SWTPc TV Typewriter II

SHARE	Return to Threads
SWTPc TV Typewriter II	by Bill Degnan - 07/29/2009 09:16
The original TV Typewriter on display in the MARCH museum was donated by Bill Dudley. I don't have any pictures at the moment. Bill showed up on Wednesday and we took the thing apart to see if it could be made functional again. Here are some project notes for future reference, I will come back to this post asap. 1. The keyboard attached to the unit is a George Risk. See same keyboard model I have the tech doc for the keyboard, but this was not necessary, all of the voltages checked out OK. The keyboard voltages and connections are good. We checked the repeat key mod Bill had made. We monitored the voltage coming into the TV Typerwriter from the keyboard. Each time a key was pressed, the voltage jumped 15 millivolts (from 80 to 95). We did not go so far as to verify that when you press an "a" you get an "a" sent by the keyboard. When powered on and attached to a display the TVTW dumps the jibberish contents of RAM to the screen, and presents a blinking cursor. We were unable to get keyboard keystokes to output to the screen however. No response generated at all. The TV Typerwriter has a baud rate adjustment knob which we used to attempt to get positive results at various bauds, but no improvement. Using the keyboard we can occasionally clear the screen. We tured to the TV Typewriter itself to trace key pulses to find a fault in the wiring or power. The -5V( neg or pos.?) sources were registering -4.72V Bill thought that this was NOT a problem. I suggest checking that however against the chips tolerances. Next we can 1) try hooking the keyboard up to another display terminal using a null modem cable to verify that it's working correctly (can send correct chars to screen?) 2) using the schematic/book check the voltages until we find problem chips or ICs, replace and see if that fixes the problem. 3) I believe we found no faults, but re-check power supply. 3) Make sure RAM is OK, test voltages coming into RAM, 1488, UART, etc. Reply
TV Typewriter II Restoration Update	by Bill Degnan - 08/08/2009 20:11
Hi Bill, Evan and Bill Degnan asked me to take a look at your TV Typewriter II, as they know I like fooling with old electronics stuff. Whoever built it really did a great job on keeping things nice and neat, even the mods. I'm impressed. The TV Typerwriter II in MARCH's museum. Click image for larger version. When I first turned it on, I wasn't getting a clock to the UART, which also kept the character from getting sent to the display RAM. The keybd strobe was fine. I removed a "rate multiplier" chip that adjusts the clk to get the different baud rates, and then plugged it back in. Now I had a clk to the UART. That got me characters on the screen. Then I tried the cursor keys and noticed that "right" was moving the cursor maybe 10-20 positions. I found a one-shot on the cursor bd that wasn't working. Replaced that chip and the cursors started to do the expected thing. Then I did a little work with the ohmmeter to figure out what pins on the RS-232 cable were tx and rx, and made an adapter for the PC's 9-pin connector. Now I can run a terminal program and send and receive chars from TVT II. I had to change the PC to 7N1 instead of 8N1 to get it to work. Your unit appears to have these mods: 1) 64 columns instead of 32. 2) Extra RAM chip added to store the 7th bit. 3) Extra character generator ROM chip added to support lower case. 4) Speaker for control-G beeping. (works) I am not sure what the circuitry on the extra board that the speaker is hooked to is for. Maybe that's used for some of the above mods? Evan mentioned something about auto-repeat for the keyboard. Not sure where that is done. One strange thing I have noticed is that the bit map for lower case j looks wrong. The character ROM is supposed to have one blank row per character. When it's time to blank the display (vertical or hor blanking) the circuit selects the blank row, and the display is blank, even though the display RAM is being accessed during the blanking time. The lower case j seems to have a bit set in the row that is supposed to be all 0's. As a result, if there are j's in the display memory, they show up as vertical lines in the part of the screen above and below the 16x64 normal character display area. Does this sound familiar? I was wondering if you had this problem before, or if maybe the character generator ROM has gotten a little corrupted from age. All of the other characters look fine and have correct blanking. Best wishes, Bob Grieb --------------------------------------- P.S. To fix the lower case j blanking issue, one would copy both 2513 ROM's (all of 64 x 8 x 5bits!) and program a new CMOS EPROM with both images, after fixing the j pattern of course. Any EPROM with at least 1K bytes would be big enough. Something like a 27C64 or 27C32 would work. I would attempt it, but the pinout of the 2513 is totally non-standard and it takes 3 supply voltages, so I don't think my EPROM programmer could read it. It works fine as-is, just has some extra pixels in the blanked area of the screen if there are lower case j's on the screen. Also, the lower case j looks weird: * * * * * * * ** * and it should be: * * * * * * * * * * * Bob Reply