Power Handling of Rogers Board

Hi,

you have to estimate/simulate the expected preveailing temperature and keep it within the maximum operating temperature of the used PCB technology/material, see.

BR

in DUBUS 1/2015 (www.dubus.org) was an article of a ham who built a
50W amplifier with a TriQuint module on 0.635 mm Rogers 3210 material.
It burnt eventually, and TriQuint is now Qorvo. device == TGA2312FL
The board was soldered to a metal block and the output stripline was only a
few mm long with some lowpass/matching structure.

He mentions < www.globalcommhost.com/rogers/acm-techsuppothub/calculatorMWI.php >
as a calculator for temp rise, but that was in 2010 and I get error 404.

The calculator estimated the temperature as 200°C .
It seems he then contacted Rogers. With 0.5mm R4003 the rise was 80°C
and with 0.5mm 6035HTC it was 45°C .

Extrapolating that to your 400W promises problems unless it is pulsed only.
Try to get that calculator and maybe Al2O3 ceramics. Or move to waveguides.

regards, Gerhard

400W @ 50 Ohm, it makes 2.82A_rms current.
If you have low di-electric coefficient substrate, the lines will be larger but I cannot say anything for substrate power handling capability.
Waveguide Power Combiner can also be considered.

one issue is that the microstrip line will heat up with ohmic losses. So a thicker copper thickness helps by reducing the line resistance.
Also there is a trade off. If you use a thick substrate the line can be wider, but there is more thermal resistance in the dielctric layer since it is now thicker. And vice versa.

Rogers TC series boards have a ceramic filled teflon dielectric that purportedly has lower thermal resistance. That would allow a thicker substrate with a wide line, but still not end up with too high a conductor temperature

biff44 said:

one issue is that the microstrip line will heat up with ohmic losses. So a thicker copper thickness helps by reducing the line resistance.

Click to expand...

Due to skin effect, most current really pushes to the edges, and there is very little current in the middle of the trace.

The benefit of using wider lines would be heat transfer insider the copper, but as you already mentioned, wider lines requires thicker dielectric for the same line impedance, and Teflon thermal conductivity is rather poor. It is not easy to find the best trade-off.