15 Jun
2014
15 Jun
'14
4:23 p.m.
All,
I have an (untested) H7140 here with the bias/interface module from my
dead one...took me a bit to get it in...apparently the aligning rails got
knocked loose.
Now, it comes up ad the DC OK light blinks (there is no load, I haven't
done voltage checks and I'm not ready for a load yet). However...the fans
don't spin and I measure 0VAC there. Last night the wires had gotten
caught on something, cut a bit and ended up shorted together...I hadn't
noticed as the tag covered up the damage...but I fixed that today. I
wouldn't imagine there was protection missing there...I never heard
arcing, anything frying, and I didn't see current draw shoot up.
Any ideas? I don't want to put ANY load on this until I have the fans
working...not even a dummy load.
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
16 Jun
16 Jun
1:12 p.m.
All,
I have an (untested) H7140 here with the bias/interface module from my
dead one...took me a bit to get it in...apparently the aligning rails got
knocked loose.
Now, it comes up ad the DC OK light blinks (there is no load, I haven't
done voltage checks and I'm not ready for a load yet). However...the fans
don't spin and I measure 0VAC there. Last night the wires had gotten
caught on something, cut a bit and ended up shorted together...I hadn't
noticed as the tag covered up the damage...but I fixed that today. I
wouldn't imagine there was protection missing there...I never heard
arcing, anything frying, and I didn't see current draw shoot up.
Any ideas? I don't want to put ANY load on this until I have the fans
working...not even a dummy load.
YEs...
The H7140 is a bit strange. No, let me put that anotehr way. It's
totally uncovnentional.
The first thing is to have a clear idea od the secions of the PSU :
Incoming mains is rectifieed/smoothed to give about 350V DC This supplies
3 other secisons
There is a little SMPSU (of a fairly conventional design) to provide the
power for the cotnrol circuitry. The transformer is on the packplane
board, the chopper and control circuitry are on the bias/interface board
There is the 'logic' suppky. A fairly normal SPMSU with all control
circuitry o nthe output ( not mains) side of the isolation barrier. This
supplies +5V, +/-15V for the lgoic
Lastly there is the meory supply. This is another SM{SU (of a similar
design to the logic PSU) that provides 36V. This 36V can be
battery-backed. It is then regulated down to +/-12V and +/-5V for the
memory boards usign non-isolating switchign regulators.
Now for the fans. As is well-known these are special, 35V 70Hz ones. They
are run from the 36V DC rail i nthe meoroy supply via a full-H driver
circuit (4 power tranisotrs on the bias/interface board) and associaled
bits. The idea of that is that the fans are also abttery-backed.
Now, there is no protection on the fan driver circuit. So any short
between the fn leads of fro ma fan lead to chaiss will do damage. In
particular it burns out the transisotrs i nthe full-H circuit.
I would start by checkig nthat 36V supply is presnet and correct. There's
a capactior on brackets conencted to the 2 boards in the memory supply.
Check acoross that. TAKE GREAT CARE when working on this PSU, the
rectified mains is carried on some of the barrier strip blocks on the top
of the PCBs. And it is iethal. Seriously lethal. before touchign anything
i nthe PSU, swithc off, pull the mains plug, wait a few minutes, then
cleck the ovltage beween the wires on the 2-way pbarirer blocks. Only
start woking on things when that voltage is too low to be dangerous.
If that 36V supply is fine, then check the 4 apporpriate transsitors on
the bias/interface board. Most likely soem are damaged.
-tony
2 p.m.
On Mon, 16 Jun 2014, Tony Duell wrote:
YEs...
The H7140 is a bit strange. No, let me put that anotehr way. It's
totally uncovnentional.
That I know for sure. ;)
The first thing is to have a clear idea od the secions
of the PSU :
Yeah, that helps.
There is a little SMPSU (of a fairly conventional
design) to provide the
power for the cotnrol circuitry. The transformer is on the packplane
board, the chopper and control circuitry are on the bias/interface board
Ahh. Logic starts on the module where I don't know if it's good or not.
Now for the fans. As is well-known these are special,
35V 70Hz ones. They
are run from the 36V DC rail i nthe meoroy supply via a full-H driver
circuit (4 power tranisotrs on the bias/interface board) and associaled
bits. The idea of that is that the fans are also abttery-backed.
Ahhhh! Also on the bias/interface board! That module might very well be
bad...which would explain a lot of the problems.
Now, there is no protection on the fan driver circuit.
So any short
between the fn leads of fro ma fan lead to chaiss will do damage. In
particular it burns out the transisotrs i nthe full-H circuit.
Ahh. I imagine they're not uncommon power transistors?
I would start by checkig nthat 36V supply is presnet
and correct. There's
a capactior on brackets conencted to the 2 boards in the memory supply.
Check acoross that. TAKE GREAT CARE when working on this PSU, the
rectified mains is carried on some of the barrier strip blocks on the top
of the PCBs. And it is iethal. Seriously lethal. before touchign anything
i nthe PSU, swithc off, pull the mains plug, wait a few minutes, then
cleck the ovltage beween the wires on the 2-way pbarirer blocks. Only
start woking on things when that voltage is too low to be dangerous.
Can you give me a picture of that? I'm having difficulty visualising it
in my head for some reason...all the wires confuse me a bit...but when I
know what's what I'm good.
If that 36V supply is fine, then check the 4
apporpriate transsitors on
the bias/interface board. Most likely soem are damaged.
-tony
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
3:08 p.m.
On 6/16/2014 5:00 PM, Cory Smelosky wrote:
Can you give me a picture of that? I'm having
difficulty visualising it
in my head for some reason...all the wires confuse me a bit...but when I
know what's what I'm good.
Cory, Have you seen this?
<http://www.retrocmp.com/tools/pdp11gui/memory/37-decnet/61-reparing-a-h7140-power-supply>
John H. Reinhardt
7:06 p.m.
On Mon, 16 Jun 2014, John H. Reinhardt wrote:
On 6/16/2014 5:00 PM, Cory Smelosky wrote:
The problem appears to actually be in the circuitry and nothing as simple
as that, unfortunately. :(
I have managed to track down 2 transistors (seemingly NOT a part of the
H-bridge) that don't test as I would expect them to, however. That's a
starting point.
Can you give me a picture of that? I'm
having difficulty visualising it
in my head for some reason...all the wires confuse me a bit...but when I
know what's what I'm good.
Cory, Have you seen this?
<http://www.retrocmp.com/tools/pdp11gui/memory/37-decnet/61-reparing-a-h7140-power-supply>
John H. Reinhardt
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
17 Jun
17 Jun
3:04 p.m.
The H7140 is a bit strange. No, let me put that anotehr way. It's
totally uncovnentional.
That I know for sure. ;)
The first thing is to have a clear idea od the
sections of the PSU :
Yeah, that helps.
Ahh. Logic starts on the module where I don't know if it's good or not.
It is meaningless IMHO to regard the bias/interface board as one module.
It contails 3 parts of the PSU circuit, which are mostly independent of
each other :
The control cirucit and chopper for the PSU for the control circuits fo
the main PSUs . Be warend that this PSU is fairly conventional and as a
result much of the circuitry is on the live side of the isolation barrier
-- directly conencted ot the mains. Take care
The logic that eneables/disables the 2 main PSUs depending on the console
switch, etc
The fan driver circuit.
Now for the fans. As is well-known these are
special, 35V 70Hz ones. They
are run from the 36V DC rail i nthe meoroy supply via a full-H driver
circuit (4 power tranisotrs on the bias/interface board) and associaled
bits. The idea of that is that the fans are also abttery-backed.
Ahhhh! Also on the bias/interface board! That module might very well be
bad...which would explain a lot of the problems. Now, there is no protection on the fan driver
circuit. So any short
between the fn leads of fro ma fan lead to chaiss will do damage. In
particular it burns out the transisotrs i nthe full-H circuit.
Ahh. I imagine they're not uncommon power transistors? Can you give me a picture of that? I'm having
difficulty visualising it
in my head for some reason...all the wires confuse me a bit...but when I
know what's what I'm good.
On the top of some of the boards are little spade terminals soldered to
the PCB. These are fitted to plastic screw-terminal blocks, the design of
these blocks iwth braised barrier ridges between the screws leads them to
be ca;lled 'barrier strips' in the UK. I am not sure what they are called
elsewhere.
In genneral, in this PSU, these are used for mains-side voltages. Briom
what I remembrer at least the 'logic inverter' (==chopper) PCB has a
multi-way one goign to the chopper transformer for that supply. And both
the 'inverter' boards have 2 way blocks with wires going to the big
capcitors at the left side. These carry the 350V DC from the maisn
rectifier/smoothign capacitors to the choppers. Needless to say that is a
lethal voltage/current (and the connections ot the hcopper transofrmer
migth well be deadly too).
Before workign on the supply, unplug it fro mteh mains. Wait a few
minuts. Then take the cover off (if it's not already off) and measure the
votlage between the connecitosn on one of thse 2-way barrier strips. It
doens't matter which one, they are the same. DOn't sart disconnecitng
things or pulling PCBs until that voltage is too low to the dangerous.
-tony
3841
days inactive
3843
days old
5 comments
3 participants
participants (3)
-
ard＠p850ug1.demon.co.uk -
b4＠gewt.net -
johnhreinhardt＠yahoo.com