How much power can microstrip handle? 1

[groundhog1]

Somebody just asked me if we could put 100-200 watts CW through some pc board mounted relays and filters....

Would co-planar be better? Maybe stripline but that would be a more complicated pc board.

Then they asked if we could use PIN diode switches. I think I have seen some PIN diode switches handle 100 watts, but I am not totally sure.

thanks,
groundhog

 

[Noway2]

"put 100-200 watts CW through some pc board mounted relays and filters".

I am not entirely certain I understand what you are asking.

The PCB traces will disipate power as current flows through them in proportion to their resistance (assuming low enough frequency that phenomenon such skin effect can be ignored). The resistance of the traces is a function of their geometry, width and height. The combination of geomtry, time, and power disipation affect the temperature rise of the copper. Too much temperature rise and poof - you have a fuse. This would apply to the use of any switching equipment on your board too.

As far as putting devices that dissipate 100 - 200 watts on a PCB, again it depends on the temperature rise as to wether or not it will damage the board, but that sounds like an awfull lot of power for a board.

 

[groundhog1]

The 100-200 watts is only passing through the board, being transported. It is not being dissipated totally.

But on the max power that can be carried on microstrip, I see what you are saying and it makes sense. I just don't know what the actual number of watts that typically can be passed through microstrip.

I think a substrate material with low loss and high thermal conductivity would be good.

As to what what would be better - coplanar waveguide or stripline.. which would yield lower current densities?

Then microstrip has part of the flux above in air.. that sounds like a good thing. But the fringing with all that power... is that a bad thing?

groundhog

 

[Noway2]

Transporting 100w - 200w. At what voltage? At what current?

The temperature rise of the board will depend on the IR loss of the conducting traces. 200W at 200V (ie 1A) could be easier to deal with than 200W at 2V (ie 100A) .
The answer comes down to one of materials and geometry. The clearance spacing around the conductor will be necessitated by the voltage and the amount of copper and heat disipation requirements will be a factor of the current draw.

You should familiarize yourself with some formulas for the conductivity of copper and the concept of thermal resistance, which will help you calculate the required paramaters.

 

[logbook]

It is reasonable to assume that the impedance level woud be around 50 ohms. That means 2A to give 200W.

Skin effect will certainly be relevant.

You need to make the trace wider to handle the current then you need to increase the dielectric thickness to keep the impedance constant.

Sounds difficult but not impossible.

 

[groundhog1]

Do you see any benifit of using for example, stripline versus microstrip?

Microstrip of course is a ground plane on bottom, dielectric layer, and then signal trace on top.

Stripline is ground plane on bottom, dielectric layer, signal trace, dielectric layer again, then another ground plane on top.

The fields in the stripline would be enclosed and more symetric. The fields in microstrip are mostly in the dielectric, but also some in the air above.

groundhog

 

[Noway2]

Whether or not either method is of benefit depends on your design goals. For starters, both methods have calculatable characterstic impedances based on trace geometry. Microstrips, embedded in the PCB can be more noise immune as you have two solid planes surrounding them protecting from both outside interference and helping to limit radiation. On the down side, it is a hell of a lot harder to probe or modify a trace that is buried inside of a multi layer board than on that is on top, which can mean that a $1000+ prototype goes [flush]

I am not cetain as far as power disipation goes, but I would think exposure to air could be better as you will ultimately need to get the heat away from the board, to the air. If your circuit is 2A as Logbooks suggests, you probably won't have a lot of heat build up if you use sufficently sized traces.