Need to heat up very cold circuit board.

Thread Starter

dmusoke

Joined Jan 4, 2010
11
Hi all:

I have an application where I'm using a -40°C rated circuit board I'm designing in a -55°C environment. Finding parts that work down to -55°C is very difficult, expensive and frustratingly inconsistent.

I am looking for a thin heater pad that I can use to warm up my circuit board up to -40°C or higher. This would be under processor control or thermal relay or something equivalent via PWM control. The critical part is the PCB mountable servo amp that's rated only at -40°C and the manufacturer is adamant not to use it below that temperature, hence the desired to keep its temperature within spec.

My board space requirements are extremely tight. The enclosure for my ciruit board is only 6.0" W x 6.0" L by 1.0" H.

Can anyone help:confused:

Thanks,
David
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
If your board is actually going to be operating in those temp ranges, you'll be hard-pressed to even find a regulator that will work.

Have a look at Roman Black's page for an OCXO (Oven Controlled Crystal Oscillator):
http://www.romanblack.com/xoven.htm

Roman's design might be adapted to your situation, using some thick-film SMT resistors and strategically placed board traces/pads to get the heat where it's needed. You only need to get the temp above -40° before you're within the temp specs.

A problem that you'll have is adhesives that won't fall apart at such low temperatures. J-B Weld won't go that low.
 

Markd77

Joined Sep 7, 2009
2,806
You can get thermal pads for heating reptile tanks which are a thin film element encapsulated in plastic ~ 0.5mm thick. I'm not sure if they go down to 6 inches square but it might be worth investigating. They are mains powered.
 

SgtWookie

Joined Jul 17, 2007
22,230
Beware that Kapton/polyimide is not acceptable for use on aircraft that the Navy or USMC use. It was banned after a mysterious series of in-flight fires were traced to the use of Kapton insulation.

If you are making a multilayer board, you could incorporate traces for heating in inner layer/layers, and bring the heat to the surface where needed using vias.
 

Wendy

Joined Mar 24, 2008
23,415
Seems to me insulation is the big problem, you can add a couple of small resistors strategically placed in a insulated bag, no regulation needed. A 15° boost in temperature is not hard.
 

mcgyvr

Joined Oct 15, 2009
5,394
It's the cold starts that will/could be a problem. I've seen many products for telecom outside plant cabinets that have no problem with the temps after they get going but starting cold just wasn't happening right.
 

Thread Starter

dmusoke

Joined Jan 4, 2010
11
Macgyver, SgtWookie, Bill and Markd77:

Thanks for all your generous suggestions. The idea of using resistors scattered across the PCB sounds really good but how to determine how long they take to effect a 15°C change? Also the idea of using heater pads from the likes of Minco and others is also a relief. I wasn't aware such devices are available:)!

As you may have surmised, the application is an aircraft one where they could turn stuff off at 40,000ft or more for an indeterminate amount of time and then turn it back on. Stuff needs to work quickly thereafter:rolleyes:!

Due to space and volume limitations, a heater blanket isn't much of an option but the ideas provided here so far seem to me a great start.

Thank you all Gents:),
David
 

SgtWookie

Joined Jul 17, 2007
22,230
Please DO NOT use the Minko heater pads in your application. I do not care if they claim NASA approval or not. Kapton is not a safe electrical insulation for use in aircraft. It is very easily damaged, it has a "memory", it carbon-tracks easily, and will very likely cause grief if used.

If you're going to incorporate Roman Black's idea in a circuit board, then run your heater supply traces on an inner layer. The resistors should be either located directly adjacent to the pads for the ICs that require heating, or underneath them. You will need a thermistor that is thermally connected to each IC to control the transistor that supplies current to the heater resistors.

To monitor the temperature, you could flip Roman's circuit upside-down (using a PNP Darlington like an MPSA64) then use mil-spec voltage comparators to measure the voltage drop across the heater resistors; if it's high, the thermistor is cold and the transistor is turned ON warming the IC's up. If it's low, the thermistor is warmed up.
 

Mogy

Joined Apr 7, 2010
1
What about Peltier Effect Module? I've never used one, but have often seen them around.
Check DigiKey and search for "Thermal - Peltier Modules"
:)
 

John P

Joined Oct 14, 2008
2,025
Unfortunately it gets quite messy if it's a question of raising the temperature of the unit rather than just keeping it always warmed up. The temperature rise time will depend on the thermal mass of the box, how well it's insulated and how much power you put into the heating. And depending on how the heating is done, you might have to worry about hot spots and cold spots, and maybe even mechanical stresses caused by uneven temperature.

Would it be acceptable to have a dual system where the box is kept warm even if most of the electronics are off? I doubt if the power needed would be all that great, and you'd get instant availability. It would save some headaches. Note that "warm" doesn't mean you can use it as a toaster, just enough to keep the circuits within specifications.
 

SgtWookie

Joined Jul 17, 2007
22,230
Very good points, John_P.

Thermal cycling will cause metal fatigue, even inside IC's - for example, the bonding wires to the substrate. The more even the temperature is kept, the less likely the board will fail due to thermal stress over time.

A serpentine trace or traces could be wound through the board on an inner layer, and fed from a relatively simple regulator, powered from a standby circuit. Current flowing through the trace would dissipate heat. How MUCH heat would be necessary depends on how well the board is insulated from the surrounding cold.
 

mcgyvr

Joined Oct 15, 2009
5,394
Not sure on the aerospace specifications but in my world it is all about ambient temperatures and components being rated for those ambients.. Spot heating the board does not necessarily allow your board to meet specifications. Preventing the local ambient from falling below 40 does meet requirements. Thats why I mentioned (and John) pointed to a system that would keep the insides of your enclosure at the correct temp range versus just local heating of certain components/areas.
 

coldpenguin

Joined Apr 18, 2010
165
What about Peltier Effect Module? I've never used one, but have often seen them around.
Check DigiKey and search for "Thermal - Peltier Modules"
:)
Darn, I was just about to suggest that too.
If you could sandwich the board such that the hot side was touching the PCB, and the cold was attached to the side of the enclosure, I reckon it could be quite good.
Also, if it is only really required for startup, and other areas could generate heat afterwards, you could in theory use it to generate a bit more electric back.

(how about add some DS1820s to the system? on my board, they are raising the temp by 3C each [this is meant to be a digital temperature sensor, not a heat generator :( ])
 

SgtWookie

Joined Jul 17, 2007
22,230
Peltier devices aka TECs are OK if one side of the device will be within around 30° of the temperature you wish to reach, but they are pretty power-hungry.

A simple resistor is actually quite good at turning incoming power into heat. Efficiency is something that must be considered, as power budgets get used up very quickly. Re-designing an aviation power system would be hideously expensive.
 

Thread Starter

dmusoke

Joined Jan 4, 2010
11
Thanks SgtWookie, Mogy, ColdPenguin and Macgyver. I will back-off the Minco based products(Thx SgtWookie) and look into the resistive heating and Peltier devices(Thanks Mogy).

When I checked out Minco heaters, they had rated their prodcust to work down to -60°C and lower. Was this a lie?
 

jpanhalt

Joined Jan 18, 2008
11,087
Thanks SgtWookie, Mogy, ColdPenguin and Macgyver. I will back-off the Minco based products(Thx SgtWookie) and look into the resistive heating and Peltier devices(Thanks Mogy).

When I checked out Minco heaters, they had rated their prodcust to work down to -60°C and lower. Was this a lie?
Minco has Kapton and other polymer-based products. I have two of the red silicone rubber heaters and they work fine. It is a reputable company. I would consider their whole product line. I do not own any stock in it.

John
 

coldpenguin

Joined Apr 18, 2010
165
Peltier devices aka TECs are OK if one side of the device will be within around 30° of the temperature you wish to reach,
This is only really true though for cooling isn't it?

In a closed system for heating such as this, isn't it a benefit?

As they are inefficient, (nothing is ever 100% efficient), with a peltier, you get the benefit of the peltier effect, it will cool the outside, and heat the inside in this case, which will act as a good insulator if nothing else, and, the inefficiency will cause resistive heating on the inside. A win-win surely?



(I have only looked at these from a hobby point of view though, the cooling we use is standard exchangers, but these are room sized).
 

retched

Joined Dec 5, 2009
5,207
You have to think about differential. If it is -40 outside the case, the peilter will only "heat" to -10ish

To overcome this, you will need increasing power. It would be better to just use resistive heating.
 
Top