heating problem in a waveguide

[nashat1985]

Dear Colleagues,
I have a heating problem in a waveguide -to coaxial, transition the temperature exceeds 75 c degree, while the other rf station temperature in 37 c degree, do you any idea what is wrong?
I have attached ppt file show the waveguide transition and the thermal image shows the hotspots,
Thank you for your feedback.

SESAME Drive

drive.sesame.org.jo

 

[FvM]

73 °C isn't necessarily too high. If it's an off the shelf part, it should have a power rating spec. If you designed it yourself, your EM simulations should tell you how the power dissipation is brought up.

 

[nashat1985]

73 °C isn't necessarily too high. If it's an off the shelf part, it should have a power rating spec. If you designed it yourself, your EM simulations should tell you how the power dissipation is brought up.

Thank you FvM, yes the waveguide transition is off-shelf part and very old we do not have a technical data sheet for it, I other 3 RF stations with the same RF power drive, and reads lower temperatures, for comparison, I attached a thermal photo for RF station no 3 and it shows temperatures low and does not exceed 38 c degree, and this makes me concerned about the setup of RF station 4, your feedback is highly appreciated

 

[vfone]

If the RF stations are identical the reason of heating inside of one of the waveguides could be due to high VSWR. This could happen due few reasons as, poor connection of the coaxial probe or existence of intruders inside of the cavity. Is just a guess..

 

[nashat1985]

In the attached photo referring to 4 years ago, it has been taken before the assembly of the waveguide, it is highlighted the connection that I would be concerned about. The bolt for tightening this is in the inner conductor of the coaxial line. we will dismantle the waveguide setup and do visual inspection would tell you quite a bit. If this contact surface is damaged, we should see it right away.

 

[pgib8]

Any chance you can insert an attenuator in the coax output of the waveguide to see if the temperature goes down? This might be stupid of me to suggest because I don't even know what kind of setup this is or how much power is going through there. I will say that your job looks super cool and interesting. Please post more stuff as time goes on

 

[Azulykit]

It sounds like you have similar adapters. If you exchange the adapters does the problem follow the adapter? What s11 and s21 are you seeing with the adapter? Any sign of corrosion or a missing part? I would anticipate s11 to be better than -15 dB to -20 dB and s21 to be hardly anything, even for a venerable transition. It sounds like you have a number of adapters. You might make back to back comparisons to see if the "hot" transition is distinctly different from the others.

 

[biff44]

In the attached photo referring to 4 years ago, it has been taken before the assembly of the waveguide, it is highlighted the connection that I would be concerned about. The bolt for tightening this is in the inner conductor of the coaxial line. we will dismantle the waveguide setup and do visual inspection would tell you quite a bit. If this contact surface is damaged, we should see it right away.

if i understand your picture, there is a horizontal round rod going from left to right, and the feed attaches with a larger diameter round structure coming form the top of the waveguide?

the connection from that center connection to the rod is at a high electric field point. so there can be arcing there (check to see if there is any black pitting of the metal around that area).

also the RF current from that center contact to the horizontal rod is very high there. IF the metalic connection failed, like a cracked brazed joint, the ohmic lass will cause high temperature. see if they are firmly attached together, and can not be moved apart, even just a tiny gap will be a disaster.

the coaxial feed at the top probably has a dielectric, maybe teflon. it might be very dirty, and causing arcing also, so clean it up good.

i would have said to also check the connection of the horizontal rod to either side wall, but your thermal picture does not show any heating there. that is also a high rf current point

that IS a traditionally good and broadband way to transition from coax to waveguide. but...i am not sure it is the best one to chose for high power handling...there are high currents involved in those metal rods.