20 Jul
2020
20 Jul
'20
8 a.m.
I have a MicroVAX 3100 which has a H7822 power supply. The power supply
and the machine itself mostly work (there is a problem with the SCSI
interfaces but that's another story) except that the two fans in the
power supply don't run. If left on for a long time, the machine gets
too hot and a thermal trip operates, shutting it down.
The fans are DC 12V 0.2A and if I connect them to +5V or +12V, they
work fine and don't draw excessive current so there would seem to be
a problem with the section of the power supply which drives the fans.
Unfortunately, it's operation is not obvious and the power supply is
a pig to work on. It consists of two double sided PCBs connected by
short leads and having live parts on both boards making it difficult
to get access to both sides of the board where the fan circuit is when
the power is on.
I don't have an identical working power supply to compare the faulty
one to but the fan circuit looks superficially similar to the one in
the H7821 which I do have working examples of so that may be a way
to proceed.
Does anyone have a service manual for the H7822 or H7821 or know how
the fan circuit is supposed to work in these power supplies?
Regards,
Peter Coghlan.
20 Jul
20 Jul
11:03 a.m.
On 07/20/2020 07:00 AM, Peter Coghlan via cctalk wrote:
I have a MicroVAX 3100 which has a H7822 power supply.
The power supply
and the machine itself mostly work (there is a problem with the SCSI
interfaces but that's another story) except that the two fans in the
power supply don't run. If left on for a long time, the machine gets
too hot and a thermal trip operates, shutting it down.
The fans are DC 12V 0.2A and if I connect them to +5V or +12V, they
work fine and don't draw excessive current so there would seem to be
a problem with the section of the power supply which drives the fans.
Unfortunately, it's operation is not obvious and the power supply is
a pig to work on.
Yes, it probably has a temperature sensor and a fan speed
controller. If you don't care about noise, you
could probably just rewire the fans to 12 V directly.
Jon
11:55 a.m.
Jon Elson wrote:
On 07/20/2020 07:00 AM, Peter Coghlan via cctalk
wrote:
I was thinking about wiring them to 5V because it was only marginally
overheating after running for a long time. However, I started
poking around with the multimeter and discovered a low resistance
across one of the connectors for the fans. This led me to a
1N759A 12V 400mW zener diode which read about 20 Ohms in both
directions. Looking at the H7821, there was a 1N4742 12V 1W zener
diode in a similar position. It had much more plausable readings
so I borrowed it and fitted it to the H7822 in place of the dud
1N759A. The fans are spinning nicely now with about 7.5 to 8V
across each one. This was a lot easier than I was expecting :-)
I have a MicroVAX 3100 which has a H7822 power
supply. The power supply
and the machine itself mostly work (there is a problem with the SCSI
interfaces but that's another story) except that the two fans in the
power supply don't run. If left on for a long time, the machine gets
too hot and a thermal trip operates, shutting it down.
The fans are DC 12V 0.2A and if I connect them to +5V or +12V, they
work fine and don't draw excessive current so there would seem to be
a problem with the section of the power supply which drives the fans.
Unfortunately, it's operation is not obvious and the power supply is
a pig to work on.
Yes, it probably has a temperature sensor and a fan speed
controller. If you don't care about noise, you
could probably just rewire the fans to 12 V directly.
Jon
12:45 p.m.
On 07/20/2020 10:55 AM, Peter Coghlan via cctalk wrote:
I was thinking about wiring them to 5V because it was only
marginally
overheating after running for a long time. However, I
started
poking around with the multimeter and discovered a low
resistance
across one of the connectors for the fans. This led me to a
1N759A 12V 400mW zener diode which read about 20 Ohms in both
directions. Looking at the H7821, there was a 1N4742 12V
1W zener
diode in a similar position. It had much more plausable
readings
so I borrowed it and fitted it to the H7822 in place of
the dud
1N759A. The fans are spinning nicely now with about 7.5
to 8V
across each one. This was a lot easier than I was
expecting :-)
Wow, lucky it didn't smoke anything. there must be a series
resistor somewhere that probably got pretty hot.
Jon
2:18 p.m.
On 07/20/2020 10:55 AM, Peter Coghlan via cctalk
wrote:
It looks like the zener is connected between the adjust terminal
of a heatsink mounted LM337T adjustable negative voltage regulator
and the positive connection to one of the fans. Also connected
to the adjust terminal is what looks like a thermistor mounted on
the heatsink of a MBR3045, presumably a switching transistor.
Perhaps the zener didn't end up carrying much current, however if
that is the case, it is a bit strange that it failed.
Regards,
Peter Coghlan.
I was thinking about wiring them to 5V because it was only
marginally
overheating after running for a long time. However, I
started
poking around with the multimeter and discovered a low
resistance
across one of the connectors for the fans. This led me to a
1N759A 12V 400mW zener diode which read about 20 Ohms in both
directions. Looking at the H7821, there was a 1N4742 12V
1W zener
diode in a similar position. It had much more plausable
readings
so I borrowed it and fitted it to the H7822 in place of
the dud
1N759A. The fans are spinning nicely now with about 7.5
to 8V
across each one. This was a lot easier than I was
expecting :-)
Wow, lucky it didn't smoke anything. there must be a series
resistor somewhere that probably got pretty hot.
Jon
1468
days inactive
1468
days old
4 comments
2 participants
participants (2)
-
cctalk＠beyondthepale.ie -
elson＠pico-systems.com