Does anyone know if this is *fully* isolated from the gubbins?
Measurements with an ordinary digital meter suggests that it is, but I'd like to be sure.
The unit I'm developing has a grounded diecast alloy case and I'd like to use this as a heatsink, ideally with a small strip of copper in direct contact with
both, and just a smear of heatsink compound.
On 02-07-2021 23:23, Folderol wrote:
Does anyone know if this is *fully* isolated from the gubbins?
Measurements with an ordinary digital meter suggests that it is, but I'd like
to be sure.
The unit I'm developing has a grounded diecast alloy case and I'd like to use
this as a heatsink, ideally with a small strip of copper in direct contact with
both, and just a smear of heatsink compound.
It's a heat spreader, not connected to the internals electrically (of >course?). What you plan to do is what all the alloy cases do that act as >heatsinks; you'll be safe. Hardest part will be to get the distance just >right: not too close or you'll bend the board and the connections, not
too far or it won't transfer heat.
Thanks for the confirmation. I was pretty sure that was the case.
My intention is to mount the Pi upside down with long bolts and springs pushing it down (it will be fully enclosed) so that should keep everything safe.
On Sat, 3 Jul 2021 11:15:05 +0200
"A. Dumas" <alexandre@dumas.fr.invalid> wrote:
On 02-07-2021 23:23, Folderol wrote:
Does anyone know if this is *fully* isolated from the gubbins?
Measurements with an ordinary digital meter suggests that it is, but I'd like
to be sure.
The unit I'm developing has a grounded diecast alloy case and I'd like to use
this as a heatsink, ideally with a small strip of copper in direct contact with
both, and just a smear of heatsink compound.
It's a heat spreader, not connected to the internals electrically (of
course?). What you plan to do is what all the alloy cases do that act as
heatsinks; you'll be safe. Hardest part will be to get the distance just
right: not too close or you'll bend the board and the connections, not
too far or it won't transfer heat.
Thanks for the confirmation. I was pretty sure that was the case.
My intention is to mount the Pi upside down with long bolts and springs pushing it down (it will be fully enclosed) so that should keep everything safe.
On 02-07-2021 23:23, Folderol wrote:
Does anyone know if this is *fully* isolated from the gubbins?
Measurements with an ordinary digital meter suggests that it is,
but I'd like to be sure.
The unit I'm developing has a grounded diecast alloy case and I'd
like to use this as a heatsink, ideally with a small strip of
copper in direct contact with both, and just a smear of heatsink
compound.
It's a heat spreader, not connected to the internals electrically (of course?). What you plan to do is what all the alloy cases do that act
as heatsinks; you'll be safe. Hardest part will be to get the
distance just right: not too close or you'll bend the board and the connections, not too far or it won't transfer heat.
Ref.: https://www.raspberrypi.org/blog/raspberry-pi-3-model-bplus-sale-now-35/
The SoC is a good contact point but there are a few others that get hot,
too. If you don't connect to those and keep the board upside down, you
might trap the heat.
On Sat, 03 Jul 2021 14:42:09 +0200
A. Dumas <alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get hot,
too. If you don't connect to those and keep the board upside down, you
might trap the heat.
Good point! Thanks.
Does anyone know if this is *fully* isolated from the gubbins?
Measurements with an ordinary digital meter suggests that it is, but I'd like to be sure.
The unit I'm developing has a grounded diecast alloy case and I'd like to use this as a heatsink, ideally with a small strip of copper in direct contact with
both, and just a smear of heatsink compound.
On 02/07/2021 22:23, Folderol wrote:
Would the RPi-400 be relevant in your usage?
Does anyone know if this is *fully* isolated from the gubbins?
Measurements with an ordinary digital meter suggests that it is, but I'd like
to be sure.
The unit I'm developing has a grounded diecast alloy case and I'd like to use
this as a heatsink, ideally with a small strip of copper in direct contact with
both, and just a smear of heatsink compound.
On 03/07/2021 22:00, Folderol wrote:
On Sat, 03 Jul 2021 14:42:09 +0200Ventilation is important: I once designed a small audio equaliser in a >sealed 19" rack case - it was only about 15 watts total draw, but after
A. Dumas <alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get hot, >>> too. If you don't connect to those and keep the board upside down, you
might trap the heat.
Good point! Thanks.
5 hours continuous usage it packed up - inside temperatures rose to over >100°C!
WE drilled a few holes in the case and that fixed it.
Airflow is better than transmission through even metal cases. The
problem is that with no air movement the air inside becomes an effective >thermal insulator.
Furthermore *forced* air cooling can multiply heat loss by up to a 5
times or more.
On Sun, 04 Jul 2021 08:41:30 +0100, The Natural Philosopher wrote:
On 03/07/2021 22:00, Folderol wrote:
On Sat, 03 Jul 2021 14:42:09 +0200 A. DumasVentilation is important: I once designed a small audio equaliser in a
<alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get
hot,
too. If you don't connect to those and keep the board upside down, you >>>> might trap the heat.
Good point! Thanks.
sealed 19" rack case - it was only about 15 watts total draw, but after
5 hours continuous usage it packed up - inside temperatures rose to over
100°C!
WE drilled a few holes in the case and that fixed it.
Airflow is better than transmission through even metal cases. The
problem is that with no air movement the air inside becomes an effective
thermal insulator.
Furthermore *forced* air cooling can multiply heat loss by up to a 5
times or more.
Have you any comparable figures for passive 'chimney effect' cooling,
i.e. use a metal contact heat sink on the chip(s) but fit, say, a 50mm x >400mm cardboard or plastic tube rather than a fan on the enclosure. A >benefit is that chimney effect is quiet and can't go wrong (unless you
count dead moths etc accumlating inside it, but I'm curious how effective
it would be compared with a fan.
On 03/07/2021 22:00, Folderol wrote:
On Sat, 03 Jul 2021 14:42:09 +0200 A. DumasVentilation is important: I once designed a small audio equaliser in a
<alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get
hot,
too. If you don't connect to those and keep the board upside down, you
might trap the heat.
Good point! Thanks.
sealed 19" rack case - it was only about 15 watts total draw, but after
5 hours continuous usage it packed up - inside temperatures rose to over 100°C!
WE drilled a few holes in the case and that fixed it.
Airflow is better than transmission through even metal cases. The
problem is that with no air movement the air inside becomes an effective thermal insulator.
Furthermore *forced* air cooling can multiply heat loss by up to a 5
times or more.
On Sun, 04 Jul 2021 08:41:30 +0100, The Natural Philosopher wrote:
On 03/07/2021 22:00, Folderol wrote:
On Sat, 03 Jul 2021 14:42:09 +0200 A. DumasVentilation is important: I once designed a small audio equaliser in a
<alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get
hot,
too. If you don't connect to those and keep the board upside down, you >>>> might trap the heat.
Good point! Thanks.
sealed 19" rack case - it was only about 15 watts total draw, but after
5 hours continuous usage it packed up - inside temperatures rose to over
100°C!
WE drilled a few holes in the case and that fixed it.
Airflow is better than transmission through even metal cases. The
problem is that with no air movement the air inside becomes an effective
thermal insulator.
Furthermore *forced* air cooling can multiply heat loss by up to a 5
times or more.
Have you any comparable figures for passive 'chimney effect' cooling,
i.e. use a metal contact heat sink on the chip(s) but fit, say, a 50mm x 400mm cardboard or plastic tube rather than a fan on the enclosure. A
benefit is that chimney effect is quiet and can't go wrong (unless you
count dead moths etc accumlating inside it, but I'm curious how effective
it would be compared with a fan.
Have you any comparable figures for passive 'chimney effect' cooling,Not very, or we wouldn't use fans ;-)
i.e. use a metal contact heat sink on the chip(s) but fit, say, a 50mm x
400mm cardboard or plastic tube rather than a fan on the enclosure. A
benefit is that chimney effect is quiet and can't go wrong (unless you
count dead moths etc accumlating inside it, but I'm curious how effective
it would be compared with a fan.
On 04/07/2021 13:27, Martin Gregorie wrote:
Have you any comparable figures for passive 'chimney effect' cooling,
i.e. use a metal contact heat sink on the chip(s) but fit, say, a 50mm x 400mm cardboard or plastic tube rather than a fan on the enclosure. A benefit is that chimney effect is quiet and can't go wrong (unless you count dead moths etc accumlating inside it, but I'm curious how
effective it would be compared with a fan.
Not very, or we wouldn't use fans ;-)
On 03/07/2021 22:00, Folderol wrote:
On Sat, 03 Jul 2021 14:42:09 +0200Ventilation is important: I once designed a small audio equaliser in a
A. Dumas <alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get hot, >>> too. If you don't connect to those and keep the board upside down, you
might trap the heat.
Good point! Thanks.
sealed 19" rack case - it was only about 15 watts total draw, but after 5 hours continuous usage it packed up - inside temperatures rose to over 100°C!
WE drilled a few holes in the case and that fixed it.
Airflow is better than transmission through even metal cases. The problem
is that with no air movement the air inside becomes an effective thermal insulator.
Furthermore *forced* air cooling can multiply heat loss by up to a 5
times or more.
The Pi reports a CPU temperature of around 60 deg C which I
understand is within safe limits.
The Pi is in a standard plastic case (as
sold by Raspberry Pi Foundation) with no fan;
about 45-48 C.
On Sun, 4 Jul 2021 15:54:57 +0100 The Natural Philosopher <tnp@invalid.invalid> wrote:
On 04/07/2021 13:27, Martin Gregorie wrote:
Have you any comparable figures for passive 'chimney effect' cooling,Not very, or we wouldn't use fans ;-)
i.e. use a metal contact heat sink on the chip(s) but fit, say, a
50mm x 400mm cardboard or plastic tube rather than a fan on the
enclosure. A benefit is that chimney effect is quiet and can't go
wrong (unless you count dead moths etc accumlating inside it, but I'm
curious how effective it would be compared with a fan.
You can always move more air with a fan, but sometimes a chimney
moves enough and when it does it is quieter, more reliable and larger
than a fan.
"The Natural Philosopher" <tnp@invalid.invalid> wrote in message >news:sbronb$p04$1@dont-email.me...
On 03/07/2021 22:00, Folderol wrote:
On Sat, 03 Jul 2021 14:42:09 +0200Ventilation is important: I once designed a small audio equaliser in a
A. Dumas <alexandre@dumas.fr.invalid> wrote:
The SoC is a good contact point but there are a few others that get hot, >>>> too. If you don't connect to those and keep the board upside down, you >>>> might trap the heat.
Good point! Thanks.
sealed 19" rack case - it was only about 15 watts total draw, but after 5 >> hours continuous usage it packed up - inside temperatures rose to over
100°C!
WE drilled a few holes in the case and that fixed it.
Airflow is better than transmission through even metal cases. The problem >> is that with no air movement the air inside becomes an effective thermal
insulator.
Furthermore *forced* air cooling can multiply heat loss by up to a 5
times or more.
My Pi4 is on 24/7 because I use it as a PVR for recording TV programmes. The >recordings are written to an external USB HDD (ie not to the SD card "system >disc"). The Pi reports a CPU temperature of around 60 deg C which I >understand is within safe limits. The Pi is in a standard plastic case (as >sold by Raspberry Pi Foundation) with no fan; the case feels warm but not >hot to the touch. The Pi (plus a Pi 3 for logging data from a weather >station) and the external HDD are kept in a small cupboard of a TV cabinet, >which only has a hole cut in the back to take cables. The temperature inside >the TV cabinet is about 27 deg C - so a little warmer than normal room >temperature.
One thing I noticed about the Pi4 compared with the Pi3 is that its CPU >temperature remains a lot more constant; the Pi3's CPU is normally about 40 >deg C but rises to about 80 deg C as soon as I run Firefox (so I don't run >FF on it - partly because of the temperature rise and partly because it runs >like a wounded snail on the Pi3).
On Sun, 04 Jul 2021 19:41:11 +0100, Ahem A Rivet's Shot wrote:
You can always move more air with a fan, but sometimes a chimney
moves enough and when it does it is quieter, more reliable and larger
than a fan.
Yep, pretty much what I've understood, but I was hoping that somebody
would have numbers to quantify the difference between fan and chimney cooling.
Really hard to do - fans are easy(ish) you can rate their flow in
free air, max static pressure and measure the noise and pretty much
compare them from that. Chimneys are quite sensitive to design and
conditions - a bit like the transfer ports of a two-stroke engine - a
small change in geometry can make a large difference to the airflow.
| Sysop: | Rempala |
|---|---|
| Location: | Richlands, NC |
| Users: | 103 |
| Nodes: | 10 (0 / 10) |
| Uptime: | 71:56:51 |
| Calls: | 157 |
| Files: | 6 |
| Messages: | 110,465 |