Pansat 9200 HD Heatsink Modifications
03/05/2010
Extended 03/15/2010
Extended 01/30/2012
Extended 10/08/2016
SKEW/POLARITY BOARD WEBSITE
Extended 10/29/2016
The
MPEG II and the MPEG 4 chips on the Pansat 9200 HD STB run hot and the
provided heat sinks are attached with a gob of glue that is very
thick. This glue acts
as an insulator such that the chips are considerably hotter than the
sinks. I removed
the heat sinks and the glue on them and replaced it with Silicone heat
sink compound.
This compound is not a glue and the heat sinks must be retained by
another means.
For the larger MPEG II sink a beam structure was created using existing
motherboard
holes and a phospher bronze sheet to constantly apply pressure.
The alignment was
maintained by the aluminum beam as seen below in the pics.
For the MPEG 4 chip, which was much smaller, two small dots of
superglue were
placed at two diagonal corners of the chip for retension.
The first six pictures below are of the MPEG II chip and its
heatsink. The next three
are of the MPEG 4 chip and its heatsink.
The last 2 pictures show the arrangement used to retain the MPEG II
heatsink.
To date this along with a small fan on the top of the case has reduced
freezing of the
Pansat 9200 HD by about 20 times. Initial trials with the fan
alone were virtually
ineffective. Fan speed is set so that it is not audible with a
small variac.
To separate the heat sink from the chip I used an Exacto knife (thin
flat end blade)
and a small hammer with some very sharp light taps. Yeah, it can
be nervous time in
doing such. I believe I cleaned the glue off with lacquer thinner.
This change certainly kills the warranty but it is worth it to make it
usable. It has been
in operation for a year or two this way. It certainly is still a
quirky piece of hardware
in other ways but at least under normal usage it is stable. I
believe any DVB-S2 upgrade
board will still fit. (Wrong - a
better design follows for the Pansat DVB-S2 Plus to fit).
The tools to part the heatsinks are in the last picture. A single
hit with the hammer
should part the heat sink. Confirm location of the blade each hit.
Clicking on
the below pics provides a 4 times enlargement of such:
The
following pictures are of a second Pansat 9200 HD which I obtained on
Ebay.
The Exacto blade (only) was used with the tack hammer to crack off the
heat sinks.
The scorch on this MPEG-2 Decoder glue is clearly evident.
The last picture shows an improved design of the clamping hardware
where both
the MPEG-2 and MPEG-4 decoder heat sinks are clamped. Two 1.5"
#4-40 bolts
and one 2" #6-32 bolt were used to secure the three legged clamp.
Brass tubing was
resistance soldered to the ends of a 3/4"x7.94"x.092" bar and to one
end of a 6.25"
length bar soldered to the longer one.
Existing board holes were again used and the Pansat DVB-S2 Plus Kit
does fit ok.
A third approach to attaching the heatsink is by using a tiny spot of
superglue on each
corner between the chip and the heatsink. Of course, the
insulative thick glob of glue
used in the original manufacturing process is replaced by a very thin layer of Silicone
heat sink compound, essentially only present where the heatsink does
not acturally
contact the chip. This provides the best possible conductivity
between the two. This
not only provides good cooling but is clean and simple.
A tiny fan can then be mounted on the heatsink. Using a 12v fan
with a 7 to 9volt DC
supply eliminates any audio noise. This is done by using six to
nine power diodes in
series from the positive source to the fan. As in the below
picture these are covered
with some 1/8" shrink tubing. If the 9200HD does not have the S2
card installed then
half the socket it normally uses on the power supply board can be used
as the 12v
source.
One More Time on the Pansat
9200HD Heating Problem
One more iteration was applied to end the heating problem of the Pansat
9200HD, not to mention the failure of four (4) can capacitors in its
power supply. The latter capacitors can be easily identified by
their
rounded tops and in my case are identified in the following picture and
sketch. Three of these are 2200uf 16v can caps and the third a
3300uf
6.3v can cap. I salvaged some from failed pc motherboards.
The problem with earlier attempts to limit the heating of the processor
was that the area on the chip which needed to be in contact with the
heatsink was from .001" to .003" below the surrounding epoxy encasement
and ended up having a layer of heat sink compound filling that
void.
Metal to metal contact is the best possible means of transmitting the
heat away from the processor.
To remedy this problem the bottom of the stock heat sink was milled to
a depth of .003" so the center portion would contact the metal part of
the processor as can be viewed in the following picture:
Next it was solidly clamped to the pc board processor (with a very thin layer of heat
sink compound) as in this next picture with a small fan installed:
To power the fan a header was installed on the power supply board and
the regulated 12vdc was supplied through 6 power diodes (dropping the
voltage to about 9v) to insure its noise level was not audible.
The
result was that the processor heat sink was observed to run at about
20-22° above room temperature. In contrast, the transformer
on the
power supply board was about 85° above room temperature and the can
capacitor closest to the transformer (which failed) ran at about
55°
above room temperature.
While I recognize that few FTA fans have access to a milling machine or
a resistance soldering unit this serves as a reference as to one effort
necessary to remedy this manufactured shortcoming. The
Pansat 9200HD
otherwise works very well and it does not appear that the many
overheating occurances, which this unit was exposed to, has permanently
damaged the processor. Evidence of overheating shows up simply in
the
processor's slowing down to a crawl in exercising commands.
It should be noted that an alternative to the brass clamping of the
heat sink is to apply a spot of super glue to the corners of the
processor chip. I did this on another unit as noted above but I
do not
know how permanent this solution is for the main processor though it
has been effective for the MPEG-4 chip on this unit.
----------------------------------------------------------------------------------
PANSAT SKEW/POLARITY CONTROL BOARD SITE
