• Please review our updated Terms and Rules here

My lucky blinken light day.

Hello DECfans - I figured I'd pop into this thread instead of starting a new one. Last weekend, I brought this home:

https://picasaweb.google.com/102190732096693814506/CharmillesPDP1105?authuser=0&feat=directlink

While the OP found his in Europe, where Charmille presumably configured these CNC rigs, this one made it out to suburban Chicago, USA, some ~35 years ago. I was fortunate to catch the shop owner before the 11/05 went to the scrapper with the rest of the rack it was in.

It's very similar to Mattis's find, although his seems to have more cards and I didn't get much documentation. There was an 11/04 along for the ride as well which came with some engineering schematics (I will scan everything!) and lots of diagnostic (MAINDEC) paper tape to go with the two Infranor tape-readers.

My first obstacle - after cleaning it up - is power. Being a Swiss machine, it has a BC05-U power module which, if I read the correct docs, is 230V only. It's also had its cord adapted to the odd Charmille power backplane connector (see the pics for an example.)

Is converting to 110V as simple as finding a BC05-T module? I'm new to Unibus but if I understand the way these chassis work, the H750 PSU accepts DC input from the transformer, so anything from there forward would be the same as a European model.

If that's correct, how hard will it be to find a BC05-T or other 110V drop-in replacement? The alternative, of course, is to just hack the Charmille plug off, replace it with a standard plug, buy a beefy step-up transformer and run it as-is.

The machine was demonstrated to us - at least as far as powering up and blinking a few lights - while still attached to the CNC rig power supplies. I am hoping there won't be too much more to do to it. Looking forward to demonstrating my first blinkenlights machine!
 
Great find. If you look at all the shows on TV these days about collecting and restoration it looks like that’s the ideal condition to find things in. It is a lot better than paying top dollar for something on EBay that’s sometimes has a questionable past and no accessories. The paper tape drive along with its card will be a big help. As far as the voltage requirements why not keep it original and install a 220 volt plug like you would use for an air conditioner? Don’t know if the 50 cycle thing will be an issue though.
 
As far as the voltage requirements why not keep it original and install a 220 volt plug like you would use for an air conditioner? Don’t know if the 50 cycle thing will be an issue though.

It's an option - I already have 220 installed in the basement, although that would tie the 11/05 to that location. I'm not sure about the line frequency issue either but I think it would be OK, as it's being rectified to DC anyway. I wouldn't feel like replacing a DEC part with a DEC part would harm the originality too much - it's just "localization." :)
 
Yes, sorry to also pop in on the OP's thread. I have to opposite problem to SIlent700, my 10-1/2" PDP-11/05 is 115V and I would like to find or exchange the AC input box for a 230V one. Frustratingly I have the BA-11KE 22-slot chassis and the PCU is the 7009811-1 115V whereas I need the 7009811-2 version for 230V. I guess I'll rewire it as SIlent700's unit won't be the right one.
Great finds, for both you guys.

Steve.
 
Now more than two years after the acquisition, this machine has started to execute its first instructions correctly. It needed cleaning. the PSU had several bad capacitors, new (old) fans are in place, wiring has been repaired, a PROM on the CPU control board has been replaced, a 74H00 on the core memory XYdriver board went caput. The M9301 also had a bad PROM and two DEC8881 chips failed during fixing the machine.

I am still missing the top cover and bottom cover and there is a intermittent problem loading the top four bits from the SR sometimes. But I am pretty sure I succeeded to identify a bad N8266 in the B-leg of the Data Path using cooling spray. When chilled the problem was less prominent.

I have been running through diagnostics and also loaded BASIC from PDP11GUI.

I put together a web page and a short video on it.

Now I will go on connecting a high speed reader / punch and eventually a TC11 with TU56.
 
Thanks for the very informative web and video work. You mention that the large filter cap was hard to replace - did you actually rebuild it by stuffing new components into the old shell (as your page suggests) or did you find a replacement? This power supply is also used on many other DEC devices, e.g. PDP-8/F and M, 11/05 in both chassis, some versions of 11/35, etc. I was able to find OEM replacement caps on eBay a few years ago but not sure if I still have the vendor data.

Jack
 
No, I never found a replacement. Well. I searched around for a while and then I thought that I just replace it with a modern. Finding values that were similar was not such a big problem, but then also something with the same size was more difficult. So I removed the guts of the old capacitor and put the new one inside. It was a hard job to remove all the things inside...
 
Going bigger on the output caps is OK. If they are below 6000uF for the +5V or 3000 uF for the -15V the switching circuit will not work.
 
These are the input filter caps. 24000mfd 50V. It seems like Mouser actually carry them but minimum quantity is 49 and the price is more than 50 USD. It seems though that Nebraska Surplus sales also has them: Price 18USD.

The size are 2" x 5-5/8" (51mm x 143mm)

cfe-024000R00ABA_1_lg.jpg


But I didn't know of Nebraska Surplus sales when starting to work on the PSU. Since I now have at least 4 more of this PSUs in various boxes it could be a good idea to get a few spare capacitors.
 
Which power supplies use this specific cap?

The one in low PDP-11/05, the 11/05-NC (BA11-D with H750 PSU), the 11/35 in BA11-D, MF11 in low chassis. PDP-8/F, PDP-8/M.

A lot of places in early 1970-ies machines. An ex DEC engineer said this PSU was a real pain.

VXPug0d.png


It is the PSU to the right with big caps.

Micheal Thompson has done a spice model on it. Maybe we should have a new thread on this PSU so Michael could give some info regarding his SPICE model. I recognized that the one manufactured in 1978 differs in components and values from the one in made in 1973.
 
I have had one of these supplies used in an 8/M fail in the input rectifier diode bridge; one of the diodes shorted. The failure mode is dramatic, very high input AC current causing circuit breaker to trip. Took quite a while to find the fault, as power on debugging was not possible. Lots of unsoldering to isolate subsystem components. Luckily bridges with compatible footprints are still available from major suppliers (Mouser, etc).

Debugging this supply in an 8/M is a royal pain. Not possible just to pull the supply and replace, need to pull the backplane to slide the supply out. Basically means complete disassembly/reassembly of the box to get to the power supply, and you better have fixed it before you put it back in, otherwise rinse and repeat.

Don
 
Debugging this supply in an 8/M is a royal pain. Not possible just to pull the supply and replace, need to pull the backplane to slide the supply out. Basically means complete disassembly/reassembly of the box to get to the power supply, and you better have fixed it before you put it back in, otherwise rinse and repeat.

Don

Since the DW8E is nearly the same chassis as an 8/M, we had the same experience with repairing and debugging. After about the 10th time we could remove the power supply in just a few minutes.

An failure mode that we observed was that the -15 regulator will behave like a linear and will almost work, but can't regulate under a normal load. This is an indication that the regulator is not switching, usually due to too small capacitance output caps.

If you are reviving one of these power supplies that has not been used in a long time, it is very important to reform the output caps for several days.
 
Back
Top