9 Apr
2013
9 Apr
'13
5:04 p.m.
Started tearing into the 11/780 and pulled the 11/03 and the RX01 out to clean and test.
Looks like the PS is bad in the 11/03. It powers on and the fans run because they are AC
but I am getting sometimes 4.5 volts on the 5v line and 0 volts on the 12v. I have 34
volts coming out of the diode rectifier. I suspect switching transistors.
Anyone have experience with these?
Brian.
9 Apr
9 Apr
5:50 p.m.
On Apr 9, 2013, at 20:04, Brian Roth <abacos_98 at yahoo.com> wrote:
Started tearing into the 11/780 and pulled the 11/03
and the RX01 out to clean and test. Looks like the PS is bad in the 11/03. It powers on
and the fans run because they are AC but I am getting sometimes 4.5 volts on the 5v line
and 0 volts on the 12v. I have 34 volts coming out of the diode rectifier. I suspect
switching transistors.
Anyone have experience with these?
Not H780s specifically, but switching supplies in general.
Are you putting any kind of load on the supply? Lots of
switchers don't like running without some minimum load
(I think DEC often specced the minimum load in the
manual).
For low voltage out of the rectifier diodes, the switching
transistors aren't a bad bet. Filter caps are also a possible
culprit, though they're more likely to fail open or lose capacitance.
Check them with an ohmmeter anyway.
- Dave
10 Apr
10 Apr
3:46 a.m.
Thanks Dave,
?? I checked first with a load and then without. I have the main board out of it now so I
was going to replace the transistors and caps. The problem is the schematics on Bitsavers
is just a little blurry so I cannot make out the names and the transistors themselves(a
few of them) are unreadable.
Brian.
________________________________
From: David Riley <fraveydank at gmail.com>
To: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
Sent: Tuesday, April 9, 2013 8:50 PM
Subject: Re: H780 power supply
On Apr 9, 2013, at 20:04, Brian Roth <abacos_98 at yahoo.com> wrote:
Started tearing into the 11/780 and pulled the 11/03
and the RX01 out to clean and test. Looks like the PS is bad in the 11/03. It powers on
and the fans run because they are AC but I am getting sometimes 4.5 volts on the 5v line
and 0 volts on the 12v. I have 34 volts coming out of the diode rectifier. I suspect
switching transistors.
Anyone have experience with these?
Not H780s specifically, but switching supplies in general.
Are you putting any kind of load on the supply? Lots of
switchers don't like running without some minimum load
(I think DEC often specced the minimum load in the
manual).
For low voltage out of the rectifier diodes, the switching
transistors aren't a bad bet. Filter caps are also a possible
culprit, though they're more likely to fail open or lose capacitance.
Check them with an ohmmeter anyway.
- Dave
2:09 p.m.
Started tearing into the 11/780 and pulled the 11/03 and the RX01 out
to clean and test. Looks like the PS is bad in the 11/03. It powers on
and the fans run because they are AC but I am getting sometimes 4.5
volts on the 5v line and 0 volts on the 12v. I have 34 volts coming out
of the diode rectifier. I suspect switching transistors.
Anyone have experience with these?
Probably not that particualr PSU, but I know that some of the smaller DEC
pSUs used on Qbus systems do not work correctly unloaded. But the normal
behaviour then is for the 5V rail to come up but the +/-12V rails to be
very low -- they 'tag along' with the 5V rail so if the latter is not
loarded the PSU is running at very low power, the choppers are on for
very little time so the 12V liens don't come up.
I would try loading the 5V line (a 6V 12W bulb should do it) and see what
happens.
-tony
5:21 p.m.
Probably not that particualr PSU, but I know that some
of the smaller
DEC pSUs used on Qbus systems do not work correctly unloaded.
It wasn't a DEC PSU, but I once fried a PSU by running it with too
little load. (It was a Sun SPARCstation SLC or ELC, I forget which,
and I was using the PSU without the monitor load it was designed to
always drive - http://www.sunhelp.org/~mouse is a page I wrote up for
sunhelp.org which, among other things, mentions the experience.)
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
6:27 p.m.
On Apr 10, 2013, at 8:21 PM, Mouse <mouse at rodents-montreal.org> wrote:
Yeah, if you run some of the less intelligent ones with two low a
load, the duty cycle on the switcher shrinks to near zero, which
means lots of sharp pulses up/down next to each other on an
inductor. The end result is not nice. The smarter systems will
just hold off for a period if they detect an insufficient load;
the even smarter ones will switch in a dummy load automatically.
- Dave
Probably not
that particualr PSU, but I know that some of the smaller
DEC pSUs used on Qbus systems do not work correctly unloaded.
It wasn't a DEC PSU, but I once fried a PSU by running it with too
little load. (It was a Sun SPARCstation SLC or ELC, I forget which,
and I was using the PSU without the monitor load it was designed to
always drive - http://www.sunhelp.org/~mouse is a page I wrote up for
sunhelp.org which, among other things, mentions the experience.)
15 Apr
15 Apr
6:24 p.m.
From: David Riley <fraveydank at gmail.com>
To: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
Sent: Wednesday, April 10, 2013 9:27 PM
Subject: Re: H780 power supply
?
Yeah, if you run some of the less intelligent ones with
two low a
load, the duty cycle on the switcher shrinks to near zero, which
means lots of sharp pulses up/down next to each other on an
inductor.? The end result is not nice.? The smarter systems will
just hold off for a period if they detect an insufficient load;
the even smarter ones will switch in a dummy load automatically.
- Dave
?? Well my new caps came in tonight. They were kind of? odd values so I had
to order them. I put them in, assembled it and success! DC light right on and after
getting the baud right, a prompt!
http://www.vintage-computer.com/vcforum/attachment.php?attachmentid=12819&a…
Brian.
7:46 p.m.
On Apr 15, 2013, at 9:24 PM, Brian Roth <abacos_98 at yahoo.com> wrote:
I didn't realize until your reply that my phone auto-corrected
"too low" to "two low". How embarrassing.
In any case, congrats! I'm really excited to see the whole beast
come together.
- Dave
From: David
Riley <fraveydank at gmail.com>
To: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
Sent: Wednesday, April 10, 2013 9:27 PM
Subject: Re: H780 power supply
Yeah, if you run some of the less intelligent
ones with two low a
load, the duty cycle on the switcher shrinks to near zero, which
means lots of sharp pulses up/down next to each other on an
inductor. The end result is not nice. The smarter systems will
just hold off for a period if they detect an insufficient load;
the even smarter ones will switch in a dummy load automatically.
- Dave
Well my new caps came in tonight. They were kind of odd values so I had
to order them. I put them in, assembled it and success! DC light right on and after
getting the baud right, a prompt! 
16 Apr
16 Apr
4:46 a.m.
I didn't realize until your reply that my phone auto-corrected
"too low" to "two low". How embarrassing.
It's done it again. It changed "auto-goofed" to
"auto-corrected".
Orwell was not far wrong about television and newspeak but he failed to predict
that nearly everyone would have a personal tracking device which automatically
censored their communications :-)
Regards,
Peter Coghlan.
17 Apr
17 Apr
12:42 p.m.
I didn't realize until your reply that my phone auto-corrected
"too low" to "two low". How embarrassing.
It's done it again. It changed "auto-goofed" to
"auto-corrected". 
6:17 p.m.
On Wed, 17 Apr 2013, Tony Duell wrote:
I do hate machines that think they know better than I
do what I actually
want. It's fine to warn me that a word was not found in the dictionary,
but not to change it. After all, a lot of what I type involves signal
names, variable names, assembly language mnemonics, etc. Those are
unlikely to be found in a (smartphone) dictionary, and changing them will
make nonsnse of the text.
About three decades ago, I designed a spell checking program with
heuristically developed lexicon.
'course when you switch to another set of mnemonics that you haven't used
before, it would call your attention to each.
--
Grumpy Ol' Fred cisin at xenosoft.com
18 Apr
18 Apr
1:23 p.m.
About three decades ago, I designed a spell checking
program with
heuristically developed lexicon.
'course when you switch to another set of mnemonics that you haven't used
before, it would call your attention to each.
I have no problem with a machien claleing attention to a possible error
(in this case, that it can't find the word in its dictionary). I do have
a problem if it chaged 'LDA' into 'Lad', 'muClk' into
'mulch' and so on
without telling me.
-tony
5:16 p.m.
On Thu, 18 Apr 2013, Tony Duell wrote:
I run in to that issue when my name gets changed to "Cory
Smells"...sometimes I skim when checking for spelling errors and I don't
expect my name to be changed without asking me. ;)
About three decades ago, I designed a spell
checking program with
heuristically developed lexicon.
'course when you switch to another set of mnemonics that you haven't used
before, it would call your attention to each.
I have no problem with a machien claleing attention to a possible error
(in this case, that it can't find the word in its dictionary). I do have
a problem if it chaged 'LDA' into 'Lad', 'muClk' into
'mulch' and so on
without telling me.
-tony
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
17 Apr
17 Apr
9:55 p.m.
On Apr 17, 2013, at 21:42 , Tony Duell <ard at p850ug1.demon.co.uk> wrote:
As an aside, One 'famous' autocorrect (which
may not actually have
happeend) was that 'Dear Sir or Madam' (standard opening for a letter)
became 'Dear Sir or Madman'
Two autocorrects (or spell correction suggestions) that really *did* happen were
Ericsson Telecom -> Erection Talcum
whitehouse -> whorehouse
4266
days inactive
4274
days old
13 comments
8 participants
participants (8)
-
abacos_98＠yahoo.com -
ard＠p850ug1.demon.co.uk -
b4＠gewt.net -
cctech＠beyondthepale.ie -
cisin＠xenosoft.com -
fraveydank＠gmail.com -
mouse＠Rodents-Montreal.ORG -
rwiker＠gmail.com