Where are we now?


netlinker

Member
Joined
May 21, 2015
Messages
44
Location
Bavaria, Germany
I know, white is cool because it reflects all light. So it would seem white is the best colour for cooling.
That's not entirely true out of the box. White does not absorb Light but reflects it instead, which is good, but it also does not emit light in this wavelength very well, and this is bad.
Every body at room temperature radiates at a wavelength of several µm and thereby looses energy. The amount of heat they loose is dependent on the surface area as well as on the emissivity (black -> good emissivity, white-> bad emissivity). Thats also the reason why most heatsinks are anodized in black.
This radiating loss of energy goes with the temperature difference in T^4 so rises rather quickly. At lets say 80°C cpu temp (wild guess based on a calculation I did a while ago) the Pyra would probably loose more heat by radiance than convection (Airflow).
To sum it up, you would want a material that absorbs as less light as possible (white) while radiating as much as possible.
As sunlight contains most of its energy above 3µm wavelength and the emitting wavelength of the Pyra has a peak at about 7µm, you would need a color that has high emissivity at high wavelengths and low emissivity at low wavelengths. This is exactly like radiator paint works, as its visible white, but infrared black.
If you operate the Pyra in shade (almost no external light reaches the pyra -> no absorption -> no heat) you can also paint it just black to enhance emission and run it cooler.
BUT: this is only true if the case is black in infrared wavelengths. A lot of plastics are more or less transparent to thermal radiation (thats actually the reason why a cheap infrared thermometer has a plastic fresnel lens). Therefore it is not the case that is emitting, but probably the pcb beneath it.
So it could be very well that a plain white plastic case would be the best for cooling, as it reflects all the visible light and lets the radiation of the pcb pass through (depending on the plastic). But this is maybe something azouls knows about the plastics if the manufacturer releases a datasheet containing transmissivity curves.

for explanation of black body radiation see:
Also explanes human body radiation and a nice image showing the translucency of plastics.
BTW: A human body looses approximately 100W via radiation only according to the article.
Post automatically merged:

Or brushed aluminium :)
From my experience brushed aluminum heats up quite significantly in the sun, better make it chrome, or even better ... GOLD :)
 

Phlyra

Active Member
Joined
Jul 25, 2019
Messages
300
With a 3D printer you could use a high % content Aluminium-fill plastic filament and print in ridges/cooling fins, like on motor casings but rather smaller so more appropriate for our use case (if you’ll pardon the pun) - melt the plastic content away and turn the case (or parts of it) into a textured surface heat sink?
 
Last edited:

netlinker

Member
Joined
May 21, 2015
Messages
44
Location
Bavaria, Germany
With a 3D printer you could use a high % content Aluminium-fill plastic filament and print in ridges/cooling fins, like on motor casings but rather smaller so more appropriate for our use case (if you’ll pardon the pun) - melt the plastic content away and turn the case (or parts of it) into a heat sink?
My wild uneducated guess would be that the Aluminum filling has no structural integrity, as it was printable before. So if you melt away the plastic, the aluminum remaining would be plain powder and melt away with the plastic, but why would you want to melt away the plastic? Thermally conducting plastic would be good enough.
 

Phlyra

Active Member
Joined
Jul 25, 2019
Messages
300
My wild uneducated guess would be that the Aluminum filling has no structural integrity, as it was printable before. So if you melt away the plastic, the aluminum remaining would be plain powder and melt away with the plastic, but why would you want to melt away the plastic? Thermally conducting plastic would be good enough.
No, no it’s a deliberate process to obtain pretty much 99% purity of Aluminium in the final product. I don’t know all the ins-and-outs of it as I’m just learning about it myself but you can also do the same with Copper, Brass, Bronze, Steel etc.
 

netlinker

Member
Joined
May 21, 2015
Messages
44
Location
Bavaria, Germany
No, no it’s a deliberate process to obtain pretty much 99% purity of Aluminium in the final product. I don’t know all the ins-and-outs of it as I’m just learning about it myself but you can also do the same with Copper, Brass, Bronze, Steel etc.
Ok, but then it would need some kind of sintering, does it?
Sounds cool though, even if you need the sintering, as this avoids a sls printer and FDM printers are easily available :).
 

Phlyra

Active Member
Joined
Jul 25, 2019
Messages
300
Ok, but then it would need some kind of sintering, does it?
Sounds cool though, even if you need the sintering, as this avoids a sls printer and FDM printers are easily available :).
Sort of - not with a laser but with an oven/kiln
 

Yogendra

Still Fresh
Joined
Aug 3, 2017
Messages
24
Age
71
I hope you're all doing fine and healthy during these weird time in our lives.

Unfortunately, Corona has hit the Pyra as well, but we're still moving forward.
I've been extremely busy taking care of the shop. Adding new features (like a "pay what you like"-section), doing marketing, adding new products, etc.
Sales have dropped - so I needed to take care to make at least enough money to keep the shop alive. And while our profits are not as good as before, it's still profit and not a loss, which is the most important thing.

Money is still tight, so this was an important task... and it's not over yet.

That was the main reason I wasn't seen that much here or elsewhere - I'm basically working all the time.
So if I made some of you worried: My health is fine :)

And now for something completely different: The Pyra :)

1. What's still missing?

As you probably know, I have 1000 full case sets in my office. Batteries, LCD Cables, Nubs, LCDs... it's all here :)
The heatpipe-material has also been ordered and is planned to be delivered in 5 weeks.
The heatsinks itself also have been ordered - these should arrive within the next 1 - 2 weeks.
Boxes and manuals also haven't been finished, but the production of those happens in Germany and takes around 2 weeks.
I am also waiting for the samples of small silicone cutouts which should shield the battery from the CPU heat a bit. These come from Poland and have been shipped a while ago - but have not arrived yet.
The mass production itself can be done in one week, as these are simple laser cutouts, which is cool :)

Everything else is already here. Yay!


2. What needs to be done?

Well, we still need to coat the cases and make a final color decision.
As I've posted last time, the coating company closed down for a while (most probably because Audi stopped production for a while as well, and they're mainly doing stuff for Audi... pretty normal here in Ingolstadt :D)
They resume work tomorrow, so the samples shouldn't be too far away.

And then only the shells need to be coated.

Global Components has also already started populating the PCBs. These will be finished in about 2 - 3 weeks.

Then it's time to assemble. To save money, we'll assemble them here in our company (well, it will probably be mostly me).
Addtionally, we can see if the assembly can be improved in some ways. This practice has worked well with the Pandora as well!

The biggest pile of work is the OS though.
It seems with Corona, the already sparse free time of developers is even more sparse than before.
That's why Toastbucket and another developer I have here in Germany who wanted to work on remaining issues can't do much at the moment.

Luckily, aTc still builds me some images, so since today, I have an image that can at least boot on the mass production model (which had some changes compared to the final prototype).

The TILER Code needs to be ported to a more recent kernel version though, and the audiodriver seems to be fully broken at the moment... but yeah, at least it boots :)

Hopefully, we will be able to fix this somehow until we start with the delivery... otherwise, I hope that we will have a few interested and motivated developers who ordered one of the first 500 units.
That would help tremendously :)

Okay, that's it for today.
Stay safe and healthy, let us stick together and get things to work :D

As usual: Any questions are welcome here.
Post automatically merged:

I hope you're all doing fine and healthy during these weird time in our lives.

Unfortunately, Corona has hit the Pyra as well, but we're still moving forward.
I've been extremely busy taking care of the shop. Adding new features (like a "pay what you like"-section), doing marketing, adding new products, etc.
Sales have dropped - so I needed to take care to make at least enough money to keep the shop alive. And while our profits are not as good as before, it's still profit and not a loss, which is the most important thing.

Money is still tight, so this was an important task... and it's not over yet.

That was the main reason I wasn't seen that much here or elsewhere - I'm basically working all the time.
So if I made some of you worried: My health is fine :)

And now for something completely different: The Pyra :)

1. What's still missing?

As you probably know, I have 1000 full case sets in my office. Batteries, LCD Cables, Nubs, LCDs... it's all here :)
The heatpipe-material has also been ordered and is planned to be delivered in 5 weeks.
The heatsinks itself also have been ordered - these should arrive within the next 1 - 2 weeks.
Boxes and manuals also haven't been finished, but the production of those happens in Germany and takes around 2 weeks.
I am also waiting for the samples of small silicone cutouts which should shield the battery from the CPU heat a bit. These come from Poland and have been shipped a while ago - but have not arrived yet.
The mass production itself can be done in one week, as these are simple laser cutouts, which is cool :)

Everything else is already here. Yay!


2. What needs to be done?

Well, we still need to coat the cases and make a final color decision.
As I've posted last time, the coating company closed down for a while (most probably because Audi stopped production for a while as well, and they're mainly doing stuff for Audi... pretty normal here in Ingolstadt :D)
They resume work tomorrow, so the samples shouldn't be too far away.

And then only the shells need to be coated.

Global Components has also already started populating the PCBs. These will be finished in about 2 - 3 weeks.

Then it's time to assemble. To save money, we'll assemble them here in our company (well, it will probably be mostly me).
Addtionally, we can see if the assembly can be improved in some ways. This practice has worked well with the Pandora as well!

The biggest pile of work is the OS though.
It seems with Corona, the already sparse free time of developers is even more sparse than before.
That's why Toastbucket and another developer I have here in Germany who wanted to work on remaining issues can't do much at the moment.

Luckily, aTc still builds me some images, so since today, I have an image that can at least boot on the mass production model (which had some changes compared to the final prototype).

The TILER Code needs to be ported to a more recent kernel version though, and the audiodriver seems to be fully broken at the moment... but yeah, at least it boots :)

Hopefully, we will be able to fix this somehow until we start with the delivery... otherwise, I hope that we will have a few interested and motivated developers who ordered one of the first 500 units.
That would help tremendously :)

Okay, that's it for today.
Stay safe and healthy, let us stick together and get things to work :D

As usual: Any questions are welcome here.



Please do not use black rubber paint anywhere because over time it becomes sticky and deteriorates
 

Confuzzled

Active Member
Joined
Oct 1, 2018
Messages
70
So black would be cooler in the shade? I think I once heard something like that; someone saying that radiators would be more effective if they were black instead of white.
The emissivity of materials in the IR specturm is often wildly different to the visual spectrum colour. The colour of household heating radiators is probably negligibly correlated with their infrared emissivity, and the same is likely true of the Pyra coating.

To quote the Wikipedia article:

With the exception of bare, polished metals, the appearance of a surface to the eye is not a good guide to emissivities near room temperature. Thus white paint absorbs very little visible light. However, at an infrared wavelength of 10x10⁻⁶ metres, paint absorbs light very well, and has a high emissivity. Similarly, pure water absorbs very little visible light, but water is nonetheless a strong infrared absorber and has a correspondingly high emissivity.
 

netlinker

Member
Joined
May 21, 2015
Messages
44
Location
Bavaria, Germany
Yes, it's also worth considering that a pyra case geared to the greatest emissivity should also be cold to the touch at rest, and would also absorb hand heat when holding it.
Unless you have a big sink for infrared radiation, like a clear sky, the pyra will reach equilibrium when it reaches the temperature of its surroundings. So most probably it won't feel cool to the touch. Thats mainly a property of materials with a high thermal conductivity in conjunction with a relatively high thermal capacity. For example aluminum or ceramic reinforced plastics.
 

levi

Still fresh, damnit!
Joined
Oct 6, 2008
Messages
13,001
Location
Somewhere off the coast of the EU
That's true. My thinking was clouded by this talk of radiators, which are cold to the touch when off, largely because they're full of cold water. Other metal things like door handles don't feel noticeably cold when touched. Although my enamelled cooker does feel noticeably cooler than the surroundings probably because of the large amount of steel in there wicking away my body heat when I touch it. As you say, things need a high thermal conductivity combined with a large mass to feel significantly cooler.

But a metal pyra would still wick away body heat reducing its thermal headroom, although to what degree I'm not sure. My (plastic housed) laptop has a hot spot that develops when it's running anything beyond mp3 decoding that's continuous, so I put my hand on it and wick heat into my circulation a few minutes after starting to play a video which helps it to not miss key frames when playing back.
 

netlinker

Member
Joined
May 21, 2015
Messages
44
Location
Bavaria, Germany
When I get my pyra I will eventually try to mill me a aluminum case to keep it cooler, as I tend to do heavy stuff on underpowered devices ... my Raspberry pi can tell stories ;). Has gotten a custom milled case meanwhile. Will take some time to design for the pyra tough I guess.

Have you checked your cpu cooler for dust? Mine usually just runs hot when it's dirty.
 

Phlyra

Active Member
Joined
Jul 25, 2019
Messages
300
When I get my pyra I will eventually try to mill me a aluminum case to keep it cooler, as I tend to do heavy stuff on underpowered devices ... my Raspberry pi can tell stories ;). Has gotten a custom milled case meanwhile. Will take some time to design for the pyra tough I guess.

Have you checked your cpu cooler for dust? Mine usually just runs hot when it's dirty.
Do you leave spaces open for signal to reach the antennas or do you only connect to your RPi via Ethernet cable?
 

mrtdnb

Still Fresh
Joined
Dec 31, 2012
Messages
3
"Well, actually all radiators are black, just in the infrared spectrum "
That is interesting, I have learned something new thanks
 

netlinker

Member
Joined
May 21, 2015
Messages
44
Location
Bavaria, Germany
Do you leave spaces open for signal to reach the antennas or do you only connect to your RPi via Ethernet cable?
I'm planning to have plastic covers where the antennas are. My raspberry has gotten an external antenna for this purpose, but for the pyra a sma sticking out at the back is not really cool.
 

Phlyra

Active Member
Joined
Jul 25, 2019
Messages
300
Does titanium interfere with EM-signals? Or, what about strong ceramics?
It’s still electrically conductive, a lot less than Aluminium but more so than Stainless Steel, so I imagine it would most likely still create a Faraday Cage effect?
 
Top