Hi,


I have here a CyberPowerSystems 1500VA (950W) UPS which has failed in a most peculiar way. Actually, it's not the failure that seems peculiar to me, but that it was ever designed this way to begin with. I've chosen not to move forward with any replacement parts until I can nail down exactly why they are there and what kind of intoxicants the original engineer was on at the time of design.

First off, it doesn't know what model it is. The front says 1500AVR, the sticker in the microcontroller says OP1500 (which is another CPS model), the board says OP2200A (yet another). Ignoring that...

There are two large transformers, completely parallel with each other, each with three windings, two with taps. Of interest is the non-tapped winding, which gives me 24V, and connects back to the main board. It hits first a 6..3A fuse, then a full-wave rectifier (made of discrete diodes), expected filter cap, and then... a mosfet. a 60V 80A n-channel mosfet. Source and drain. The gate goes back to a collection of surface-mount parts which my eyeballs can't seem to follow.

I can't see any reason for this mosfet OTHER than to blow that fuse (which is exactly what it did. the mosfet is failed short).


Anybody have any familiarity with these? I've read that they're not very reliable, they must fail fairly often... so I can't be the only person with a broken one...

Am I correct in understanding that your question is... Did the MOSFET fail of
its own accord, or is there something else kaput that's likely to take out the
replacement?

In article ,
wrote:
I have here a CyberPowerSystems 1500VA (950W) UPS

Not the same model, but I have a PowerCOM PCM "Black Knight" 500VA that I
roughly traced the circuit of.

.3A fuse, then a full-wave rectifier (made of discrete diodes), expected fi=
lter cap, and then... a mosfet. a 60V 80A n-channel mosfet. Source and drai=
n. The gate goes back to a collection of surface-mount parts which my eyeba=
lls can't seem to follow.

Similar, but the "fuse" seems to be a 0.1R resistor, bridge, then an FET
right across the supply feeding onward to the battery charger section.

I can't see any reason for this mosfet OTHER than to blow that fuse (which =
is exactly what it did. the mosfet is failed short).

Your posting reminds me of the same puzzlement I had. On mine, a pin on the
MicroController operates an intermediate transistor that feeds the FET,
and seems to be an emergency shutdown to forcibly stop the charger dead.

Some kind of controlled crowbar. Or a "warranty expired, time to replace me"
timer

--
--------------------------------------+------------------------------------
Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk

--- news://freenews.netfront.net/ - complaints: ---

Den 09-05-2013 22:12, Mike skrev:
In article ,
wrote:
I have here a CyberPowerSystems 1500VA (950W) UPS

Not the same model, but I have a PowerCOM PCM "Black Knight" 500VA that I
roughly traced the circuit of.

.3A fuse, then a full-wave rectifier (made of discrete diodes), expected fi=
lter cap, and then... a mosfet. a 60V 80A n-channel mosfet. Source and drai=
n. The gate goes back to a collection of surface-mount parts which my eyeba=
lls can't seem to follow.

Similar, but the "fuse" seems to be a 0.1R resistor, bridge, then an FET
right across the supply feeding onward to the battery charger section.

I can't see any reason for this mosfet OTHER than to blow that fuse (which =
is exactly what it did. the mosfet is failed short).

Your posting reminds me of the same puzzlement I had. On mine, a pin on the
MicroController operates an intermediate transistor that feeds the FET,
and seems to be an emergency shutdown to forcibly stop the charger dead.

Some kind of controlled crowbar. Or a "warranty expired, time to replace me"
timer

Crowbar is the correct term. Here is an explanation of the function:

http://en.wikibooks.org/wiki/Practic...rowbar_circuit



--
Uffe

Den 09-05-2013 20:55, skrev:
Hi,


I have here a CyberPowerSystems 1500VA (950W) UPS which has failed in a most peculiar way. Actually, it's not the failure that seems peculiar to me, but that it was ever designed this way to begin with. I've chosen not to move forward with any replacement parts until I can nail down exactly why they are there and what kind of intoxicants the original engineer was on at the time of design.

First off, it doesn't know what model it is. The front says 1500AVR, the sticker in the microcontroller says OP1500 (which is another CPS model), the board says OP2200A (yet another). Ignoring that...

There are two large transformers, completely parallel with each other, each with three windings, two with taps. Of interest is the non-tapped winding, which gives me 24V, and connects back to the main board. It hits first a 6.3A fuse, then a full-wave rectifier (made of discrete diodes), expected filter cap, and then... a mosfet. a 60V 80A n-channel mosfet. Source and drain. The gate goes back to a collection of surface-mount parts which my eyeballs can't seem to follow.

I can't see any reason for this mosfet OTHER than to blow that fuse (which is exactly what it did. the mosfet is failed short).


Anybody have any familiarity with these? I've read that they're not very reliable, they must fail fairly often... so I can't be the only person with a broken one...

Crowbar circuit. Found on Wiki:

http://en.wikibooks.org/wiki/Practic...rowbar_circuit



--
Uffe

In article ,
=?ISO-8859-1?Q?Uffe_B=E6rentsen?= wrote:

MicroController operates an intermediate transistor that feeds the FET,
and seems to be an emergency shutdown to forcibly stop the charger dead.

Some kind of controlled crowbar. Or a "warranty expired, time to replace me"
timer

Crowbar is the correct term. Here is an explanation of the function:

http://en.wikibooks.org/wiki/Practic...rowbar_circuit

It just seems unusual for the crowbar to be operated by a
microcontroller I/O pin. Normally, it would be a zener or voltage
reference, so that if supply voltages were exceeded it forcibly
held things down until something went pop. A standalone safety
circuit.

And this is stopping off the *input* to a current limited, voltage
regulated lead-acid charger designed for 12v 7-9Ah batteries.
--
--------------------------------------+------------------------------------
Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk

--- news://freenews.netfront.net/ - complaints: ---