• Please review our updated Terms and Rules here

Homebrew TVT I picked up

falter said:
As usual, I have confused things. Chuck said character generator... I crosswired that in my head to EPROMs. But yeah, when I did the testing I did find good continuity between the battery holder and pin 15 on each of the EPROMs. So I'm not sure what is going on there.
Click to expand...

The character ROM is the MCM6571. Pin 1 should go to the negative lead of the battery. I suspect you saw all the rest of the parts connected across +5V and ground as the current path. The battery should have one lead connected to ground ( measured as a TTL ground pin ) and the other negative lead to pin one of the MCM6571.
I'm not sure why they created a separate supply for the MCM6571 as they need a -9V supply for the 1702s. The 1702s actually need a 14V supply ( 5V plus the -9V is 14V total ). Measure carefully with a low ohms meter.
For the MCM6751, they could have used a voltage divider from the -9V and that would have been enough a 10ma divider would supply the 10ua easily. I'd most likely have used a 1K and a 500 ohm to create the divider. The error would be so small as to not worry.
Dwight
 
I suspect the 1702s are to create a different data for the AY-5-1013. It may be that they'd been using the TVT for something other than an ASCII serial input. I'm not sure what the 1702 on the board at the right does but the two that are close to the AY-5-1013 look to have their data lines connected to the AY-5-1013 ( Falter can you measure these? ). It is possible that the AY-5-1013 has its data line connected to the address lines of the 1702s. Maybe one is used to translate data input while the other is to translate data output. The third one on the separate board is a question still.
So, check the AY-5-1013 data lines. I suspect one set go to one of the 1702s address and to the other 1702s data lines.
Dwight
 
Chuck(G) said:
EBCDIC<->ASCII perhaps? But what of the third 1702?
Click to expand...

That makes sense. He'd need to trace the address and data line of the third one to try and figure out. As I recall the IBM had some funny special displayed characters. Maybe it is an additional character ROM for the things that are not typical ASCII to be displayed.
Dwight
 
That's strange. I have EBCDIC character generator ROMs from the early-mid 70s. They weren't rare--and given that this thing was circa 1979, "rolling your own" from a 2716 was a very practical option. Same for keyboard encoding.

On the other hand, if this thing was EBCDIC, the Moto character generator would be inadequate for the full EBCDIC set, as printable EBCDIC from cent to vertical line (0x4a - 0xfa) is 177; more than the 128 afforded by the Moto character generator.

Check the wordsize jumpering on the UART; if it's 8 data bits, then EBCDIC makes some sense.
 
Last edited:
Ok. Pin 1 to the negative battery terminal resistance is zero. So there does appear to be a link there.
 
Okay, I pulled the 1702s and decided to test the PSU and then reconnect to the computer if nothing went sparky sparky.

And nothing did.. was a total non-event, PSU seemed fine. So I put it all back together with the 'computer' (boy, that was a chore) and then fired up, still with the 1702s out. So far voltages look okay, but definitely not getting anything useful onscreen.. just these weird diagonal lines:

https://drive.google.com/file/d/14GzWIsHYGJcZB28bolabM3jcVulVTfAw/view?usp=sharing

Maybe the EPROMs come into play somewhere here. I don't know. I'm going to finish testing all the voltages for the EPROM power pins, if they line up, might as well give'er. Good chance they're dead anyway.

One question - this thing has one of those BNC style connectors.. I'm assuming the pinout on those is the same - ground on the outside and VID signal on the inner pin?
 
Okay so checking voltages on the 1702 sockets:

Pin 12 - VCC - +5.05V
Pin 15 - VBB - +5.05V
Pin 16 - VGG - -8.98V
Pin 22 - VCC - +4.98V
Pin 23 - VCC - +5.05V
Pin 24 - VDD - -9V

I think that's correct yeah?
 
Yes, that's the way that BNCs are used. Doing it a different way would be nonsense. Those diagonal lines could be straightened out to vertical lines, if you'd adjust the vertical sync. But really, this is a job for a 'scope.

The voltages on the 1702A look to be correct.
 
The connection from EPROM to the UART may be through buffers. Look up the specs on the UART as this part was not software configured, like the 8251 would be. It is configured by connecting the pins to ground or pullups ( in rare case I/O latches ). You can find out things like baud rate, parity and data width with an ohm meter and the spec sheet.
Dwight
 
Thanks Dwight. Yes I did follow some of the wires from EPROM and they lead up to what appear to be buffer ICs above.

I think tonight I will try to see if I can get a stable video display and kind of work from there. Also trying to trace that SWTPC like connector to see if by chance it follows SWTPC convention.
 
Tested the MCM6571A in my digital group box. Seems to form letters, although they are all screwy because it's a 6571A rather than the plain 6571 that machine uses.

Next up I'm going to check the usual bandits - the 74123s of which there are.. ulp.. lots. And then 161s, anything to do with video generation.
 
Okay so, I discovered that the batteries are being drained pretty quickly... they went from -3V to -2V in about 24 hours. I'm assuming that's not normal, and that the way this fellow built this thing, he wouldn't have wanted to be having to remove the batteries every time he shut down? I'm trying to trace the circuit more to figure out where else it might go, or what's draining it.
 
They seem back to normal today.. -3V.

I did a bit of probing around. The crystal appears to go into a 74H04, I'm not sure if that drives the system clock, but I can see a pulse on pin 4 of that chip.

Definitely going to be tricky with this one.. lots of surface rust up top. I'm hoping it hasn't spread deep into the sockets.
 
Progress!

I swapped out some 7473s that didn't seem to be working properly, and now we have:

https://drive.google.com/file/d/14JZjXVwuCVaOLa7LRdKcbdEsTTHWW_ZF/view?usp=sharing

Doesn't look like much, but we can see distinct rows instead of just screen snow, and an attempt at forming characters. And I notice that little line at the bottom is always there through multiple power cycles. I think that may be the cursor?

It's going to be really difficult figuring this thing out further... there is a lot of surface rust on several of the pins, and I do *not* have the skill to remove and replace wirewrap sockets. I can never really be sure if we've got a bad socket or bad IC unfortunately. And then there's the issue of not knowing what this thing was connected to originally, and if it depends on it to operate.
 
I think the basic circuit (absent the EPROMs) was quite likely cribbed from some source. Perhaps you should start your search there. Perhaps some of the variations on Don Lancaster's TVT? The TVT II used 6 2102 SRAMs; was there an upper-lower case later version that used 8?
 
The fact that the screen is stationary would indicate that it is reading uninitialized data. I've not looked at the original schematic but it may not have had a clear of the RAM on power up. Try some returns to see if it moves things down.
Dwight
 
Late to the party...

But it looks distinctly to me as if the horizontal frequency is wrong, or simply the horizontal hold isn't locking.
 
I did try hitting return. Also tried probing the outputs of the serial board to see if anything is getting output when typed, but nothing.

The keyboard is doing something... probing some of the fingers of its edge connector, when keys are pressed the state of the pin changes (ie low or high), although it does not change back once the key is released.
 
You must log in or register to reply here.
Back
Top