Stick a mini-pc in an enclosure? (crudest pedal ever)

Started by SnRa, November 01, 2018, 11:28:29 AM

Previous topic - Next topic

SnRa

I was given an Ockel Sirius B mini-PC and I want to use it to run a complex looper patch using Max/Msp on Windows.

I was thinking of finding a nice enclosure, securing the PC in there and using a small but decent USB audio interface for stereo in/out

(also adding a pedal switch input and some pots to the enclosure using an arduino chip)

I'm worried about ventilation and heat.  Will this ever work if the PC is not designed to be enclosed?

Thanks!

vigilante397

Welcome to the forum :)

If you're worried about heat I would recommend getting a slightly oversized enclosure and adding a fan and some vents. But it's a neat idea and I hope you can get it working ;D
  • SUPPORTER
"Some people love music the way other people love chocolate. Some of us love music the way other people love oxygen."

www.sushiboxfx.com

marcelomd

Hi,

Why not go with a two part solution?
One mini PC such as the Intel NUC and one enclosure with the audio card+arduino+controls. You can bolt them to a base and treat as one unit.
Sounds easier if you need to change something later (I know, overengineering =) )

blackcorvo

Tips to avoid a hot enclosure:

- Heatsink the ICs that get hot;
- Add a fan to the back of the enclosure, pushing air out;
- Make vent holes on the bottom of the enclosure, and keep it off the floor by using some rubber feet, so that way cold air can get sucked into it by the fan;
- Keep vents and the heatsink blades oriented in parallel to each other, that way air can flow thru everything more easily;

That's essentially it.
She/They as of August 2021

SnRa

Thanks for the tips!

I'll try vent holes and rubber feet. Where would you recommend getting cheap fans of this size? 

I hadn't thought of a 2 part solution, it might work although I don't think I can really bolt the Ockel to anything as it's already in its own plastic enclosure.  I haven't opened it up but I doubt it has mounting holes.


anotherjim

It may be ok to space the enclosure lid a little with washers over the screws causing an air gap all around.
Then it needs a thru airway. It would be easy, and maybe even quite neat, to saw slots across the top edges at 45deg. Say 2 a side.
As for fixing the pc in there, maybe a couple of holes in the bottom lid and use nylon ties.
Definitely put rubber feet on the bottom.
With the pc snug against the lid, maybe saw & file slots in the bottom edge of the enclosure sides to coincide with the port/power connector positions.

Phoenix

#6
Quote from: blackcorvo on November 01, 2018, 03:15:38 PM
- Add a fan to the back of the enclosure, pushing air out;
Pushing air over something rather than pulling air over something (like you've suggested) is much more efficient - pulling will always result in air currents that have lower turbulence, ie: flowing around heatsink fins rather than through them. Directed air however goes where you point it (mostly). For a real-world analogy, consider how much easier it is to extinguish a candle by blowing rather than sucking.

Ben N

Even if you pull, the airflow is forced by the enclosure to go where it is needed (unlike trying to suck a candle out, in which case there is nothing imposing directionality on the airflow); the advantage of a side or top-mounted fan pulling is, as I understand it, that less dust is drawn into the enclosure.
  • SUPPORTER

Phoenix

Quote from: Ben N on November 04, 2018, 03:24:26 AM
Even if you pull, the airflow is forced by the enclosure to go where it is needed (unlike trying to suck a candle out, in which case there is nothing imposing directionality on the airflow); the advantage of a side or top-mounted fan pulling is, as I understand it, that less dust is drawn into the enclosure.
No, there really is a difference. If you mount a fan directly to a heatsink with a heat source attached (active device, resistor, doesn't matter), and only flip the direction the fan is pointed, blowing air onto the heatsink always results in lower temperatures. This is due to the "blow" side air being faster moving and more turbulent, which imparts greater thermal transfer from heatsink to air.

amptramp

Are you using the display and touchscreen?  If so, you should move it to the top of the box.  It does not look like they have made any internal provisions for cooling but that does not mean it will not benefit from it.

I prefer centrifugal fans to axial-flow propeller fans since they do not lose flow with back pressure and the condition of the impeller is not really important as compared to axial-flow fans.

Rob Strand

#10
When I look at the dimensions and specs  of Ockel Sirius B mini-PC.  It's pretty small 124x80mm and 13mm high.  The power is quoted at 10W.

In normal operation the thing is designed to be sitting on a table.  Some heat will sink into the table and the majority will be removed through *natural convection* and radiation from the top and side surfaces.  *Assuming 10W is realistic*, and with that size enclosure, an upper estimate for the surface temperature would be about 40C above ambient.  This looks far too high to me.  So perhaps the unit doesn't pull the full 10W.

The point is the surface temperature is pretty high.  If that unit was placed in a moving air stream the surface temperature would drop dramatically,  possibly down to 15C above ambient, because heat is now being removed by forced convection.  So the shear fact you place it in a moving air-stream means you can operate in an ambient 40C - 15C = 25C higher than before.  So if the enclosure fan has enough air flow to keep the temperature rise in the enclosure below 25C above ambient with 10W of heat generated inside then you are at least equal to the design specs of the unit.  I suspect this should be possible.

A caveat is the numbers I gave are broad estimates but you could measure the surface temperature of the unit under normal operating conditions on the bench.   When you put it in the enclosure you measure the surface temperature again and make sure it's not higher than the base-line measurement.

The thing about fans is they create noise, which isn't great for audio, so you might want a larger diameter fan blowing down onto the unit then have the exhaust holes to the sides.
-----------------
Edit: A fan-less idea would be to bolt the top surface onto the inside of a metal enclosure.  Then the enclosure gets rid of the heat through natural convection.    Some care would be required to make sure the enclosure was able to remove the heat.

Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Transmogrifox

You can probably monitor the PC on-board temperature through software.  Often this plays into the CPU governor with some operating systems (I don't know what M$FT has done with these little PC's but seems reasonable you can get to the onboard temp sensor).

If anything is going to make your computer work at the full 10W it will be processing realtime audio with a general purpose PC :)  No doubt there is thermal concern.

Regardless it is certainly worth trying it.  You have little to lose as long as you monitor PC temperature for the first little bit until you gain confidence in how it will work out.
trans·mog·ri·fy
tr.v. trans·mog·ri·fied, trans·mog·ri·fy·ing, trans·mog·ri·fies To change into a different shape or form, especially one that is fantastic or bizarre.

SnRa

Wow, thanks everyone for the advice!  This is clearly a great forum.  For this simple project I want to waste as little time as possible for the build so I can start making music, so I think I'll see if I can find some way to bolt it to a metal enclosure, following Rob Strand's suggestion, instead of adding a fan to the mix 

I won't have a display but I'll be able to display information and communicate with the looper via OSC using touchosc on android.