PDP-8/e
Rod Smallwood
I recently aquired a complete PDP-8/e. As it had not run for a while I took
the usual precautions of reforming the electrolytics in the PSU as well as
dismantling the whole system and cleaning everything.
I was rewarded with a working system. I started to run some simple toggle in
programs and got to the stage of driving a VT 320 with a display of whole
the character set at 110baud.
I was about to toggle in the next test which would have been a keyboard echo
routine when I discovered I could no longer load anything into memory.
Load address works but setting the data to load into memory in the switch
register and raising DEP no longer writes the data to memory.
.
RUN works so it looks like the timing plus the RUN/STOP flipflop are running
ok.
However one shot operations like DEP and EXAM are difficult to diagnose as
is hard to see a one off 500nS pulse.
I don't think its the memory itself ( 4 x 4K of core) as I have tried each
4K bank on its own with the same result.
I do have the CPU manual and a decent Tektronix scope.
Has anybody any suggestions?
Rod Smallwood
David Gesswein
Rod Smallwood <rodsma...@btconnect.com> wrote:
>
>Load address works but setting the data to load into memory in the switch
>register and raising DEP no longer writes the data to memory.
>
the power good from the power supply. Memory operation is shut down
when the signal indicates power is bad.
Also check the dep switch itself. I had to clean the contacts on mine
when it got difficult to get it to deposit.
>However one shot operations like DEP and EXAM are difficult to diagnose as
>is hard to see a one off 500nS pulse.
>
For some of the signals in the sequence all you want to know is did the
signal occur or not so you can either use a logic probe or set the scope
to normal trigger and see if it triggers when you push the switch.
Quadibloc
wrote:
> Hi
>
> I recently aquired a complete PDP-8/e. As it had not run for a while I took
> the usual precautions of reforming the electrolytics in the PSU as well as
> dismantling the whole system and cleaning everything.
>
> I was rewarded with a working system. I started to run some simple toggle in
> programs and got to the stage of driving a VT 320 with a display of whole
> the character set at 110baud.
Wow and congratulations.
Also, I wasn't familiar with what "reforming the electrolytics" meant,
but now that I knew of it, I could do a search and find the details:
http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm
John Savard
Peter Flass
Very interesting.
Rich Alderson
Quadibloc <jsa...@ecn.ab.ca> writes:
> On Mar 23, 7:55 am, "Rod Smallwood" <rodsmallw...@btconnect.com> wrote:
>> I recently aquired a complete PDP-8/e. As it had not run for a while I took
>> the usual precautions of reforming the electrolytics in the PSU as well as
>> dismantling the whole system and cleaning everything.
> Also, I wasn't familiar with what "reforming the electrolytics" meant,
> but now that I knew of it, I could do a search and find the details:
> http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm
That's a good explanation, as far as it goes.
My colleague Keith Perez and I maintain a large collection of mainframe
computers which use a large number of electrolytic "cans". Through our
experience and research, we have learned that it is better to discard a
capacitor more than 14 years old rather than trying to reform it. The
electrolytes used go through irrevocable chemical changes rendering them
useless, and potentially dangerous if the failed capacitor is used.
If the capacitor does not develop a dead short (and end up destroying part
of the circuit in which it resides), it will allow an AC waveform onto a DC
circuit, playing havoc with signals. (We found a 5V ripple on the 20V lines
in a PDP-11/40 memory, traced back to a dead cap.)
We no longer attempt to power up any system before we replace all aluminum
electrolytics in the box. In some cases we have pulled caps with date codes
from 1977.
I urge anyone who is running an older system with reformed caps to check the
date codes and replace anything older than 14 years. If testing indicates
that a newer cap should be changed, please do that as well. Safety first!
(Keith sent a version of this same post to the maintainer of the web page
cited by J. Savard.)
--
Rich Alderson "You get what anybody gets. You get a lifetime."
ne...@alderson.users.panix.com --Death, of the Endless
Mensanator
wrote:
> [added alt.sys.pdp11 because there are people doing the same thing there]
>
I once saw a mere thumb sized cap explode sending
a trail of aluminum foil six inches in front of my
nose. That put the fear of God into me when seeing
beer can or larger caps!
>
> (Keith sent a version of this same post to the maintainer of the web page
> �cited by J. Savard.)
>
> --
> Rich Alderson � � � � � � � � �"You get what anybody gets. You get a lifetime."
> n...@alderson.users.panix.com � � � � � � � � � � � � � --Death, of the Endless
CBFalconer
> Rich Alderson <n...@alderson.users.panix.com> wrote:
>
... snip ...
>>
>> I urge anyone who is running an older system with reformed caps
>> to check the date codes and replace anything older than 14 years.
>> If testing indicates that a newer cap should be changed, please
>> do that as well. Safety first!
>
> I once saw a mere thumb sized cap explode sending a trail of
> aluminum foil six inches in front of my nose. That put the fear
> of God into me when seeing beer can or larger caps!
We had the same problem 50+ years ago. Except that then it
happened with brand new capacitors out of the storeroom. Always
fire up the PS first. Then call the maintenance staff to fix the
roof.
--
[mail]: Chuck F (cbfalconer at maineline dot net)
[page]: <http://cbfalconer.home.att.net>
Try the download section.
--
Posted via a free Usenet account from http://www.teranews.com
Chuck Harris
> [added alt.sys.pdp11 because there are people doing the same thing there]
>
> Quadibloc <jsa...@ecn.ab.ca> writes:
>
>> On Mar 23, 7:55 am, "Rod Smallwood" <rodsmallw...@btconnect.com> wrote:
>
>>> I recently aquired a complete PDP-8/e. As it had not run for a while I took
>>> the usual precautions of reforming the electrolytics in the PSU as well as
>>> dismantling the whole system and cleaning everything.
>
>> Also, I wasn't familiar with what "reforming the electrolytics" meant,
>> but now that I knew of it, I could do a search and find the details:
>
>> http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm
>
> That's a good explanation, as far as it goes.
>
> My colleague Keith Perez and I maintain a large collection of mainframe
> computers which use a large number of electrolytic "cans". Through our
> experience and research, we have learned that it is better to discard a
> capacitor more than 14 years old rather than trying to reform it. The
> electrolytes used go through irrevocable chemical changes rendering them
> useless, and potentially dangerous if the failed capacitor is used.
I have not observed this phenomenon in the large numbers of very old
electrolytic capacitors I have currently in use. Most of the capacitors
in the equipment I use are more than 30 years old.... some as old as
60 years.
Do you have a reference for this supposed "irrevocable chemical changes"?
In my 25 years as an EE, that reforms electrolytic capacitors, I have
found 4 different cases:
1) the capacitor reforms, and the current drops to near zero.
2) the capacitor's leakage current will not drop.
3) the capacitor is dried out, and has little or no capacitance
4) the internal connection to the + or - terminal has been etched
into an open circuit.... a common problem on "computer grade"
electrolytic capacitors... the big caps with screw terminals.
I reform capacitors using the following sequence:
0) Never power up the unit before reforming! To do so will destroy
any hope you have of returning the capacitors to new function.
1) I test for capacitance. If the capacitance is drastically wrong
(eg. less than 1/2, or more than 5x), I discard the cap.
2) I test for ESR, and make note of the reading.
3) I charge the capacitor to WV using a current limited supply for 1 to 2 hours.
If the leakage current doesn't drop to near zero in that time, I discard
the capacitor.
4) I retest the ESR, it should be within manufacturer's specified limits, and
lower than the original test. If it is not, out it goes.
What does *not* work:
0) firing up the unit and hoping for the best.
1) bringing up the supply with a variac at reduced voltage.
2) hoping that a capacitor that finishes reforming too leaky will get
better in use.
3) hoping that a capacitor whose ESR is too high will do the job anyway.
I suspect that you and your colleagues have a bad opinion of reforming
capacitors because you have only tried #1 in the "What does *not* work
column. Most people like #1 because it requires little or no effort.
Those that have seen success using the variac have been lucky to have
capacitors that didn't actually need reforming... which I would guess
to be better than 70% of the capacitors out there in older equipment.
[#1 does not work because the capacitor will go through moments of very
low resistance during the reforming process. If you try and reform using
a variac, those moments will cause the capacitor to draw excessive current
and dig a pit (hole) into the aluminum foil where it needed reforming.]
I have had long success reforming capacitors, and using reformed
capacitors.
-Chuck Harris
Morten Reistad
CBFalconer <cbfal...@maineline.net> wrote:
>Mensanator wrote:
>> Rich Alderson <n...@alderson.users.panix.com> wrote:
>>
>... snip ...
>>>
>>> I urge anyone who is running an older system with reformed caps
>>> to check the date codes and replace anything older than 14 years.
>>> If testing indicates that a newer cap should be changed, please
>>> do that as well. Safety first!
>>
>> I once saw a mere thumb sized cap explode sending a trail of
>> aluminum foil six inches in front of my nose. That put the fear
>> of God into me when seeing beer can or larger caps!
>
>We had the same problem 50+ years ago. Except that then it
>happened with brand new capacitors out of the storeroom. Always
>fire up the PS first. Then call the maintenance staff to fix the
>roof.
I was always in awe to the Prime-OEM'ed Kennedy 9 track tape drives,
with their unbelievable accelration of the quite substantial mass
of a tape reel from standstill to fast in less time than you could
register. Now it moves. Now it doesn't. Now it moves again.
It had 6x1F (yep, farad, without prefix) 480 volt capacitors to
help store the energy. We had 4 of those drives right next to
each other. That is 24 Farad. We planned for the inevitable
big bang, but it never materialised.
I had left that PPOE when they were decommishioned. Shame. I
wanted to make a flashlight of of those caps and a neon light.
It probably would have made a direct phase transition from solid
to plastma.
-- mrr
glen herrmannsfeldt
(snip)
> My colleague Keith Perez and I maintain a large collection of mainframe
> computers which use a large number of electrolytic "cans". Through our
> experience and research, we have learned that it is better to discard a
> capacitor more than 14 years old rather than trying to reform it. The
> electrolytes used go through irrevocable chemical changes rendering them
> useless, and potentially dangerous if the failed capacitor is used.
If the vent seal has broken then the capacitor is bad. If it dries
up it will be either open or closed circuit, both are not good.
> If the capacitor does not develop a dead short (and end up destroying part
> of the circuit in which it resides), it will allow an AC waveform onto a DC
> circuit, playing havoc with signals. (We found a 5V ripple on the 20V lines
> in a PDP-11/40 memory, traced back to a dead cap.)
> We no longer attempt to power up any system before we replace all aluminum
> electrolytics in the box. In some cases we have pulled caps with date codes
> from 1977.
Doesn't sound like a bad idea. Probably especially important for larger
ones which can make a bigger mess when they explode.
> I urge anyone who is running an older system with reformed caps to check the
> date codes and replace anything older than 14 years. If testing indicates
> that a newer cap should be changed, please do that as well. Safety first!
If a reformed cap has a sufficiently low leakage current it should be
pretty much as good as can be. If the electrolyte is bad, then it
likely won't have a good oxide and will have leakage current.
My now 29 year old stereo amplifier has had two capacitors replaced
in the power on circuit. (That delays closing a relay to the speakers
until the power supply is stable.) I presume those have a higher than
usual AC current through them and go bad faster. (10uF and 100uF).
I haven't felt like replacing the rest of them. It isn't easy to
take apart, either.
-- glen
Christian Corti
> computers which use a large number of electrolytic "cans". Through our
> experience and research, we have learned that it is better to discard a
> capacitor more than 14 years old rather than trying to reform it. The
> electrolytes used go through irrevocable chemical changes rendering them
> useless, and potentially dangerous if the failed capacitor is used.
I've seldom read such a non-sense...
Through our experience and research here, we have learned that
capacitors more than 14 years are mostly like new. And never has any
electrolytic cap exploded... the oldest caps in a computer here are
from 1958, and are still fine. When we got our second LGP-30 we tried to
reform the electrolytics in the power supply (300V/2A and other
voltages) which has many big caps rated at 100uF/350V, but this was
completely senseless as they were (and are, the machine runs fine) like
new. You will have more problems with oil filled MP capacitors though.
You may be thinking of electrolytic caps used in 30s radio sets that
tend to have current leaks which will e.g. burn out the rectifier tube.
> If the capacitor does not develop a dead short (and end up destroying part
> of the circuit in which it resides), it will allow an AC waveform onto a DC
> circuit, playing havoc with signals. (We found a 5V ripple on the 20V lines
> in a PDP-11/40 memory, traced back to a dead cap.)
A dead short (happened a few here) will open the fuse in the power
supply (like in the power supply bricks from DEC). The only caps that
I've seen "exploding" (but harmlessly) are tantalum caps and foil caps
in line filters.
You can either try replacing the cap (a new one is *very* expensive,
around EUR 20), or you can try to remove the short with a high current
pulse (done successfully several times) and measure the capacity and
other parameters afterwards. I have never had any trouble with such
"repaired" capacitors in a system.
On the other hand, dead caps (i.e. without any sign of capacity) should
obviously be replaced.
> We no longer attempt to power up any system before we replace all aluminum
> electrolytics in the box. In some cases we have pulled caps with date codes
> from 1977.
Why don't you just junk everything in the box and rebuild it from ground
up? We are not talking about modern PC power supplies with high
frequency high power load changes and leaky cheap chinese "low-ESR" caps.
> I urge anyone who is running an older system with reformed caps to check the
> date codes and replace anything older than 14 years. If testing indicates
> that a newer cap should be changed, please do that as well. Safety first!
I urge anyone not to do that but use your brain and experience. *If* a
cap is evidently bad, then replace it. If not, then don't replace it.
Christian
Tim Radde
stuttgart.de> wrote:
I have to agree. I have several older systems from the early 70s and
have only had
to replace on of the big caps in the PS. I had some serious ripple on
the +5v
line and check the caps and this one just wouldn't hold a charge very
long.
Replaced it with another of the same vintage and everything is fine.
Why
chuck them if they can be reformed?
Tim R
glen herrmannsfeldt
(snip)
> I've seldom read such a non-sense...
> Through our experience and research here, we have learned that
> capacitors more than 14 years are mostly like new. And never has any
> electrolytic cap exploded... the oldest caps in a computer here are
> from 1958, and are still fine. When we got our second LGP-30 we tried to
> reform the electrolytics in the power supply (300V/2A and other
> voltages) which has many big caps rated at 100uF/350V, but this was
> completely senseless as they were (and are, the machine runs fine) like
> new.
It might be very different for low voltage electrolytics than
for high voltage ones. You will find that the size of an
electrolytic depends on the product of capacitance and voltage.
The construction is very much the same, the difference being the
thickness of the anodic oxide on one of the plates. Higher voltage
means thicker oxide, which means lower capacitance per unit area.
Also, the higher voltage power supplies will usually have a larger
resistor in series. Powering up slowly with a variac may work if
done slow enough and the series resistor can limit the current
appropriately. A true current limited power supply is a better
choice.
-- glen
Philipp Hachtmann
> I've seldom read such a non-sense...
> Through our experience and research here, we have learned that
> capacitors more than 14 years are mostly like new. And never has any
> electrolytic cap exploded...
Never? I have had that a few times..
> A dead short (happened a few here) will open the fuse in the power
> supply (like in the power supply bricks from DEC).
"to open the fuse" - a great euphemism :-)
>> We no longer attempt to power up any system before we replace all aluminum
>> electrolytics in the box. In some cases we have pulled caps with date codes
>> from 1977.
Why? Burn them! :-)
>> I urge anyone who is running an older system with reformed caps to check the
>> date codes and replace anything older than 14 years. If testing indicates
>> that a newer cap should be changed, please do that as well. Safety first!
>
> I urge anyone not to do that but use your brain and experience. *If* a
> cap is evidently bad, then replace it. If not, then don't replace it.
>
I fully agree!
I have never reformed a capacitor or used a variac. When I get a
(valuable) new unit, I disconnect the PSU and fire it up with full
voltage. Then I measure the outputs. If everythink looks fine, I simply
reconnect it. Sometimes a capacitor goes bad, with more or less
explosion and smoke. Sometimes taking a few diodes along its way to
hell.. But... The overall damage (to my collection) by less than one
exploded cap/year is far smaller than the impact (not at last the
financial) of testing and changing masses of capacitors...
Best wishes,
Philipp :-)
kreu...@gmail.com
Hi Rod,
look for a digitizer with a usb2.0 interface. There a lot around. Its
like a modern digital skope and you get a fine picture of your single
shot. This boxes arent very expensive and a Laptop is always around.
Gerhard
kreu...@gmail.com
Hi,
I got a name now: Handyscope Hs3 or Hs4, but there still others around
on the market, its a simple digital storage oszi, the displayunit is
your laptop.
With best regards
Gerhard