• Please review our updated Terms and Rules here

TV Typewriter replica project back on track

falter

Veteran Member
Joined
Jan 22, 2011
Messages
6,556
Location
Vancouver, BC
After a whole bunch of diversions I am finally getting back to building my TV Typewriter prototype replica. Don Lancaster has been great with supplying info about the original's design. I've finally secured an MDS keypunch keyboard that is correct (keys and pattern all match that of the prototype photos, just not colour). I'm going to feel bad about disassembling a vintage piece like that, but the machine it was attached to is long gone anyway.

I have an opinion question though: should I strive for complete accuracy and see if I can obtain photos of the prototype circuit boards somewhere and try to make the replica internals as close to the original as possible? Never seen photos of them before and not sure if whoever has it would be willing to open the original. Option B is use the plans printed in Radio Electronics. Then I could go further and socket the chips so I can swapnout quickly where they fail..
 
I think that Augat machined-pin sockets came a bit later; I do remember that most of the IC sockets found were those awful TI models--which I used in my TVT back when.

Just wondering where all of the plain 74xx (not 74LSxx) logic came from, not to mention the old 1000 nsec. (very slow) 2102s (21L02s/91L02s came a bit later). Those 6 buggers ran very hot. The right character generator couldn't have been easy to find.

If you're doing this thing strictly period-correct, my hat's off to you.
 
I think that Augat machined-pin sockets came a bit later; I do remember that most of the IC sockets found were those awful TI models--which I used in my TVT back when.

Just wondering where all of the plain 74xx (not 74LSxx) logic came from, not to mention the old 1000 nsec. (very slow) 2102s (21L02s/91L02s came a bit later). Those 6 buggers ran very hot. The right character generator couldn't have been easy to find.

If you're doing this thing strictly period-correct, my hat's off to you.

Well, not 100% right off the bat, no. Some chips of the right vintages are impossible to find. The goal was to make anything I could period correct, get it running with the closest chips I could find for now and then work on finding period correct pieces.

I have had some luck, like these char gens:

1440023509331.jpg

And these guys, which I think, if I'm reading date codes correctly, are 1974 or 1979 vintage:

1440023599534.jpg

These 7400s are 79 vintage, but they'll do I hope:

1440023799923.jpg

They pretty much all came from ebay. There was a chip place in California I understand closed its doors and dumped a ton of NOS and working used chips there. Still have a bunch more to get, but I'm sure I can get what I need to build it.. then just spend the rest of my days slowly accumulating correct date chips. I like neverending projects.
 
I think that Augat machined-pin sockets came a bit later; I do remember that most of the IC sockets found were those awful TI models--which I used in my TVT back when.

Just wondering where all of the plain 74xx (not 74LSxx) logic came from, not to mention the old 1000 nsec. (very slow) 2102s (21L02s/91L02s came a bit later). Those 6 buggers ran very hot. The right character generator couldn't have been easy to find.

If you're doing this thing strictly period-correct, my hat's off to you.

Well. I certainly remember using the GP IC ones in the '70s, and as for machine-pin sockets in general (not IC), the tube variety has been around since the late '30s...

BUT, my "instant-recall-memory" definitely need a trip to ReKall - ReKall - Rekall, so YMMV!

PS : I also did a lookup in a '76 Fairchild handbook, and it lists 2102's (plain) as being 350ns (where on earth were these 1000ns chips???)
Although my handbook (and a 1975 Intel one) only list 2102*A*, so maybe that's the difference....


gwk
 
Last edited:
I didn't think my mind was playing tricks on me--have a look at the late-1973 Intel Memory Design Handbook. PDF page 129, "Selection Guide". See the timing for the 2102? About the same speed as the 1702A EPROM. 350 mW too.

You can see that, right about that time, where dynamic MOS shift registers looked pretty good for screen refresh. Even in 1976, the 2102-2 was 850 nsec.

If you wanted to go fast, you could use bipolar RAM, but have a fresh supply of eggs to fry on them.
 
I just found a bunch more chips I needed on ebay, 1973 vintage (!?). The supplier (acpsurplus) seems to be dumping a lot of vintage chips on the market. I seem to recall reading around here somewhere about a major chip wholesaler or something in California going bust or shutting down, and these being the result of that? Anyone know the story? Can't find it now via search. Seems odd that they'd have 1973 NOS chips still hanging around. I got 15 or so 6540 ROMs that way. These chips are becoming a collector obsession for me all on their own.
 
This weekend I will finally begin a bit of assembly; of the case, just to have something to inspire me to keep moving forward. To match Don Lancaster's prototype, I've got what I think is the correct wood and of course just recently, the Mohawk keyboard.

I'm currently working on a blog as I progress: http://www.bradhodge.ca/blog and welcome comments/suggestions/factual corrections. But one thing I was curious about -- did Mohawk customize their keyboards according to customers? When I was working on one of the blog posts I found a pic online of the keyboard from the first auction I missed out on, and then noticed that it was slightly different than mine. Mine has a rectangular key at the far right, and far fewer switches up top.

Also wondering if anyone has a pic of the complete MDS key to tape system that used these keyboards. I can't find one anywhere and I've really been looking. Just curious.
 
Thanks Chuck,

I literally have no idea what model of system my keyboard belongs to. That's kind of the problem.. this stuff predates my existence by a few years. Just figuring out who made my keyboard was hard enough. :) I'll take a look at what you have there.. but unless I see a photo of the system with my actual keyboard being used with it I'll probably never know if it's correct.


Are you talking about the MDS 1101 system. Bitsavers has some stuff on it.

Brochure
Handbook
 
Well, there were several early 60's models with that type of strange keyboard. Later (70s) Mohawk key-to-tape systems had more conventional-looking keyboards.
 
Question for sharp eyed folks -- if you had to take a guess as to what wood the sides of the TVT prototype were made of, what would you guess (I'm kind of hoping maybe I get lucky and someone here has been to one of the shows or the museum and actually seen the prototype up close and very unlikely noted what wood it was :))

IMG_3579.jpg

Initially I guessed oak. Depending on photo it can be a bit tricky. I'm still fairly certain it's hardwood and based on another photo I picked up some fir. But I'm having doubts now that I have that photo the museum was kind enough to send me of the side with the knot in it. I'd kind of like it to be exact.

I'm also wondering if anyone has *any* guess as to those switches up top (clear/run, on/cursor) etc. I have *no* idea who made those and Don hasn't answered email recently.

Many thanks!
 
I'm not sure about hardwood--the sides to me look like a softwood--probably pine or Doug- fir (notice the "tearing" at the corners), The front looks like a 2x4 of perhaps western hemlock. The top is a puzzle--could be hard maple.

The switches look like plain old panel-mount rocker switches. You can get them in any sort of configuration you'd like--momentary, center-off, etc. Make by a bunch of manufacturers.

G17286B.jpg
 
Great stuff -- thanks Chuck. I was hoping to try and match the switches as exactly as possible (at least, from a look standpoint). What you have there is pretty close.

Yeah I figured fir or pine. I went with fir. When I was at Home Depot the pine just didn't look right somehow.

20150926_141451.jpg

To try and get a better scale of things I took the direct overhead shot the computer history museum has on their site and cropped it, then printed it to life size on my plotter. After checking the keys on it against mine and confirming the measurements were the same, I discovered the front piece of wood was only about 1.25" rather than 2" wide at the top. Kind of a weird size.. 1x4? I think there's another piece like it on the other side of the keyboard, followed by plywood the rest of the way back. I got a back photo from the museum that shows what appears to be metal on the bottom? Thinking the case bottom might be metal like my CT-1024. Also trying to figure out what those screws are holding up in the back.

IMG_3572.jpg
 
Well, a 2x4 is 1.5 x 3.5 finished standard. (you can only get a true 2x4 when it's unfinished rough cut). The same way that 1" is actually 0.75". Up there in BC, you have plenty of mills and lots of Doug-fir (not a true fir; belongs to the pine family--or weeds, depending how close you live to them :) ).

http://mistupid.com/homeimpr/lumber.htm

Doug-fir has changed a lot in 40 years, though. I've got some 2x4 studs that you'll break a hammer trying pound a 10 penny nail through--old-growth stuff with tiny dense annular rings--nothing like what you'd buy off the rack today.


Screws might be holding the power transformer...
 
Ok. I guess the top view of the case is just messing with my eyes. I should know from model railroad experience that photos warp things. I know my keys are .75in and it looks like I can fit two abreast across the width.. implying 1.5". But it still looks wrong up close. And then there't the top.. I keep vacillating on whether the top os a single piece of 3/4" plywood.. or a 2x4 or 2x2 followed by plywood. I'm sure there is plywood part way.. there is one good photo where the vinyl wrap has pulled away. But a 2x4 there leaves almost no room in the case.

The museum guys actually offered to try opening it (after all, the important stuff is inside). But they couldn't find a safe way in. Surprised Don didn't make provisions for that. They'll not ever likely open it.. could be a pair of old socks in there for all anyone knows. :)

That's interesting about wood. I've run into that with wall studs in my 40 year old house. You don't think that much changes in 'just' 40 years.. not stuff like that. But it sure does.

Well, a 2x4 is 1.5 x 3.5 finished standard. (you can only get a true 2x4 when it's unfinished rough cut). The same way that 1" is actually 0.75". Up there in BC, you have plenty of mills and lots of Doug-fir (not a true fir; belongs to the pine family--or weeds, depending how close you live to them :) ).

http://mistupid.com/homeimpr/lumber.htm

Doug-fir has changed a lot in 40 years, though. I've got some 2x4 studs that you'll break a hammer trying pound a 10 penny nail through--old-growth stuff with tiny dense annular rings--nothing like what you'd buy off the rack today.


Screws might be holding the power transformer...
 
The bottom definitely looks like an aluminum plate, which wouldn't make it unusual.

The wood on the back (and probably on the top) looks like fiberboard (MDF)--you can see in the shot of the rear, where it shows its texture. I've seen the stuff used in furniture making for non-appearance stuff like drawer sides. It usually has a top layer of paper bonded for appearance. It has the characteristic that, while it's comparatively weak, it's very stable dimensionally. Sometimes called "furniture board".
 
I didn't catch the mdf at the back, but now that I'm looking at it.. yeah. Kind of does look like that. It's all going to be covered by vinyl anyway.

The real decision piece is the circuit boards themselves. I asked Don about them and it kind of sounded like the prototype boards were part of a run of them that they made and debugged. The RE article doesn't indicate if the pic of them shown there are in fact the ones in Don's prototype or later SWTPC kit boards. I'm guessing the basic design would have been the same, minus the SWTPC and parts location silkscreening up top?

Kind of wonder what happened to the TVT unit that was on the magazine cover. Nobody seems to know where that went or how functional it was. I'd wanted to do that one originally but the keycaps apparently were only made for a little while by Mechanical Enterprises. Haven't seen any that have the correct lettering.
 
Took my best guess at it. After a day of sawing and learning how to not use a router...

20150927_171659.jpg

20150927_171547.jpg

Not bad if I do say so myself! Really kind of neat to be able to see it come to life from pictures.

One more question for the guessers.. how do you think Don did the key, switch and front labelling? I'm wondering if the front label is metal?
 
Here's a side by side shot -- on the left, Don's prototype, on the right, mine (the keyboard on mine is not pulled fully up into its space. Looks like my width is just a bit off. Don rebuilt his keyboard to be 13.5 keys wide, mine appears to be 14 at the moment. Not sure yet how to rearrange the keys. Having seen another MDS keyboard with a slightly different config than mine (same foot print, just keys arranged differently and mine has that extra rectangular key), I'm assuming I should be able to change the key positions around. I think MDS must have placed keys according to customer needs, so in theory this should be re-configurable. It looks like the switches slide lock into place along twin rails inside. The wood I used for the sides is fir -- looks a fair bit different in shade than Don's. The plywood and 1x2 piece I used is totally based on guesswork and hours looking at photos. :) I've no idea what's really under there although as Chuck posited MDF would be a good guess.

tvt-sidebyside.jpg
 
Back
Top