Altair 183C pro

[knobby]

Ok, so it's windy / cloudy / raining ...

I'm bored, fancied a project, turns out that a 30 X 30 X 3.2 peltier will fit under the heatsink ! Mmmmmmm 

Got one and some compound coming next week along with longer (but not too long) screws.

Gotta be worth a bash ?

Will a dew controller be able to vary the cooling by PWM ?

[Gina]

Peltier TECs don't like PWM and it reduces the efficiency as well as working life.

[andyboy1970]

Watching with interest.

How are you going to deal with sensor icing/ optical window icing?

[skybadger]

Peltiers respond to linear current control  - ie big current transistors sourcing 5v at 7A. They didn't use to like power cycling quickly either since they delaminate due to thermal shock. I gather they are better these days. 

[knobby]

54 minutes ago, andyboy1970 said:

Watching with interest.

How are you going to deal with sensor icing/ optical window icing?

Just dabbling, but I'm aiming for around 5 to 10°C only so shouldn't frost up ... If it even works ! The heat sink is pretty small so not expecting much.

[knobby]

Just found this calculator, might not be too bad actually.

 

[Adam J]

Really don't over do it with the power, start out at 10 watts or so of it was me. 

You can 100% get a larger heatsink in there, it will help to move the fan outside the case to make room. 

I am actually quite worried about this as even in comparison to a dslr this is quite open to the cold side being exposed to air flow....little condensation on those electronics will leave you with a fancy paper weight.

Hence why people use hot glue to protect DSLR electronics from condensation when they mod them for cooling.

I was always willing to give it a shot with a 450D or a 550D as second hand your only losing £120-£180 pounds if you brick it, this is a whole lot more expensive.

Adam

[knobby]

1 hour ago, Adam J said:

Really don't over do it with the power, start out at 10 watts or so of it was me. 

You can 100% get a larger heatsink in there, it will help to move the fan outside the case to make room. 

I'm not sure how to adjust the power (been googling) but only intend to go down to 5°C so won't be pushing things .

I might try using a 9v cell as a test but really need to research it before bits arrive.

This is the peltier model

https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.co.uk%2Fulk%2Fitm%2F392193270647

[Adam J]

1 minute ago, knobby said:

I'm not sure how to adjust the power (been googling) but only intend to go down to 5°C so won't be pushing things .

I might try using a 9v cell as a test but really need to research it before bits arrive.

This is the peltier model

https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.co.uk%2Fulk%2Fitm%2F392193270647

I put down some additional comments above just before you posted. I think a better goal as you say is to reduce the temperature by about 5c but more than that to control the temperature and keep it at 5c under ambient that way its more of a mini set point cooler that allows for constant temperature for the entire imaging session.

I still think this is risky with the proximity of the electronics to the cold side of the TEC.

Adam

[skybadger]

Two things: reduce load by surrounding Peltier in form fitting insulation foam and coat the electronics in conformal coating - varnish if you like. 

 

 

[knobby]

When I take it apart again, I think I might be able to silicon seal a barrier of some sort between the main board and the heat sink. The fan / heatsink/ peltier will be separate from the electronics ... Well fairly separate ?

Just to reiterate, I'm not aiming for freezing temperatures, just trying to make the cooling a bit more effective.

[knobby]

Assembled at lunch today, running peltier at 12 v direct the reported sensor temp drops from a very dubious 28°c (it's not that warm in here) to 23°c in about 2 mins then stabilises at around 24. It's very tight inside and I guess the heatsink is warming up enough to 'fight' the cooling.

Put some cut up transparency sheet between boards as protection.

I'll give it a bash outside when possible.

Good fun anyway ?

[knobby]

 

 

So, it seems to drop about 6 °c in seconds then creeps back up a degree or 2 then stabilises. That's 6°c more than the fan alone.

[Adam J]

The fact it bounces back like that is a sure indication that your TEC is too powerful for your heatsink. 

[knobby]

9 hours ago, Adam J said:

The fact it bounces back like that is a sure indication that your TEC is too powerful for your heatsink. 

I have got a temp sensor / controller coming on Tuesday so maybe it will work better when it's not full on all the time ?

[knobby]

I won't bore with another video but I've put some loose polystyrene to insulate hot from cold side and placed some strategically to force the airflow through the fins on the heatsink.

Using 6v it went from 21.3 c to 12.5 c and stayed stable.

Any more than 7 v and it starts to rise again. Not too shabby for what is a completely reversible mod.

Moving forward I'm not sure how to control the temperature as the controller I've ordered just turns on and off the power but there is so much overshoot when the supply is cut.

[knobby]

So, to bring this to a close, a 5 minute dark at 12° and 18° stretched the same look virtually identical. Maybe stacking will show more of a difference but for the sake of this experiment and having very poor control of the temperature ( which would have been its biggest advantage) it's all been put back to normal.

But I had fun ?