PDP 11 gear finally moved
lbickley at bickleywest.com
Thu Jul 23 02:04:06 CDT 2015
On Wed, 22 Jul 2015 09:01:35 -0400 (EDT)
jnc at mercury.lcs.mit.edu (Noel Chiappa) wrote:
> > From: Tothwolf
> > How do you know those aluminum electrolytic capacitors are functioning
> > just as good as they did when they were new? Unless you've tested them
> > out of circuit ...
> > ... aluminum electrolytic capacitors by their very electrochemical
> > nature degrade as they age and as they are used.
> I am way out of my knowledge range in this discussion, but here's something I
> wanted to ask about: how do you reconcile this observation (assertion?) with
> the observations from several people (e.g. the PDP-1 people) that they _have_
> measured the electrolytics in their power supplies, and despite being N
> decades old (where N ~= 5), they are _still_ within specs? If the very nature
> of electrolytics mandates that they degrade, how are these still meeting
> I'm very confused...
This is the last time I'm going to discuss this subject.
IMHO, there's been way too much emotion expressed on what is a rather straightforward subject.
The PDP-1 team is composed of seasoned engineers. Combined we have over 250 years of experience in high tech work and restoration. One Team member is a senior engineer at Tesla, two others are lead engineers in a different high tech medical startups/firms. Others of us are consultants to multiple Silicon Valley startups and companies. Almost all of us are serious vintage computer collectors.
The PDP-1 at the CHM is demoed regularly for the public - and we run Spacewar! tournaments every quarter which are "sold out" (there is no fee - just signup). The PDP-1 is also regularly demoed regularly by some of us for friends and special guests of the Museum (or personal friends).
Since the PDP-1 at the CHM is the only running PDP-1 in the world - and the fact that the Museum considers it a highly significant artifact - we made restoration decisions with those factors in mind.
In accordance with standard Museum protocol, we tagged every bad part we removed from the PDP-1, bagged it and tagged it with the specific location it was removed from. We tagged all new (old stock) parts with a red dot. (In Museum protocol, one has to be able to restore an artifact to it's original condition - even if it is anticipated that will never occur). The CHM as other major Museums (like the Smithsonian) follow this protocol.
We spent months laying out the restoration plan for the PDP-1 - and then spent about six months restoring all the power supplies. We made the easy decision to reform the capacitors in the PDP-1 as they were all computer grade - and had not visually leaked or "dried out".
We used a very careful automated process to reform the capacitors which included a programmable power supply, current sensors and a laptop scripted to monitor the reformation progress. It was probably overkill - but we wanted to make sure we didn't overload any capacitor in the reformation process.
After reforming every capacitor in every power supply, we tested each capacitor for capacitance and ESR. Only four of the reformed capacitors did not pass these tests.
Two of us created an identical load test system that DEC had used to test the PDP-1 power supplies (multiple voltages, various currents, etc.). We then tested each power supply per DEC factory spec. starting at 85V and going as high as the spec. stated. Each power supply had to meet it's specifications in terms of voltage linearity and ripple (every test was logged and documented).
IIRC, the above process took almost six months. (While we were working on power supplies, others were working on restoring fans (total strip down and restoration), checking every soldered connection, etc., etc.
After we got the system up and passing all PDP-1 diagnostics - we went to work on the peripheral I/O gear - which took as long to restore as the PDP-1 CPU. The Model 30 display was particularly challenging. (We also reformed all the capacitors in the Model 30 display).
Subsequently to the complete restoration, We ran regular demos on the PDP-1 - and during the first year we had a few minor bugs crop up - mostly memory and display related.
The PDP-1 had been running demos for about 10 years and we've probably only had a handful of bogs over that period. We have NEVER had a P/S capacitor fail.
Every year we do a complete DEC specified preventative maintenance on the PDP-1 which includes testing every power supply for voltage, stability and ripple. Of course we do many other tests as well - including running of all PDP-1 diagnostics. Almost every year the PDP-1 passes all tests flawlessly.
Now to my own systems. For those of you who have checked my website, you'll see that I have a modestly large collection of systems. When I do restorations, I typically keep a logbook as we did on the CHM's PDP-1. I also reform all capacitors on my systems using a careful (but not as sophisticated as the PDP-1) methodology (which I've described elsewhere on this list). I then test the capacitors for capacitance and ESR before reinstalling them. In my many years of restoration, I have never had to replace any capacitor that passed my reform process.
Now to the "flip" side. I also restore vintage radios (I am an Amateur Extra Class HAM). Vintage receivers do NOT have computer grade capacitors. Most used the best "low cost" parts available. It is my common practice to wholesale replace their electrolytic and paper capacitors - as most will fail just about any test you throw their way - and seldom (if ever) does reforming work.
So IMHO, the story is: One can't get "religious" about forming capacitors or not. Each restoration has it's own characteristics that must be taken into consideration...
Bickley Consulting West Inc.
"Black holes are where God is dividing by zero"
More information about the cctech