Scored a VT100

[segaloco]

Something I've been biding my time on for a while has finally paid off, I've landed a DEC VT100 (pictured).  I really lucked out in that it was pretty discounted since the seller didn't have the original keyboard.  By pure happenstance I already had a VT100 keyboard from an earlier project that fell off (building a VT100 keyboard->USB interface, someone has specs online, I just kinda lost interest.)

Well, since I had the keyboard, a missing keyboard on the terminal was no issue, and now the two are joined and I'll soon start tinkering on it.  My intent is to get a RISC-V SBC I'm currently experimenting with hooked up to it properly and verify it all works, then see if I can embed said SBC inside the casing in a non-invasive way with a switch so I can use this either as a standalone UNIX computer or as a true hardware serial terminal depending on what I'm working on.  This really leaves only two computing displays I'd consider spending money on (save for whenever my LCD monitors die and I need to replace them):

- The Dataspeed 40/2 Terminal - One of the last terminals made by the WECo/Teletype partnership, the Dataspeed 40/2 is part of the Dataspeed 40 family and presents the ANSI asynchronous interface expected in typical time-sharing computing environments.  There are other Dataspeed 40 models, but the 40/2 is the one I'd probably get the most out of.

- The AT&T DMD 5620 Blit Terminal - This was a terminal developed specifically for UNIX, which while offering "dumb terminal" capabilities, also features a graphical bit blit mode and one of the first "desktop environments" for UNIX.  X11 ultimately won out, but this terminal is an integral part of the UNIX story and is also just cute.  Like other computer raster displays of the time, the aspect ratio is taller than it is wide, a trend that would eventually reverse.  This same "tall" display can be seen for instance on the Xerox Alto.

Seeing as I've never seen one of the former pop up at auction ever, and there is currently *just the mouse* for a DMD 5620 on eBay for somewhere around $25,000, I really, really don't think I'll own either of these ever, not that I really intend to unless a perfectly ideal situation lines up *and* I have a museum in the area that would rather have it for display the 99% of the time I wouldn't be using it.  That's my constant struggle with this sort of thing, I very rarely want something because I want it to be in my possession 24/7 and to be permanently using it for things.  I much rather just want access to the thing when I need it and then for it to be benefiting others when not.  Unfortunately the closest we have to anything like that in the immediate area is an electrical history museum that has more than once expressed their disinterest in any of the research or artifacts I've brought to them since they don't consider anything newer than like the 40s or 50s or something like that.  History of electricity stops at 1960 everyone, we can all go home.

[paul-f]

Congratulations on your find.

It brings back fond memories of working for DEC many years ago and having my VT-52 dumb terminal replaced by a VT-180 (VT100 with embedded two-board Z80 computer and dual diskettes). I was a happy camper.

You should hopefully find room for your project.

[segaloco]

Follow up now that I've dusted it off and hooked everything up, as expected, it isn't 100% perfect.  I currently cannot get the built-in CRT to activate, but have not yet traced where the fault is.  However, there is also a standard CVBS BNC connector on the back and I was able to run this into my Trinitron and confirmed the computational/logical components of the terminal appear to be functioning properly.

Attached is a photo of a suspect component, that little grey component with the very blobbly solder points on either side of it.  Couple that with what looks like some burning/charring around the contact points.  That said, that could be something that broke previously and was replaced just shoddily but should work.  The only other thing I spotted in a quick visual inspection without demounting everything was that the red thing to the right of it (dunno what it is, I'm still learning my circuit-board jargon) has a hairline fracture on the other side just out of the pic.  I don't see any evidence of say a dielectric leaking out of it or anything, so can't really say for sure, I don't think it's a capacitor.  I know the specs for the VT100 are floating around out there so I'll do some formal circuitry checks once I've done thorough diagnostics on everything else (as I don't want to rabbit hole on the display if some data pathway is royally screwed on this thing.)

Admittedly CRTs also give me the willies, almost zapped the daylights out of myself once when trying to remove an anode button.  This was early on and I hadn't done my safety research to understand precisely how much voltage the CRT retains...but decided against trying to disconnect it on that project since I didn't actually need to do anything under there, I was just getting familiar with the various connections on a CRT at the time.  I learned only days later about having to discharge the anode first by sticking a grounded flathead screwdriver up under the suction cup.  Scary stuff, I'm more careful now but certainly have a bit of trepidation around the anode button on CRTs.

Anywho, like many of my projects, I'm moving slowly on this, so if you have some fantastic VT100 diagnostic advice I don't take right away, my apologies, I'm probably tinkering inside my KSU at the time or wrapped up in some software disassembly thing.  Let's just say my proc table has copious entries but my scheduler algorithm can be a bit biased...

[SUnset2]

Nice find.  It brings back memories for me, too.  I used VT-100s at work and in college.  I also picked up one surplus for cheap, and used it with my S-100 Z80 system running CP/M.

The round device that looks like the board got hot from it is probably a diode.  The electrolytic capacitors are the most likely part to go bad over time.  You will need to get a schematic.  Other than the high voltages, the monitor section shouldn't be too difficult to troubleshoot, it basically just a B&W TV.

[tubaman]

Nice - brings back memories from work of using one of these in the late '80s, and it was considered a bit old then. It was connected to a Genrad 2271 In Circuit tester that I was responsible for. Also used a VT220 on a Marconi 800X tester - much more modern!
 

[segaloco]

Replaced that board finally, unfortunately it wasn't the immediate improvement I had hoped for.  The filaments in the electron gun now seem to warm, but I don't get a picture on the screen, or any evidence of electrons striking phosphor anywhere.  After it sits for a few minutes I start to hear a bit of quiet, squeaky noise from the flyback here and there, but nothing resembling HV.

I can feel the warmth coming off of the components in the back so I know the new board has livened some things up...but no dice on the display.  Luckily the technical manuals are pretty thorough so hopefully between that and my oscilloscope I should be able to at least diagnose the culprit...although seeing this thing light up would be the real reward.

[segaloco]

I'm now just about convinced it's a short inside the flyback requiring replacement.  Attached are a couple shots with slightly different lighting.  Notice on the left side of the coil where the ferrite core enters, just on the edge of what I assume is some sort of polymer or other potting compound you notice some white little nodules all around that interface.  I suspect this is galvanic corrosion due to a short inside the induction coil there.  It certainly gets hot and starts to give off a strange smell after being on for several minutes...but I don't know how much of that is natural induced heat vs indicative of a short or other breakdown in galvanic isolation.

Anywho, my prognosis is I'm probably going to have to replace the flyback too.  I'll pull it out next time I have a dry day outside to do a discharge.  Then I'll carefully remove it and see if there's any testing I can do on it in isolation with lower voltages.  I'm pretty convinced that's corrosion indicative of a short though...but I'll feel that much better putting eyes on that.  Heck, if the flyback is all sorts of broken, it may not even be holding high voltage in the tube, but better safe than sorry...

[segaloco]

Word on another forum is the corrosion may have a simpler culprit, that polymer compound sitting right next to it.  Someone elsewhere suspects its one of those electronics glues that turns corrosive over time and has started attacking the ferrite core.  Either way, all signs point to this poor flyback as beyond hope.  Now to play the waiting game for a working one to pop up somewhere that isn't attached to another VT100 already, I've already got one more CRT than I'd like to have sitting around at present...

[segaloco]

It lives!

Turns out yeah that flyback also needed replacing.  Sadly still not out of the woods.  While the terminal appears to be properly functional now, I haven't gotten correct serial communication from it yet.  I plugged it in via a RS-232/USB adapter I have, set the interface to 9600 baud, and started to cat(1) output from it.  Keypresses do result in characters populating on my screen, but they're all the wrong ones, and then characters sent to the terminal don't show up at all.  The little light on the USB adapter that flashes on transmit does flash as I send characters, so communication is going somewhere.

It's not the keyboard as when the terminal is in local mode, it does print the correct characters on its own screen, but then the data sent over the line seems to be the problem.  Haven't messed with it too much yet, apparently there are some dip switches in the back that may be relevant.  Either way, if I can't figure it out from that angle, guess it's time to plug in the oscilloscope and see what the signal actually looks like...

[5415551212]

Nice I am sure you have found vt100.net
this might be the manual:
https://vt100.net/c.itoh/cit101-um.pdf

What are you trying to connect it to? A linux box?

[SUnset2]

Great job fixing the monitor.  Have you tried a simple loopback test (just connecting pin 2 to pin 3)?
It might also need some of the handshake lines (like connecting CTS to RTS, and maybe others).

[segaloco]

I think it's a level issue.  The technical manual states the TX/RX lines are EIA levels, meaning 6-12V range (- for 1, + for 0) rather than even the 5V serial adapter I was using.  I recall learning in the past that today's 3.3V TTL levels are lower than what used to be used, and that 5V was the previous level before 3.3V...but I guess I've gone far enough back that I'm now in the 12V range.  These adapters are proving harder to find than the little dual 5V/3.3V ones.  Hopefully a proper 12V adapter is all that is left.

I think it's promising though, this is the only thing left that I can tell is wrong.  I'm hoping to get this thing to 100% because I then want to use it to drive the Data Set 103J I've put together and see if I can get real modem noise out of it.  If I'm understanding that side of things correctly, I need to simulate an "answer" of the remote modem along with the tone(s) it makes to then get the local one to start sending beep boops.  Currently the line circuit will stay active for a little while after depressing the data key, but the lamp goes off after about 5 seconds or so, presumably because some sort of await for the answer signal had timed out.  If it winds up needing an answer signal, maybe I can get this modem to "answer" something and rattle off the necessary noises.

[ka1axy]

All you need for an answer tone is 2225Hz (mark)
That should trigger the 1270 mark from the originating modem, and then when you send data from the terminal, you'll hear 1070 (space) alternating with 1270.

Your terminal will (should) send a mark when no keys are pressed and itmis online. The 103 modem will probably want all the modem control signals,so DTR, RTS should be asserted (you can tie those to DSR, which asserts when the dataset is powered up) You should see CD when the 2225 tone from the answer modem asserts and is detected, once you have CD,DTR and DSR, CTS should assert in response to RTS.

[5415551212]

Look for a RS232 to usb converter based around a Max232 chip. If I remember correctly they go to 10V which is enough for older RS232 equipment.

https://download.mikroe.com/documents/add-on-boards/other/interface/max232/max232-board-manual-v100.pdf
https://components101.com/modules/max3232-rs232-to-ttl-serial-port-converter-module
https://www.vt100.net/docs/vt100-tm/chapter4.html
https://vt100.net/dec/mp-00633-00-rev-initial.pdf

[segaloco]

Got a component on the way for the level shifting.  I'm realizing that technically doesn't matter for the modem since that should accept the RS-232 line out of the terminal.  Indeed the SD light illuminates with each key press when the line is active, so it is acknowledging characters from the terminal correctly.  If nothing else I've got a sine generator I can just tune to those frequencies to supply the signals.