take a look at your power supply design cookbooks -- you will notice current
balancing resistors on each of the transistors when they are paralleled.

if you purchase all of the transistors in one lot -- the MJ15003's come 100
to a package in a polyfoam container -- chances are they are going to line
up pretty closely. if you are buying them randomly you can use a DMM with a
transistor checker to measure beta. for current mirrors you measure beta and
VBE, for MOSFETS you measure the VGS such that IDS = a specified current
(100ma or 1 amp). International Rectifier has a section on their website on
MOSFET measurement, but I haven't seen anything similar for BJT's from
On-Semi or ST.

i spent a rainy afternoon "binning" transistors -- that is, I took a couple
hundred MOSFETs and ran the VGS and Transconductance test, marked each unit
(they were TO-247) and put the results in a spread sheet. From each "tube"
of IRFP9240's the distributions were tight and normal. Transconductance was
closely related (I also ran linear regression, r^2 was in the 90's) to
Vgs(t).

i do have MJ15003's and MJ15004's from the same package, email me privately
if you need them.

"bobinphx" wrote in message
news:05CYa.40443$Ne.25582@fed1read03...
To all,
I am working with a DC motor that goes forward and reverse. The motor is
controlled by a heat-sinked and fan cooled bank of 5 mj15003
transistors( 5
each forward and reverse). On occasion the motors build up a high amount
of
copper on the armatures, from the brushes and this causes one of the 5
transistors to heat up and fail. The manual and the manufacture suggest
matched transistors. What is a matched transistor????? I am guessing that
they are tested and all have similar characteristics. But what
characteristics am I looking for?? especially in a DC motor control.

Anyone have an answer for a semi technically literate person???? (i.e., I
know the hot end of the soldering iron, from the cold end and have the
scare
to prove it!!!)

Thanks in advance....

Bob in PHX