Excerpts from the FAB Notebook:
(Since there is more than one thread pondering the issue)
Running Mosfets without degenerating resistors or complicated biasing
schemes is not all that difficult. As I have said elsewhere, it takes
adequate hardware and a little patience.
Not all that long ago I did a thumbnail analysis on power Mosfet current
sharing without degeneration. I matched up a group of IRFP Mosfets with
varying degrees of Vgs match at 25 deg C, 1A and put them on the same
sink and ran them with a fixed Vgs for an initial 1 amp current until the
sink reached 50 deg C.
At 50 deg, the current through each had drifted up about 30%, or 1.3 A.
The parts matched to 2% (about 10 mV Vgs) at 25 deg showed about
a 5% match at 50 deg.
From this you can conclude that reasonable matching of Vgs at room
temperature is adequate to hold a match at normal operating temperatures,
and the drift with temperature, while significant, does not necessarily lead
to thermal runaway.
In this case you could adjust the initial bias at 75% and let it warm up.
Don't try this with Bipolar devices. There is more than one manufacturer
who has learned that the hard way...
(Since there is more than one thread pondering the issue)
Running Mosfets without degenerating resistors or complicated biasing
schemes is not all that difficult. As I have said elsewhere, it takes
adequate hardware and a little patience.
Not all that long ago I did a thumbnail analysis on power Mosfet current
sharing without degeneration. I matched up a group of IRFP Mosfets with
varying degrees of Vgs match at 25 deg C, 1A and put them on the same
sink and ran them with a fixed Vgs for an initial 1 amp current until the
sink reached 50 deg C.
At 50 deg, the current through each had drifted up about 30%, or 1.3 A.
The parts matched to 2% (about 10 mV Vgs) at 25 deg showed about
a 5% match at 50 deg.
From this you can conclude that reasonable matching of Vgs at room
temperature is adequate to hold a match at normal operating temperatures,
and the drift with temperature, while significant, does not necessarily lead
to thermal runaway.
In this case you could adjust the initial bias at 75% and let it warm up.
Don't try this with Bipolar devices. There is more than one manufacturer
who has learned that the hard way...
conclusion for all interested ..... you can do that even with pucks
you only need a drekload to choose from
(here goes small ? )
you only need a drekload to choose from
(here goes small ? )
I've already done this numerous times.
Pretty lonely in the abfab club, or maybe people just want to over think this stuff. No idea.
When ya dumb like me the solutions are easy.
Hahaha
Pretty lonely in the abfab club, or maybe people just want to over think this stuff. No idea.
When ya dumb like me the solutions are easy.
Hahaha
- Status
- Not open for further replies.