On 1/19/2011 2:25 PM, Meat Plow wrote:

Great dollops of that white goo between metal to metal thermally
conductive surfaces.

Overuse of that stuff is worse for thermal conductivity than none at all.
I've clean up gobs of it since they started using it decades ago.

This is something I'm currently wondering about. I have a fridge-like
thermo-electric cooler than has two sections, top and bottom, with
different temperatures. The top suddenly stopped getting cool at all, so
I took it apart to figure out why. The fans and voltages were all there
so I broke down the heat sinks on the bad one to get to the Peltier
device. With it isolated, I powered it up briefly and much to my
surprise the Peltier device got hot real quickly with the opposite side
getting cooler. So the device works, it has to be something with the
heat sinks?
They did use white goop on both sides, but very little and it was
already dried. The heat sinks are milled flat where they make contact
with the Peltier device, so my thinking is they need new goop.
Looking around I found that Star heat sink compound is about the best
you can get, so I ordered some. It just arrived the other day so I'm
planning to clean up the old goop, put on some new goop and hope for the
best. I don't think too much would be an issue in this case, I want it
as cold as possible.





--
-Scott

"Lab1" .@... wrote in message
...
On 1/19/2011 2:25 PM, Meat Plow wrote:

Great dollops of that white goo between metal to metal thermally
conductive surfaces.

Overuse of that stuff is worse for thermal conductivity than none at all.
I've clean up gobs of it since they started using it decades ago.

This is something I'm currently wondering about. I have a fridge-like
thermo-electric cooler than has two sections, top and bottom, with
different temperatures. The top suddenly stopped getting cool at all, so I
took it apart to figure out why. The fans and voltages were all there so I
broke down the heat sinks on the bad one to get to the Peltier device.
With it isolated, I powered it up briefly and much to my surprise the
Peltier device got hot real quickly with the opposite side getting cooler.
So the device works, it has to be something with the heat sinks?
They did use white goop on both sides, but very little and it was already
dried. The heat sinks are milled flat where they make contact with the
Peltier device, so my thinking is they need new goop.
Looking around I found that Star heat sink compound is about the best you
can get, so I ordered some. It just arrived the other day so I'm planning
to clean up the old goop, put on some new goop and hope for the best. I
don't think too much would be an issue in this case, I want it as cold as
possible.





--
-Scott



Too much of the stuff will be a problem whether you are trying to heat or
cool. I have some major doubts that a thin coating of that stuff would be
enough to make any major difference in the performance of peteler junction.
Now on the other hand, if the heatsink is loose... That could give you some
real issues.

In article ,
Michael Kennedy mike@com wrote:

They did use white goop on both sides, but very little and it was already
dried. The heat sinks are milled flat where they make contact with the
Peltier device, so my thinking is they need new goop.

Too much of the stuff will be a problem whether you are trying to heat or
cool. I have some major doubts that a thin coating of that stuff would be
enough to make any major difference in the performance of peteler junction.
Now on the other hand, if the heatsink is loose... That could give you some
real issues.

Michael is quite correct.

The thing about heatsink compound, is that you should only use a
*very* thin layer, and use it between surfaces which are already flat
and well-fitting. Adding a thicker layer of heatsink compound than is
necessary, will actually reduce thermal conductivity.

You want as much direct metal-to-metal or metal-to-ceramic contact as
you can get - enthusiasts who "overclock" their PCs will often flatten
and polish the top of the CPU and the bottom of the heatsink to
increase direct contact. A *thin* smear of heatsink compound is
appropriate... just enough to fill the remaining gaps between the
heatsink and the heat-sunk :-). You almost want to smear it on, and
then wipe most of it off gently with a single-edged razor blade, so
that there is no excess buildup between the two surfaces.

And, yes, if the heatsink actually comes loose from the Peltier
junction (e.g. if it was originally spring-clipped in place, and the
clips are loose or have fatigued and lost pressure) then you've got
problems... you'll get a layer of air between the two surfaces, and
thermal conductivity will become quite poor. Adding a thicker layer
of goop to try to fill the gap isn't the right thing to do - instead,
fix whatever caused the devices to become loose, clean the surfaces,
reapply a *thin* layer of compound, and secure the devices back
together with the proper amount of pressure.

If there was (apparently) nothing holding the two surfaces together -
no clips or retainers - then you're probably dealing with a "thermally
conductive adhesive". Some of these are good, some are poor... and
you'll have to strip off all of the remains, and then reapply (again)
a very thin layer of a suitable thermal adhesive, and fasten the parts
back together with appropriate pressure until the adhesive cures.

--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

On 1/20/2011 2:03 AM, Dave Platt wrote:

And, yes, if the heatsink actually comes loose from the Peltier
junction (e.g. if it was originally spring-clipped in place, and the
clips are loose or have fatigued and lost pressure) then you've got
problems... you'll get a layer of air between the two surfaces, and
thermal conductivity will become quite poor. Adding a thicker layer
of goop to try to fill the gap isn't the right thing to do - instead,
fix whatever caused the devices to become loose, clean the surfaces,
reapply a *thin* layer of compound, and secure the devices back
together with the proper amount of pressure.

I agree. The assembly goes like this:
Small heat sink (cold side) - square plastic gasket with embedded rubber
seal that doesn't physically touch anything - square block of Styrofoam
with a square cutout in the middle - Peltier device - foam tape around
the styrofoam - large heat sink. To screws go through everything on
either side of the Peltier to sandwich it all together.
And oddly they hot-glued the ends of the screws and nuts.

This is obviously made in China, everything is pretty crudely
manufactured and assembled, heat sink fins were mashed together in
spots. I didn't think to check the tightness of those two screws when I
took it apart, but I bet you are right and they weren't nearly tight
enough. I'm going to rebuild the 2nd one while I'm at it and will check
the tightness after I pry off all the hot glue...


--
-Scott

On Jan 20, 2:38*pm, Lab1 .@... wrote:

This is obviously made in China, everything is pretty crudely
manufactured and assembled,

Most consumer electronics stuff these days is made in China! From low
end kit to decent, so you can't generalise that something made in
China is automatically shoddy.
Although a fair proportion is indeed cheap and disposable, that is the
case because they are responding to a demand. They only supply what
the west is prepared to pay for anyway!

Pet hates: excessive amounts of screws holding covers of TVs etc.
together. this seems to have got worse with flat panels.Often you
spend as much /more time assembling and re-asembling than the repair!

-B

b wrote:

On Jan 20, 2:38 pm, Lab1 .@... wrote:

This is obviously made in China, everything is pretty crudely
manufactured and assembled,

Most consumer electronics stuff these days is made in China! From low
end kit to decent, so you can't generalise that something made in
China is automatically shoddy.
Although a fair proportion is indeed cheap and disposable, that is the
case because they are responding to a demand. They only supply what
the west is prepared to pay for anyway!

Pet hates: excessive amounts of screws holding covers of TVs etc.
together. this seems to have got worse with flat panels.Often you
spend as much /more time assembling and re-asembling than the repair!


Some large color TV consoles built in the '60s had 20 to 30 screws.


--
You can't fix stupid. You can't even put a band-aid on it, because it's
Teflon coated.

On Fri, 21 Jan 2011 04:37:26 -0800 (PST), b
wrote:

Pet hates: excessive amounts of screws holding covers of TVs etc.
together. this seems to have got worse with flat panels.Often you
spend as much /more time assembling and re-asembling than the repair!
-B

Screws are fine. I can live with having too many screws because half
of them are usually stripped out and the remainder hold things
together. What bugs me are snap together clamshell cases, such as LCD
monitors and TV's, where you have to remove a mess of screws *AND* pry
the case apart. While snap together plastic is probably easier to
assemble because it doesn't require much fixturing to position the
robotic screwdriver, it does suggest that the case was never intended
to be opened or the unit repaired.

I recently repaired a Yamaha P70 electric piano. I didn't count, but
I'll guess about 60 large "sheet metal" type screws holding it
together. I don't use an electric screwdriver, but this is one time
that I wish I owned one. Even with switching hands, my hands ached
after I as done. The problem with such "sheet metal" screws is that
they offer high frictional resistance on every turn, while more
conventional screw threads, only offer high resistance when tight (or
smeared with thread lock).

My guess is the large number of screws was to prevent mechanical
resonances in the plastic case or to keep them from falling out from
vibrations. Still, metal thread inserts, screw threads, and steel nut
plates, would have been much easier to handle.

On the silicon grease front, I agree with most of the comments. Thin
works much better than globs of silicon grease. One should remember
that the purpose of silicon grease is NOT to bridge gaps. It's to
fill in the surface roughness, groves, and crevasses. Most of the
heat transfer is metal to metal contact, not through the silicon
grease.

In a past life, I used to design marine radios. The problem was that
the power xsistor packages of the day (1970's) were generally thermal
disasters. Either there was insufficient contact area to obtain
sufficiently low thermal resistance, or they were not flat. I solved
the first by building pyramid like structures of copper washers to act
as a heat spreader. I solved the latter by polishing the mounting
base of the power transistors on fine emery cloth. I hated to polish
away the gold plating, but that's what it took to get the heat out. I
made numerous tests and measurements trying to determine the optimum
amount of silicon grease, and eventually concluded that ultra thin is
best. Instructions were to smear a tiny amount onto the area, and
then wipe ALL of it off with a plastic scraper. What remained was
silicon grease in the remaining surface roughness, which was all that
was necessary.

I recently repaired an IFR-1500 service monitor. The power supply
section was intermittent. The 0.062 aluminum power supply case, was
butted up against the large aluminum heat sink that covered the entire
rear panel. In between was a huge amount of silicon grease. The
sandwich was held together by two large 10-24 screws, which probably
explains the silicon grease overdose. Two screws is not going to bend
the aluminum case so that it lays flat. So they tried to fill in the
lack of flatness with silicon grease. That doesn't work.

It took me considerable effort and alcohol to clean up the mess, but I
still managed to get it all over everything on the bench. After the
repair (large copper wires on torroids were not soldered properly), I
reassembled it with only a little silicon grease around the two large
screws, and left the rest to it's own devices. Works fine with no
obvious overheating (checked with an IR thermometer and thermocouple
probe). My guess is all that silicon grease did nothing useful.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Jeff Liebermann wrote in message
...
On Fri, 21 Jan 2011 04:37:26 -0800 (PST), b
wrote:

Pet hates: excessive amounts of screws holding covers of TVs etc.
together. this seems to have got worse with flat panels.Often you
spend as much /more time assembling and re-asembling than the repair!
-B

Screws are fine. I can live with having too many screws because half
of them are usually stripped out and the remainder hold things
together. What bugs me are snap together clamshell cases, such as LCD
monitors and TV's, where you have to remove a mess of screws *AND* pry
the case apart. While snap together plastic is probably easier to
assemble because it doesn't require much fixturing to position the
robotic screwdriver, it does suggest that the case was never intended
to be opened or the unit repaired.

I recently repaired a Yamaha P70 electric piano. I didn't count, but
I'll guess about 60 large "sheet metal" type screws holding it
together. I don't use an electric screwdriver, but this is one time
that I wish I owned one. Even with switching hands, my hands ached
after I as done. The problem with such "sheet metal" screws is that
they offer high frictional resistance on every turn, while more
conventional screw threads, only offer high resistance when tight (or
smeared with thread lock).

My guess is the large number of screws was to prevent mechanical
resonances in the plastic case or to keep them from falling out from
vibrations. Still, metal thread inserts, screw threads, and steel nut
plates, would have been much easier to handle.

On the silicon grease front, I agree with most of the comments. Thin
works much better than globs of silicon grease. One should remember
that the purpose of silicon grease is NOT to bridge gaps. It's to
fill in the surface roughness, groves, and crevasses. Most of the
heat transfer is metal to metal contact, not through the silicon
grease.

In a past life, I used to design marine radios. The problem was that
the power xsistor packages of the day (1970's) were generally thermal
disasters. Either there was insufficient contact area to obtain
sufficiently low thermal resistance, or they were not flat. I solved
the first by building pyramid like structures of copper washers to act
as a heat spreader. I solved the latter by polishing the mounting
base of the power transistors on fine emery cloth. I hated to polish
away the gold plating, but that's what it took to get the heat out. I
made numerous tests and measurements trying to determine the optimum
amount of silicon grease, and eventually concluded that ultra thin is
best. Instructions were to smear a tiny amount onto the area, and
then wipe ALL of it off with a plastic scraper. What remained was
silicon grease in the remaining surface roughness, which was all that
was necessary.

I recently repaired an IFR-1500 service monitor. The power supply
section was intermittent. The 0.062 aluminum power supply case, was
butted up against the large aluminum heat sink that covered the entire
rear panel. In between was a huge amount of silicon grease. The
sandwich was held together by two large 10-24 screws, which probably
explains the silicon grease overdose. Two screws is not going to bend
the aluminum case so that it lays flat. So they tried to fill in the
lack of flatness with silicon grease. That doesn't work.

It took me considerable effort and alcohol to clean up the mess, but I
still managed to get it all over everything on the bench. After the
repair (large copper wires on torroids were not soldered properly), I
reassembled it with only a little silicon grease around the two large
screws, and left the rest to it's own devices. Works fine with no
obvious overheating (checked with an IR thermometer and thermocouple
probe). My guess is all that silicon grease did nothing useful.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com

I suspect we are as engineers are just as likely to be mislead by
manufacturer's claims as joe public and general ads. I could not convince
myself that silipads were better than mica (no patents so litterally dirt
cheap) - so experiment called for. These days I own a matchbox size remote
IR pyrometer and would use that to compare device body temps in before and
after situations , not fingertips

Mica versus Silicone pad insulators
I was not convinced that for an existing used amp with 4x TOP66 power output
devices that the silipads were better than mica.
Each of the 4 white insulating pads had shrunk about 5mm at the tops
(hottest)
compared to bottoms , ruffling the original outer edges, heat damage ?.
I'm wondering if they can chemically change over time and/or excessive
temperature , downgrading to be more of a thermal insulator.
They are not discoloured or hardened or anything different in the
ex-compressed area by sight or flexing, just permanently deformed , the
ruffling is permanent.
I replaced all 4 with mica and thin films of thermal grease.
Before doing so I powered up the amp with 400 Hz continuous sine giving 20
watts in a dummy load. No fan cooling for this amp, just
convection/radiation.
Laid a brass barrel protected thermometer on the heatsink and took
measurements. Stabilised at 33 deg C over ambient after 50 minutes.
Replaced with mica and redid the load test.
For same ambient , same testing position/attitude, power in load etc it now
took 30 minutes to stabilise at plus 32 deg C over ambient.
More graphically , but less scientific, - the finger test.
After half an hour of heating with the mica setup I could hold a fingertip
on each tranny for about 5 seconds before finding it uncomfortable.
Previously half a second of fingertip touch was enough.
I think I will rely on the evidence of my own observations and not
performance tables produced by the manufacturer's with an obvious vested
interest.
I've no reason to believe the original silipads had aged, been affected by
WD40 or anything.
I will assume they are , all manufacturers, all generically bad until a
similar personally conducted experiment, in a real situation, proves to me
to be otherwise.

Jeff Liebermann wrote in message
...
On Fri, 21 Jan 2011 04:37:26 -0800 (PST), b
wrote:

Pet hates: excessive amounts of screws holding covers of TVs etc.
together. this seems to have got worse with flat panels.Often you
spend as much /more time assembling and re-asembling than the repair!
-B

Screws are fine. I can live with having too many screws because half
of them are usually stripped out and the remainder hold things
together. What bugs me are snap together clamshell cases, such as LCD
monitors and TV's, where you have to remove a mess of screws *AND* pry
the case apart. While snap together plastic is probably easier to
assemble because it doesn't require much fixturing to position the
robotic screwdriver, it does suggest that the case was never intended
to be opened or the unit repaired.

I recently repaired a Yamaha P70 electric piano. I didn't count, but
I'll guess about 60 large "sheet metal" type screws holding it
together. I don't use an electric screwdriver, but this is one time
that I wish I owned one. Even with switching hands, my hands ached
after I as done. The problem with such "sheet metal" screws is that
they offer high frictional resistance on every turn, while more
conventional screw threads, only offer high resistance when tight (or
smeared with thread lock).

My guess is the large number of screws was to prevent mechanical
resonances in the plastic case or to keep them from falling out from
vibrations. Still, metal thread inserts, screw threads, and steel nut
plates, would have been much easier to handle.

On the silicon grease front, I agree with most of the comments. Thin
works much better than globs of silicon grease. One should remember
that the purpose of silicon grease is NOT to bridge gaps. It's to
fill in the surface roughness, groves, and crevasses. Most of the
heat transfer is metal to metal contact, not through the silicon
grease.

In a past life, I used to design marine radios. The problem was that
the power xsistor packages of the day (1970's) were generally thermal
disasters. Either there was insufficient contact area to obtain
sufficiently low thermal resistance, or they were not flat. I solved
the first by building pyramid like structures of copper washers to act
as a heat spreader. I solved the latter by polishing the mounting
base of the power transistors on fine emery cloth. I hated to polish
away the gold plating, but that's what it took to get the heat out. I
made numerous tests and measurements trying to determine the optimum
amount of silicon grease, and eventually concluded that ultra thin is
best. Instructions were to smear a tiny amount onto the area, and
then wipe ALL of it off with a plastic scraper. What remained was
silicon grease in the remaining surface roughness, which was all that
was necessary.

I recently repaired an IFR-1500 service monitor. The power supply
section was intermittent. The 0.062 aluminum power supply case, was
butted up against the large aluminum heat sink that covered the entire
rear panel. In between was a huge amount of silicon grease. The
sandwich was held together by two large 10-24 screws, which probably
explains the silicon grease overdose. Two screws is not going to bend
the aluminum case so that it lays flat. So they tried to fill in the
lack of flatness with silicon grease. That doesn't work.

It took me considerable effort and alcohol to clean up the mess, but I
still managed to get it all over everything on the bench. After the
repair (large copper wires on torroids were not soldered properly), I
reassembled it with only a little silicon grease around the two large
screws, and left the rest to it's own devices. Works fine with no
obvious overheating (checked with an IR thermometer and thermocouple
probe). My guess is all that silicon grease did nothing useful.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558


Another goo production failing , on a 1 to 2 yearold Fender Amp on my bench
at the moment. Uses intermediary Al block between immediate o/p h/s plate
and chassis. White goo on both surfaces is still as placed, not splurged
out. Failure to fettle/de-burr the post machining raised rims around the
machined holes so acting as thin washers so heat just going through the 3
bolts not body of Al. Amp failure nothing to do with this poor heatsinking