• Please review our updated Terms and Rules here

Homebrew TVT I picked up

I suppose it's worth a try. But my gut feeling is that if it was that great I would've heard about it already. I wonder if the guy on Facebook has stock in the company.
 
Just for kicks, I plugged the molex header into the appropriate spot on Corsham's SS50 serial card. I then plugged my serial breakout box into the DB9 on same, and then watched. Initially nothing happened, and then after a while the RX light lit up. Didn't flash as I typed or anything though.

Also, while I was probing around, something happened and the character formation became momentarily close to what it should be. But then it went out again.

Also I notice the 6571 is getting fairly warm.. is that normal?
 
Well, I had something lock up on my scope and now it won't power on at all, but in the meantime, in spite of myself, I managed to get this onscreen:

https://drive.google.com/file/d/14PcoNcPn7XBwpKSoq47g4yu0DEQ3HqTD/view?usp=sharing

This is what it looks like live:

https://www.youtube.com/watch?v=36Td__1Buoo

Bit wavy, but readable at least. The screen has expanded out now so I cannot see the cursor. Not really sure what I did - first I cleaned some really nasty sockets with a small drill bit (all I have at the moment). Flipped the board upside down to take a picture of the wiring, flipped it back. Then I started removing the RAM one by one, noticing that as I did, the characters became more readable. When I removed all of it, I was presented with a nearly perfect series of blocks on screen, except for the top.

I also removed certain ICs and powered up, just to see what the result was. I noticed that one 7493 once removed left nothing onscreen, just the cursor in the upper left.

I have no action from the device though. Typing produces nothing onscreen. That said, I'm wondering if it expects to be connected to something that would echo back. Anyway, progress is progress..
 
Yeah going to replace those. Trying to figure out what the two slim black ones are.

Also, I think I've solved the mystery of what this was used for. I had a look through the seller's previous listings and discovered this: https://www.ebay.com/itm/SWTPc-Comp...=true&nordt=true&rt=nc&_trksid=p2047675.l2557

So I emailed her and asked if they were from the same place, answer was yes. Apparently the gentleman (who passed away... sigh) had about 8 sheds worth of stuff and these were part of it. I think it's probably too much of a coincidence that these both came from the same place, and that this 'TVT' has a SWTPC style connector on it. I've been looking at the photos of the cards installed in the machine to try to glean anything I can about how it might have worked, like if it just used the MP-S, or if it was rigged up to control the TVT in a unique way. It's a later revision 6800 with built in ROMs on the CPU board. The MP-S card looks like some components were omitted.

I thought about buying that machine, but was unaware of its connection to the TVT and I already had two of them... didn't really need a third. I was only tempted by its MP-L card, which I don't have.
 
So I've decided with the video (mostly) working, I'd take a look and see what is going on when keys are pressed. I did some tracing of the actual wires. Here is what I can figure out so far.

When you hit a key, it appears to set certain bits up, which go to the first EPROM which is on a board by itself, which in turn go out the other side of that EPROM. The other side of EPROM 1 goes to a ribbon cable that goes to the board the AY-5-1013 chip is on. I've traced several of the wires from that ribbon cable to the DB1-DB8 pins of the UART. I have also traced the receiver data bits RD1-RD8 to the address bits A1-A7 of the second EPROM. The data out pins of that second EPROM -- pin 4 goes to a 7404, the remainder all have resistors on them that are tied to ground. I'm not really sure what that does, but I'm guessing the EPROM is there to translate something being received into something the TVT can understand?

When I hit a key, I can see DB1-DB8 changing depending on which key is pressed. If I understand the block diagram here, the AY-5-1013 would be looking for a strobe signal, and upon finding one, would send the data represented by DB1-DB8 to the serial output pin. I wouldn't be seeing any activity on EPROM 2 unless the TVT were receiving something from 'abroad'.

I don't see anything happening on the SO line.. but IIRC correctly my probe won't be fast enough. I'm working on getting the scope connected to it. But I'm thinking I'd want to verify data strobe is being triggered (pin 23) and/or control strobe (pin 34), and then follow SO (pin 25) to wherever it's going (I'd guess the MC1488 or such?) to see if anything is going out.

Further, I feel like I wouldn't be seeing anything onscreen, because typing is sending data out, but it needs to be echoed back to the TVT by whatever device it'd be connected to.

Sound legit so far?
 
Last edited:
I think so. Let's look at what you should be seeing on UART.

Pin 40 should have a clock signal on it that's 16 times the "baud" rate. So, if the board was set up to do 1200 bps, the frequency on pin 40 should be 16*1200 or 19.2KHz. This signal should be present all of the time.

Pins 26-33 should have the keyboard output data on them.

Pin 23 should go low with every keypress.

Data will be clocked out to pin 25 (SO).

Also, pin 1 should have +5
Pin 2 should have -12.
 
Thanks Chuck.

I'm not picking up a clock signal on pin 40.. either on the scope or the logic probe.

Pin 40 is tied to pin 17 of the same UART IC (RCP), and then that is tied to pin 4 of a 7404 on that same board. I'm guessing the 7404 generates or relays the clock signal?
 
I think your external guess is correct. Pin 10 goes to a little wire header. A black wire from that header runs to a pin on the socket for the lower card. That pin is unused on the card itself, it just serves as a point to connect the little black wire to a thicker wire that runs to the SWTPC-style Molex header. If it is in fact to SWTPC specs, the actual pin it ends up at is Clock In or Clock Out (on the connector they are tied together, so I cannot say which wire exactly with continuity testing.. I seem to recall SWTPC saying you need to do that if you do not have a self clocking tape system).

So I'm guessing it picks up the clock from the MP-S board in a 6800, assuming this was used with a 6800..
 
Now it is making a little sense. It is intended that the remote machine echos to the part that does the video. What you'd call full duplex. You need to make a loop back to see what is being typed, unless you have a computer on the other end.
While it is possible that it may be translating the keyboard to something like EPCDIC. It is more likely that the EPROM is to add parity. Most keyboards do not have parity, odd or even. Of the two odd seems to be more popular.
Dwight
 
What kind of asynchronous communication is remote-clocked?

I'd add a local oscillator. Or at least for now use a signal generator.
 
No idea. But I'm thinking I might be best to get my hands on a spare MP-S card, or make one, and modify it to the specs he has in the photo. I'm making an assumption that the card in that photo is the actual card, of course. This TVT isn't merely for video and keyboard I/O, it appears to have some kind of tape encoding capability, like a built in AC-30. I'm wondering if the reason that MP-S card is modified the way it is is to simulate paper tape handling, in conjunction with the tape circuitry on the TVT.
 
I guess my point is that if you're going to have a remote clock, you might as well be using synchronous data which is faster, and far simpler to implement. But this already has everything you need for RS232, so the I think the best thing (and wouldn't be difficult) would be to add a local oscillator. You could always add a switch to disable it if you ever use something that requires an external clock. In the meanwhile, it'll work on everything else.
 
falter said:
Thanks Dwight. But without a clock signal to drive the UART, how would a loopback work?
Click to expand...

Just make a 555 oscillator. It doesn't need to be particularly high speed. It is just to make proof of concept. Maybe 100KHz for a starter. Loop the Rx and TX, as well as any line controls needed for the UART ( look at data sheets and follow pins ).
Dwight
 
You must log in or register to reply here.
Back
Top