• Please review our updated Terms and Rules here

VAXstation issues.

> "1) So the Vax won't start at all with 0.11V. I suppose I can emulate this DCOK."
> "Sometimes the PSU starts with alternating voltages where it goes between low and high every for example 500ms."
There is no point in emulating a good DCOK if your PSU is broken
 
"There is no point in emulating a good DCOK if your PSU is broken"

Unless, as OP stated, you intend to substitute an ATX power supply for testing purposes which doesn't supply BDCOK and other needed signals.

It's a common enough procedure that there have been something like five different solutions posted on this site alone.

No point? Really?
 
Last edited:
> "No point? Really?"
I am an avid user of another forum where some of the rules are:

"Be kind. Don't be snarky. Converse curiously; don't cross-examine.
Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."

;-)
 
jplr said:
"Be kind. Don't be snarky. Converse curiously; don't cross-examine.
Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."
Click to expand...
Those are good policies to follow, online and off, whether codified as a rule or not. Keep them in mind going forward and you'll do fine.

But back to the topic at hand:


avitech.com.au

QBUS Front Panel Project

This project is an Arudino-based front panel for a bare QBUS PDP-11 chassis. It provides the switch and indicator functionality typically found on an early QBUS PDP-11 chassis (such as a BA11-MF or…
avitech.com.au

The original DEC power supplies will continue to fail. The knowledge base of how to troubleshoot and hopefully repair them is and will continue to be a problem. Things were done in early switched supplies in ways that are not common today. People will continue to try and bring up old DEC systems, even if only for testing or as a temporary solution, using an off the shelf supply, or a bench supply, and will need to generate the bus signals that would come from the DEC supply if it was working. The two solutions in those two links and the one posted by Hunta above in post #5 are examples of how it can be done.
 
MattisLind said:
I think that the -9V is used for the ethernet tranceiver. They are usually isolated from the rest so that is why you only get good readings between those specific pins.
Click to expand...
Good point.
It should be easy to tone the 9V pins to ground on the BNC connector if it has BNC ethernet, or worst case more or less just any pin on the "insulated" side of the ethernet circuit, to 100% verify this.

Also: I'm 95% sure that if you use an AUI transceiver the 9V supply isn't needed as the AUI transceiver does the insulation.

In other words, for testing purposes a PC ATX PSU would probably be fine, just ignoring the 9V output, and for long term usage either a separate 9V PSU (cheap regulated wall wart hanging out from the box for example) or just use an AUI transceiver and put a "9V missing" sticker over the BNC/TP ethernet connector.
 
If you don't want to wait on the slow boat, a cheap option is automotive light bulbs.
A rule of thumb is that when a light bulb emits light the current is almost constant, so a bulb that consumes 21W at 12V (typical brake light), i.e. almost 2A, would likely consume at least more than 1.5A at 5V.

10 ohm 10 watt seems a bit excessive for 5V, as 10 ohm only results in 0.5A, which in turn only results in 2.5W. On the other hand it would be a bit to weak for testing 12V as you'd end up with 14.4W at 12V.
 
MiaM said:
If you don't want to wait on the slow boat, a cheap option is automotive light bulbs.
A rule of thumb is that when a light bulb emits light the current is almost constant, so a bulb that consumes 21W at 12V (typical brake light), i.e. almost 2A, would likely consume at least more than 1.5A at 5V.

10 ohm 10 watt seems a bit excessive for 5V, as 10 ohm only results in 0.5A, which in turn only results in 2.5W. On the other hand it would be a bit to weak for testing 12V as you'd end up with 14.4W at 12V.
Click to expand...
Is this better:

10W 5R for the 5V
20W 12R for the 12V

That should give 1A on each.
 
I have problems with the power supply of my VAX:
  • The power supply starts now, but it takes 2 minutes from cold power on to go from 4.4V to the correct 5.10 when 5.1R 10 W is attached. A warm power on is faster ... because the capacitors are still charged I imagine.
  • DCOK does not give the good voltages, but I have not tested with pull up resistor.
  • The PSU is recapped except for 1 very small capacitor.
  • If not load is present the voltages are sometimes OK from the beginning as if everything is perfectly OK.
  • If I put 12R 20W on the 12V rail and keep the 5.1R on the 5V then the PSU can not start. It just briefly turns the fans for a second.
  • There are only some resistance that look burned, no other component.
  • Some of the tracks might be damaged from the leaking capacitors.
I am not an electronic engineer, so any help will be much appreciated :)
 
I am not an electronic engineer either but:

> "The power supply starts now, but it takes 2 minutes from cold power on to go from 4.4V to the correct 5.10"
On my VaxStation 3100 the power up is nearly instantaneous, it's DCOK that comes after a few seconds.

> "DCOK does not give the good voltages, but I have not tested with pull up resistor."
On my VaxSation 3100 DCOK is provided by an open collector, there is a pull up resistor on the main board, so there is no need for added resistor. Indeed if you don't connect the PSU to the motherboard, you need a pull up resistor to get meaningful measurements.

> "If not load is present the voltages are sometimes OK "
So it means most of the time it is not OK ;-) your PSU is broken.

> "If I put 12R 20W on the 12V rail and keep the 5.1R on the 5V then the PSU can not start. "
Idem

I would say that you have three options:
- If you want to be in the spirit of a hardcore collector, you pay someone to repair the PSU. I guess it's feasible, but that might end with a PSU with new components or new PCB (Some of the tracks might be damaged). There is a marketplace on this server.
- You make a new PSU with the original schematics, new components, same size factor, and appearance.
- You buy a new generic PSU (as other people have pointed out, creating a DCOK signal is trivial as several people have created schematics and even PCBs). You can probably manage to put it inside the PSU box. You are in Lyon isn't? I may have a DCOK PCB somewhere.

Good luck, I went down a similar rabbit hole a few years ago, it's not a good memory
 
Last edited:
Did you clean the board of the corrosion after you recaped it? Are all of the diodes good?
 
jplr said:
I am not an electronic engineer either but:

> "The power supply starts now, but it takes 2 minutes from cold power on to go from 4.4V to the correct 5.10"
On my VaxStation 3100 the power up is nearly instantaneous, it's DCOK that comes after a few seconds.

> "DCOK does not give the good voltages, but I have not tested with pull up resistor."
On my VaxSation 3100 DCOK is provided by an open collector, there is a pull up resistor on the main board, so there is no need for added resistor. Indeed if you don't connect the PSU to the motherboard, you need a pull up resistor to get meaningful measurements.

> "If not load is present the voltages are sometimes OK "
So it means most of the time it is not OK ;-) your PSU is broken.

> "If I put 12R 20W on the 12V rail and keep the 5.1R on the 5V then the PSU can not start. "
Idem

I would say that you have three options:
- If you want to be in the spirit of a hardcore collector, you pay someone to repair the PSU. I guess it's feasible, but that might end with a PSU with new components or new PCB (Some of the tracks might be damaged). There is a marketplace on this server.
- You make a new PSU with the original schematics, new components, same size factor, and appearance.
- You buy a new generic PSU (as other people have pointed out, creating a DCOK signal is trivial as several people have created schematics and even PCBs). You can probably manage to put it inside the PSU box. You are in Lyon isn't? I may have a DCOK PCB somewhere.

Good luck, I went down a similar rabbit hole a few years ago, it's not a good memory
Click to expand...
I think I would try the new PSU as well (while still trying to make the old PSU works, so mych time lost already). Can you please check for a PCB or send me the schematics so that I can order now please? Yes, I live in France and would pay for the PCB + delivery :)
 
It's a PCB designed by Frainresearch, I have several of them but I never tested them.
You can send me your postal address and I will post one of those PCBs to you.
I make no warranty that it may solve your problems.
Another good reference is:
avitech.com.au

QBUS Front Panel Project

This project is an Arudino-based front panel for a bare QBUS PDP-11 chassis. It provides the switch and indicator functionality typically found on an early QBUS PDP-11 chassis (such as a BA11-MF or…
avitech.com.au
My email is "jeanpierre" dot "lerouzic" "arrowbase" "padiracinnovation" dot "org"
 
Are there schematics for the Model H7821-00 power supply for VAX Station 3100 M38? I have found someone that can diagnose it properly.
 
Last edited:
jplr said:
My email is "jeanpierre" dot "lerouzic" "arrowbase" "padiracinnovation" dot "org"
Click to expand...
Sorry for going off on a tangent, but it's interesting how characters have different names in different languages. The Swedish word, verbatim translated, is "trunk-A", where trunk refers to the trunk of an elephant. Early on some called it "alpha curl" but that didn't really catch on. (Also W is pronounced as V when spelling out WWW in web addresses).
 
I am working on a DCOK card in KiCad for my VAX 3100. I will use an ATX Power OK as input and the output of the card will be connected to pin 7 DCOK. The other rails will be directly connected to the VAX. It will also be a test card with jumpers.

Can someone confirm that polarity is active-high:

Code: 
DCOK low, about 0 V          = power not OK / reset
DCOK high, 3.5 V to 5.25 V   = power OK / release reset

So 5V through pull resisitor will be OK to signal all is OK.
 
Last edited:
tonata said:
I am working on a DCOK card in KiCad for my VAX 3100. I will use an ATX Power OK as input and the output of the card will be connected to pin 7 DCOK. The other rails will be directly connected to the VAX. It will also be a test card with jumpers.

Can someone confirm that polarity is active-high:

Code: 
DCOK low, about 0 V          = power not OK / reset
DCOK high, 3.5 V to 5.25 V   = power OK / release reset

So 5V through pull resisitor will be OK to signal all is OK.
Click to expand...
The Qbus signal BDCOK is indeed an active high signal, as is BPOK. I think all the rest of the Qbus signals are active low.
 
The card is focused only on emulating DCOK. The card will follow this logic:

1. Contemporary (compatible) ATX PSU starts.
2. +5 V and other rails begin to rise.
3. Card waits until VAX-side +5 V is high enough.
4. Card waits an additional adjustable delay.
5. Only then DCOK_OUT goes high.

J1 - From modern PSU / ATX side (inputs):
+5V_ATX powers the card logic
GND common ground
PWR_OK_IN ATX Power Good input

J2 - To VAX side (input/output):
VAX_5V_SENSE thin wire from one VAX +5 V pin, e.g. pin 5/6/8/10
GND VAX ground, e.g. pin 2/3/4/11/12
DCOK_OUT to VAX brown wire, pin 7

Does it sound good?
 
Last edited:
I realized I do not have the current specification for my Vax 3100 M38. I tried to assemble some specification, but I am unsure of it:

Rail / signalNominal voltageEstimated max currentNotes
+5V+5.1 V15-18 AMain logic rail. For ATX replacement, should be >=18 A, preferably 20-25 A on +5 V.
+12V+12.0 to +12.1 V4.5-6.0 AMainly drives, fan, and motor loads. Most ATX PSUs provide enough +12 V.
-12V-12.0 to -12.1 V0.2-0.5 ALow-current rail, likely for serial/analog circuitry. Modern ATX -12V at 0.3 A may be enough; 0.5 A is safer.
GND0 V20-25 A aggregate returnSize ground wiring for combined return current from +5 V, +12 V, and -12 V. Uses multiple ground wires/pins.
DCOKActive-high, approx. 3.5-5.25 VSignal only, <20 mAPower-good input to motherboard.

Can somebody verify these numbers?
 
You must log in or register to reply here.
Back
Top