• Please review our updated Terms and Rules here

Fixing a VT100 power supply

BBenchoff

Member
Joined
Jun 25, 2013
Messages
15
Well I have a VT100 sitting on my workbench with a broken power supply.

I've run across this tutorial in the wiki about improvising a power supply from an old ATX supply.

Since that tutorial is about 2 years old, I was wondering if anyone had gone through the trouble to source the components to rebuild the original DEC supply.
 
Sure, you can fix the existing power supply- but first you have to know what is broken on the power supply.

Have you probed it under an attempted load to see what its behavior is?

First guesses would be the DC-side filter caps...
 
Well I busted out the scope and meter, and I'd bet I'm looking at a blown transformer here.
 
I have repaired VT100 power supplies at the component level. The OP can help himself first by downloading the print set for the VT100 and digesting the power supply section of the VT100 technical manual.

The VT100 uses a switching supply. I have never found bad iron in one. The problems were related to the main switching transistor which is the first thing to check in a switching supply.

Lou
 
Well there's my problem. I just lost a bet with myself.

You don't happen to have a supplier for this thing, do you lou? My google-fu is pretty good, and I can't seem to find a datasheet, even.
 
The VT100 uses a switching supply. I have never found bad iron in one.

Lou
Lou helped me through long series of component checks starting with the main transistors, BUT in the end I did find there was an open circuit on the primary of the main transformer - that's when I gave up and built the ATX conversion.

Rick
 
Just my 2 cents... but I think there's a market here for some industrious individual to come up with a nice clean adapter kit to replace broken VT1XX power supplies with ATX substitutes.

If you do, please consider those of us with VT103s... it's a constant worry.
 
Just my 2 cents... but I think there's a market here for some industrious individual to come up with a nice clean adapter kit to replace broken VT1XX power supplies with ATX substitutes.

If you do, please consider those of us with VT103s... it's a constant worry.

Seeing as how I'm looking at doing just that, what do you have in mind?

I'm thinking a circuit board that plugs into the power supply card edge slot that has an ATX 20+4 pin adapter and all the fun stuff to get the -23 Volt rail.

From what I can gather from RickNel's tutorial, I'm looking at about $15 USD for the components (minus the ATX supply). You can get ten 5 cm by 10 cm boards fabbed for $25. Basically, I might be able to put together an VT100 to ATX power supply adapter for about $20 a pop, minus shipping.

I've got a copy of KiCad and enough experience to put this together (it's not exactly a difficult circuit). If RickNel could send me a BOM I might be able to piece this together before the summer is out.
 
Alright, I broke out Eagle and started making a part for the VT100 power supply connector. Mechanically, it will work. Electrically, I'm going to need some help.

Could someone look at this footprint and tell me if it's correct? I'm just trying to double check everything here. Here's an imgur link to the same pic

Power Board.jpg

I forgot exactly how big this connector was. Until I get a BOM and the layout all sorted out, I'd guess this board would cost about $50 for 10 pieces. That's still about $25 for the parts and PCB for a VT100 to ATX power supply adapter. I'm happy with that.
 

Attachments

  • Power Board.jpg
    Power Board.jpg
    16.5 KB · Views: 4
Alright, I broke out Eagle and started making a part for the VT100 power supply connector. Mechanically, it will work. Electrically, I'm going to need some help.

Could someone look at this footprint and tell me if it's correct? I'm just trying to double check everything here.

A PCB sounds like a great improvement over my DIY construction!
As to BOM -
The components are just standard ones with no special ratings
[from my blog post]

The VT100 requires a -23v rail to be used together with +12v to create a 35v potential in the circuit for writing setup changes to its Non-Volatile RAM. This had to be synthesized with a small circuit employing a transformer to deliver 24vac, a diode bridge rectifier, smoothing capacitor, and a 24vdc voltage regulator. The regulator is not just to protect against over voltage. It also reduces the ~24vdc coming out of the bridge rectifier to about 22.8vdc, close enough to the 23vdc spec.

The blog also has a circuit diagram indicating values of components I used.

As to the connector, note this important reference to finger J

"During testing, I discovered an undocumented trace on the VT100 Basic Video Board that links address line LBA6 of the Line Buffer SRAM directly to finger J of the PSU connector. This is nominally 12v GND, and in the original PSU configuration 12v GND floats independently of the 5v GND (which is the TTL logic GND). The AT/ATX PSU uses a single common GND, so if finger J is grounded it will hold the Line Buffer address line 6 permanently low and thus cripples the display.

When using an AT/ATX PSU for the VT100, it is ESSENTIAL that finger J NOT BE CONNECTED to the replacement PSU GND or to chassis earth.

Rick
 
Okay. Here's what I have so far, with an imgur mirrorboard.jpg

That image shows all the hard stuff is out of the way: getting the card edge slot just right, and finding a good transformer (and building the Eagle part).

As far as the physical mounting to the existing power supply bracket, I'm thinking the easiest way would be to cut a piece of wood or fiberglass board to the size of the original power supply board and mount the ATX supply and this new PCB to that. Nothing terribly hard there.

Rick and Lou: I wanted to run by you about the DNC finger J on the card edge connector. Since this board will hopefully end up being used for the VT103s and 180s, is there anything you're aware of that would prevent Rick's circuit from working with all VT1xx boxes? The finger isn't connected to anything and there's a handy through hole for hardware hacking, but I'm just thinking documentation here.
 
Alright, everybody. Here we go:

Supply.jpg

Here's the imgur mirror.

The cheapest way for me to get this made is to send it off to a board house in China. That means a ~1 month turnaround time with shipping included. I'll get a github repo with the schematic, board file, and BOM up shortly.
 
Again, kudos for taking this forward.

1. You haven't indicated the dimensions of the board, but I'm assuming you have ensured that there is as much clearance as possible within the VT100 PSU enclosure for the separate ATX PSU (dimensions may vary), and VT100 original cable entry and switch/fuse.

2. I didn't make it explicit in my own account, but because the 24vac output transformer connects directly to mains power, the input specifications must match local mains voltage - in my case, 240vac NOT the 115vac on your suggested Mouser component.

3. The top of the PSU enclosure is quite close to the yoke of the CRT within the VT100 case. In testing, I found that emissions from the 24v transformer would interfere with CRT deflection if placed high in that enclosure. I would recommend placing the transformer as far away from the yoke as possible, at the bottom of the connector board.

Rick
 
The size is small enough. My gerbers come out to a size of 4.19x 3.26 inches (106.426x 82.804 mm), so it'll fit.

Concerning points two and three, I think I may have a solution to that. Instead of a transformer soldered to the board, how about a chassis mount transformer (something like this would do)? This would do three things: Get the size (and cost) of the board down, allow anyone to move the transformer around in the cage, and prevent the CRT from being deflected by the transformer.

Really, what I'm looking at is a balance between a 'ready to go' solution, and something that works for everyone. Going with a Hammond transformer would work for everyone on the planet, you know.

Eh, screw it. I'll just redo the board with this in mind.

EDIT:

Well gentlemen, here we go: ATX.jpg and imgur mirror

I put the transformer off the board, so I'll just mount it somewhere to the chassis of the ATX supply. The size is 2.24x 3.26 inches (56.896x 82.804 mm), so with just a tiny bit of fiddling about, I can get 10 of these boards from China for $25, or what I originally intended a few posts back. With the separate transformer this will be usable for everyone on the planet. Woo.

Rick, I'm going to get a few of these boards made and report back to you with how you integrated your circuit with the original DEC power supply switch/voltage selector, etc.... I think you have a different version of this panel than me, but it's not too dissimilar.
 
Last edited:
Ok... I tried to do this myself, but it turns out I need some help.

I just went downstairs to pull apart my VT103 and my VT100 to do a side-by-side against your concept. Much to my embarrassment, I now realize I have two VT103s and no VT100.

The supplies for the VT100 and the VT103 are the H7831 and H7835, respectively.

Output specs for each are:

H7831 (VT100)

  • +5@1.5A min to 11.0A max
  • +12@1.0A min to 2.75A max
  • -12@0.0A min to 0.5A max
  • -23@0.01A

H7835 (VT103)

  • +5@16.0A
  • +12@5.0A
  • -12@0.5A
  • -23@0.01A
  • BPOK
  • BDCOK
  • LTC

Form factor appears very similar, as one would expect being in the same terminal case. Both units have the same VT100 boards to contend with.

However, the VT103 also has a fan that is obviously powered somehow and a QBUS backplane. Models with the TU58 have it powered by separate connector from the backplane.

I believe the cables are the main interconnect difference but the supply end may be compatible. I do note however that positions for A thru V on the main supply connector are physically swapped in the drawings of the two units. The VT103 supply also has a socket to receive an additional board (H786) which creates the BPOK, BDCOK and LTC signals. This board plugs into the H7835. Cabling for those signals must be conducted via the main connector and cabling.

So, while I can co-operate with someone to check if your project can be done to be compatible with both, I can't easily do both ends without relying solely on my interpretation of the schematics, which I am reticent to do.

I'll be happy to document the VT103 here online in photos so we can do a comparison and answer this question before finalizing your order, but I'll need someone else to do the VT100 side. (Unless someone in my area has one I can document?)

My sense is that pins on the supply's output connector are indeed assigned upwardly compatible, and that your design could accommodate the needs of both VT100 and VT103 supplies by including artwork for the connector to receive the H786. (or by including logic to do that function without the H786)


It would be a shame to miss the opportunity if it exists, as the ATX supply can easily provide power sufficient for both needs.


It's kinda late to do all that tonight. Please post if I should go ahead, and I'll check back. This delay may provide time for a VT100 rendezvous to surface too.
 
Oh wow. I didn't realize the 103 was that different. That will probably need a completely different circuit and board design.

It shouldn't be that hard to do another version, though. I just found this thing that generates the BPOK and BDCOK signals with parts I have lying around in my parts drawers. The hard part - making the card edge connector - is already done. Once I can confirm the VT100 version is working, I can start work on the 103.

Go ahead and start documenting the 103 connector and supply. If there's a way to make a single board work with both, I'm all for it. That being said, making two versions isn't *that* hard. Hell, I made the VT100 version in a few hours.

I'll be happy to document the VT103 here online in photos so we can do a comparison and answer this question before finalizing your order, but I'll need someone else to do the VT100 side. (Unless someone in my area has one I can document?)

Where are you?
 
Last edited:
I agree taking the transformer off the connector PCB makes it more flexible - especially if you want a board that can accomodate the extra items for the 103. My 24v transformer is just bolted to the bottom wall of the PSU cage. That provided chassis ground without any additional wiring.

Apart from the PCB, I'm not sure what you have in mind for the construction board that would slot into the top and bottom rails vacated by the standard PSU to hold the assembly stable. The PCB needs to be mounted on something or it may work loose from the edge connector. I used common 3-ply wood laminate because it provided all the strength and insulation I could ask for at that thickness, without the expense of any of the professional PCB materials. MDF at that thickness would not be strong enough.

You need to consider whether the new assembly is to be inserted as one piece on a common baseboard attached to the external panel, or whether the PCB/24v transformer section is inserted first and homed in the edge connector, followed by the ATX/external panel section. In the second case, you will need extra length on the wires connecting the two sections.

Re the fan on the VT103 PSU - an ATX PSU will normally have its own fan which is more adequate. I also thought it advisable to orient the ATX unit with its fan motor not too close to the CRT yoke, as those fan motors emit a small EMR ripple of their own.

Rick
 
Oh wow. I didn't realize the 103 was that different. That will probably need a completely different circuit and board design. ...It shouldn't be that hard to do another version, though.
That wouldn't be a bad thing. It's just the idea of another order. You da' Boss - I just kibitzing.

I just found this thing that generates the BPOK and BDCOK signals with parts I have lying around in my parts drawers.
Shouldn't be too hard to add the LTC to that.

If there's a way to make a single board work with both, I'm all for it. That being said, making two versions isn't *that* hard.
...maybe use another edge for the VT103 / H7835 specific connector? I'm really undecided between re-using the H786 or duplicating it's function. Both require parts to be bought and some connectors can be difficult to find. Too hard to provide both options?

Where are you?
Cincinnati Ohio area.

Apart from the PCB, I'm not sure what you have in mind for the construction board that would slot into the top and bottom rails vacated by the standard PSU to hold the assembly stable. The PCB needs to be mounted on something or it may work loose from the edge connector.
Agree - but then there are several other mechanical issues to this conversion. The hobbyist will have to have some ingenuity.

Re the fan on the VT103 PSU - an ATX PSU will normally have its own fan which is more adequate.
The 103 has a fan specifically for the QBUS card cage, though powering it could be done adhoc. Thought it was worth mentioning as a difference.

Loose the transformer and just use a DC-DC upconverter from the -12 for the -23. You only need it for the EAROM.
More elegant, certainly. How much -12V can an ATX supply be relied upon to give? I know most manufacturers dropped one of the minus rails and don't usually provide it any more - can't recall which ATM.


I will post more details on the 103 this evening to cover the points raised.

Good discussion.
 
Back
Top