PCR: what to look for in a machine
Cheung C. Yue
ccy at po.CWRU.Edu
Tue Oct 20 08:56:51 EST 1992
In a previous article, fri0 at quads.uchicago.edu (Christian Fritze) says:
>We are about to buy our lab's first ever PCR machine. An industrious post-
>doc in the lab has researched many of the available offerings, and we are
>looking for some advice from experienced users about what makes the
>difference between a good and bad design.
I previously used the first generation Cetus machine. More recently I
decided to purchase the Corbett FTS-1S machine. The machine uses a
similar concept as the Idaho hot air machine, and that is to use a
capillary as the reaction vessel (in this case, a plastic positive
displacement pipette tip). The skinny tip and the small sample size
makes temperature changes to the sample *almost* instantaneous. The
tips are placed in a donut-shaped aluminum block wrapped with heating
elements on the outside, and armed with cooling fins on the inside
of the donut. Because of the geometry of the heating elements, uniformity
of temperature in different wells is almost guaranteed. The proximity
of the heating element to the sample well makes temperature changes very
fast. Cooling is just by blowing room air over the block and using
the cooling fins, and that too is fairly rapid (certainly faster than
my old Cetus with its compressor). Cycling is almost as fast as the
Idaho machine as described in BioTechniques, but one does not have to
handle glass microcapillary tube as with the Idaho machine (the last I
heard, Idaho is also coming out with a positive displacement pipet tip
model). Instructions with the Corbett machine is not the greatest.
Programming for the machine is a bit primitive, and there is no RS232
output, although there is a nice analog output for a chart recorder.
Filling capillary pipets with a manual pipettor may well create a new
clinical syndrome for the thumb (at least that's what I use to turn
the dial on the pipettor). While expensive and not tremendously versatile,
an electronic pipet designed for use with the Corbett machine is highly
recommended (about $1500). The machine costs around $7000. Other "bad"
thing about this machine is the size of the sample, which may be a plus
if you are interested only in testing samples, and not so much for cloning.
For cloning purposes, I have found I needed to run several tubes of each
sample to get enough material. Another short-coming is inability to do
Sorry for the rambling reply. I have no ties with the Corbett people. I
just happen to like their machine.
ccy at po.cwru.edu
More information about the Methods