[SOLVED] Embed a heater in PCB

Hi all,

I search some experiences and accounts about embedding a heater in PCB.
I need to keep a part of my design at 80°C and, therefore, plan to use a long a thin copper track as heating element.

What do you think about this?

Thank you in advance.

Thoma

I would suggest you use a power resistor(s). The resistance of a copper track will be quite low and not well controlled. You would need a very low voltage and high current. With a power resistor you can choose a resistance to suit the voltage(s) you have available.

Keith.

Hi Keith,

If I use power resistors, it will be difficult to get a good thermal coupling between heating elements and the components that must be thermalized.

Concerning the copper track, I estimate that a 400mm long track and 125µ large has a resistance around 2.4Ω @ -25°C and 3.5Ω @ 70°C. This "low" resistance will be driven with a chopper at 500kHz to reduce L and C size.

But, what is the behavior of FR4 with this kind of constraints ?

Thoma

It is within the operating temperature of FR4. Peel strength drops with increasing temperature. I am not sure if there are other long term effects of elevated temperatures such degradation of the base material.

You can get flat heaters which may do what you want Single Sided Heaters

Keith

Concerning the copper track, I estimate that a 400mm long track and 125µ large has a resistance around 2.4Ω @ -25°C and 3.5Ω @ 70°C. This "low" resistance will be driven with a chopper at 500kHz to reduce L and C size.

Click to expand...

Sorry, but that sounds really far off good design practice. You didn't tell much about the design requirements, so I don't want to jump into conclusions. I have implemented a number of embedded thermostats, mainly for optical sensors with insufficient thermal stability. In none of these cases, a 500 kHz chopper would be a reasonable option (for signal integrity reasons, we always used DC heaters). Also static power demand has been mostly low (< 1 W), so a few ohms heater won't fit the design.

It should be considered, that using a distributed heater doesn't necessarily guarantee uniform temperature distribution, because "sinks" usually aren't distributed. I think that copper pours and/or thick ground/power planes are a better means to achieve uniform heating. But each detail has to be tailored to your application, and your requirements may be different.

Hi FvM,

My design is powered under 3.3V.
I have strong size constraint.
I will stack my board as described below:

• top layer, I will put all my thermo sensitive stuff,
• layer 2 will be the ground plane (with thermal relief to reduce sinking effect),
• layer 3 and 4 heating coil,
• layer 5 will be the power plane,
• layer 6, I will try to put nothing in the heated area.

I will power the heating coil with a step down converter (chopper). The inductor current ripple will be around 500mA and the output voltage ripple will be around 150mV.

Maximum dissipated power will be around 3W. I am not sure to be able to thermalize at 80°C when operating temperature will be around -25°C.

But this is a proof of concept and currently I try to limit disappointment.

Thoma

Put it in a box with a heater in.

I am not sure to be able to thermalize at 80°C when operating temperature will be around -25°C.

Click to expand...

Wrapping the board in insulating foam or glass wool will considerably reduce heat dissipation. Avoid thermal bridging.

Mattylad, FvM,

This will be another big challenge, because the PCB has to be flood in resin for mechanical reasons.
And for the moment, I do not know how to reduce the heat sinking as adding glass wool will be really difficult.

Thoma

This will be another big challenge, because the PCB has to be flood in resin for mechanical reasons.

Click to expand...

Molding without thermal insulation may threaten an effective heating of the board, I fear. Excessive heat dissipation must be expected. I don't know about your mechanical requirements, but a compact foam enclosure may possibly work.

You effectively want anti-thermal-gap-pad.

They do make foam strips with adhesive on one or both sides (such as 'padded' double sided tape.) May not be intended as an insulator but could work. Adhesive would hold it in place while the resin was poured. Or tacking down a piece of foam or other insulator before the resin is poured. Or there may even be some RTV like materials that have poor heat transfer that can be applied.

Also, is there some sort of thermostat controlling the duty cycle to hold it close to the target temp even as resistances change as a function of temp?

Hi TA37,

You are right; there is a regulator (PID) that controls the duty cycle to hold it close to the target temperature.

Thoma