• Please review our updated Terms and Rules here

DEC PDP-11/10S Power Supply Repair

glitch

Veteran Member
Joined
Feb 1, 2010
Messages
5,051
Location
Central VA
full writeup here: http://www.glitchwrks.com/2019/06/05/pdp1110-psu-repair

A bit over a year ago, I finally repaired a RX211 controller while working on a friend's PDP-11/34a. The RX211 was destined to live in my PDP-11/10S. After getting the RX211 tested and running, I started bringing up the 11/10; naturally, something let go in the power supply and smoked, taking the machine down. Well, I finally got around to fixing it! The problem turned out to be a shorted bridge rectifier in the 5411086 board, which provides +15 VDC, LTC, and AC LO/DC LO. It shorted hard enough that it blew traces off the edge connector:



Ouch! I repaired the regulator, went ahead and replaced the +15 V output filter cap with a new low ESR type (the original smelled like leaking electrolyte), and patched up the burned connector. I ended up cutting the edge connector contacts off and soldering on a wire pigtail: as luck would have it, there's a Mate-n-Lok on the short pigtail between the AC transformer and the 5411086 edge connector socket.

I also took the opportunity to put a 15 A fuse inline with the AC supply to the 5411086. If your 5411086 is still working fine, you might want to do this before traces get burned away! It'd be easy to make up a pigtail with the fuse holder and two Mate-n-Lok connectors on each end, so that the wiring harness isn't permanently modified. The required connectors are still readily available (mine came from Mouser). The other regulator modules in the H765 power system could also suffer from a shorted rectifier blowing traces out, but they all use Mate-n-Lok connectors instead of edge connectors, so the damage would be easily repaired.

With the supply repaired and bench tested, I fully cleaned the rest of the 11/10 and put it back together. Success! Everything still works, the shorted rectifier didn't harm the rest of the system! Here's a few pictures:



From the front, powered off.



Minimal board set: CPU, bootstrap/terminator, console connector, core subsystem, and grant cards.



The M9312 console emulator, running on a VT220.
 
Congratulations! I found the same small supply unit in the h7420 supply I repaired for my 11/45. I saw a few already replaced rectifier bridges in the h7420, but none has failed when I have been running it. Maybe it is a weak spot.
 
Congratulations! I found the same small supply unit in the h7420 supply I repaired for my 11/45. I saw a few already replaced rectifier bridges in the h7420, but none has failed when I have been running it. Maybe it is a weak spot.

I was kind of wondering if it's a known point of failure, and if I should just preemptively replace all of the bridge rectifier modules in the other regulators. They're all the same device, by the way, so the Vishay replacement I found should work in the other DEC standard regulators from the H765 power system. Jack Rubin says some of the earlier PDP-8 era supplies used a 6-pin regulator, but all of the modules in mine use the 4-pin.
 
Jack Rubin says some of the earlier PDP-8 era supplies used a 6-pin regulator

The PDP-9 that I am working on has lots of 6-pin rectifier modules. They contain 4x diodes with 4x pins for the diodes, and 2x pins for a common anode & cathode connection for pairs of diodes.
 
The PDP-9 that I am working on has lots of 6-pin rectifier modules. They contain 4x diodes with 4x pins for the diodes, and 2x pins for a common anode & cathode connection for pairs of diodes.

Yeah, Jack was saying DEC just connected them up as a regular bridge rectifier. Not surprised that they were used on other DEC systems, they seemed to like to reuse power supply designs/components.
 
The PDP-9 that I am working on has lots of 6-pin rectifier modules. They contain 4x diodes with 4x pins for the diodes, and 2x pins for a common anode & cathode connection for pairs of diodes.

I had one of them blow in the PDP-9 many years ago. It was not connected as a full bridge so I wasn't able to replace it with those common bridge rectifiers. Made by Solitron G797-1. Anyone has a moden source for something similar? Two diode pairs in one unit.
 
Mattis,

I have a couple dozen of them; they were also used in the 8/I power supply, but jumpered to form a conventional 4-diode bridge. We replaced one in the CMA 8/I in Atlanta. I'll contact you directly and send you and Anders a couple of them.

Maybe Mike will trade me a few hold-down screws for an RK05 cover for a couple as well? :>)


IMG_6930.jpgIMG_6931.jpg
 
Mattis,

I have a couple dozen of them; they were also used in the 8/I power supply, but jumpered to form a conventional 4-diode bridge. We replaced one in the CMA 8/I in Atlanta. I'll contact you directly and send you and Anders a couple of them.

Maybe Mike will trade me a few hold-down screws for an RK05 cover for a couple as well? :>)


I can do that!
 
Mattis,

I have a couple dozen of them; they were also used in the 8/I power supply, but jumpered to form a conventional 4-diode bridge. We replaced one in the CMA 8/I in Atlanta. I'll contact you directly and send you and Anders a couple of them.

Maybe Mike will trade me a few hold-down screws for an RK05 cover for a couple as well? :>)


View attachment 53810View attachment 53811

Hello JacK!

That would be great! I haven't found these anywhere and there seems to be no modern replacement. The option would be to build one out of SMD mount parts on a small circuit board.. Looking at the PDP-9 schematic give that the DM1 is at least 12 amps and less than 24 amps.

Both the PDP-9 and PDP-8/L have several center tapped windings so the DM1 or DM2 is the only one that would fit.

I wonder why DEC so often used center tapped transformers? Was it cheaper to make a center tapped transformer than to use a full bridge back then?

/Mattis
 
I've been through my box of DM-1 and DM-2 diodes and can't really see any difference between them. I'm still trying to find specs on them, but no luck so far.

Using a Peak DCA455 component tester, they all show a relatively low Vf (range is generally 0.52V to 0.54V, a couple ran up to .6V and a few as low as .5V; If was in the 4.8 to 4.9 mA range).

I also ran a DM1 and a DM2 on a Syscomp curve tracer and got a more typical Vf of around .7V at 500 mA. Since these seem to be 100V 10A parts, I guess that makes sense. Curves from the Systron are basically identical in the range I could test -

DM1-DM2.JPG

As far as I can tell, the two parts are interchangeable.

Any other thoughts or ideas?
 
I think the difference between DM1 and DM2 is the maximum reverse voltage and maximum amps rating.

Vince lists DM1 and Solitron G797-0 and DM2 as Solitron G797-1. (http://so-much-stuff.com/pdp8/repair/subst.php) .

There is a now broken link to Halted web page: (https://web.archive.org/web/20051003221934/https://www.halted.com/ccp12601-bridge-g797-1-100v-10a-g797-1-9568.htm)

It lists the G797-1 as a 100V 10A bridge.

Looking at the PDP-9 PSU schematics it uses DM2 for the higher memory voltages and DM1 for the lower logic voltages. My guess is that DM1 is higher amps rating (two DM1 diodes handles the 24 Amp -15V in the PDP-9), but lower reverse voltage rating. Maybe 50V or so.

My guess is that DM1 is 15A 50V while DM2 is 10A 100V. But I cannot find any other info on them.
 
I bought my DM-2's from Halted/HSC before they went out of business (they closed in early 2019), where they were listed as G797-1. This is also where I got the 10A/100V rating, so it's really second-hand information. The power supply for my 8/L expansion box uses Motorola parts labeled DM-2 but with no other identifying number. I've looked through some of the Motorola catalogs from the period on bitsavers but haven't been able to come up with a match for the form factor or the part number.

I tried to get more info from Solitron but they couldn't help with 50 year old parts.

HSC says on their website that Excess Solutions (another Silicon Valley surplus dealer) has acquired all of their inventory, but a search of the Excess Solutions website doesn't pull up the part. I can imagine that integrating the two inventories will be a massive effort, but I'll call them next week to see if they might be able to locate the Solitron part.
 
Back
Top