency wire fix we had to provide for the first revision of boards - IIRC it involved creating a latch circuit out of two gates for come kind special blanking condition - that was probably what you found.
Click to expand...

Sorry I don't have a more clear response!

BTW, What is the specific scrolling behavior you see? I've debugged quite a few problems so I might be able to point you in the right direction.
 
physicsrob said:
BTW, What is the specific scrolling behavior you see? I've debugged quite a few problems so I might be able to point you in the right direction.
Click to expand...

Hi,

When I have the SOL running again, I will re-fit the VDM-1 card and document what happens. As I recall it worked fine, except that the scrolling was abnormal, otherwise the output from the VDM-1 mirrored what was seen from the video output of the Sol. But of course a VDM-1 was never intended to be run in the Sol, so I had thought it was some sort of bus incompatibility issue or conflict between the two systems trying to run concurrently. The only reason I put the VDM-1 into the Sol, was to check if the board worked. The Sol is the only S-100 bus computer that I own to try the card in.
 
Happened to come across this thread, belatedly, in researching the VDM. I acquired a Vector 1 that contained the TDL SMB and ZPU boards, a VDM, and a few memory. I'm trying to guess what the setup on this particular computer must've been, 45 years ago. Maybe a Teletype was connected into the SMB's serial port, as a keyboard? There would be no need for a terminal if you already had a video card...
 
I like the way you stuck to the basic original VDM -1 design with the vintage chips, you have my vote.

While it is possible to improve the design with more economical modern IC's (and programmable devices to cut back the chip count) as some have done with various vintage S-100 board reproductions, I think it can open a "can of worms" with non-standard behavior. Also using these more modern parts, you cannot guarantee that any remarks in the original manual will definitely hold true.

But, the vintage parts suffer none of these issues and add to the "retro-computing style" of the whole thing.
 
It's been a while since the original post about the VDM-2021 and after visiting the web site I'm wondering if the board was ever produced? The ordering web link seems to be working but it seems strange that this post was never updated. I've been looking for a vintage Video Board for my TDL Xitan S-100 system (ZCPU, SMBII, 32k DRAM Card) after finding that the original TDL VDB display card seems to have gone missing :(
 
You must log in or register to reply here.
Back
Top