The little attenuator that could - now can't.
Although I hadn't taken pics yet, I had transmitted using low power on 145 MHz through the 40 dB att. into my homemade power meter, and then into the hp 437B as a check. They both indicated 5.2 W. I was satisfied everything was working OK, I got ready to take pics and next transmission I got no power indicated on the 437B (which made me very nervous because there is a $300 sensor being used).
Switching over to my power meter, I got an indication of very low power (since mine reads much lower than the 437B.) After sanity checks with the 437B 0.0 dBm calibration output, I realized the 40 dB att. had developed a flaw.
I had tried before to get the internal element out of the 40 dB att. I didn't then because I was afraid I would break it, but now there was nothing left to do but get it out. I got out a small vise and a small punch, and after tapping on it gently for a while it finally began to move out.
Under the microscope I found where I think it developed a problem. There are three dark areas of a resistive material, and one has a flaw across it. I do not of course know how long it has been there or if it was worsened when I was testing it before when it got very hot.
There are three areas of resistive material, each connected in series. They also have a branch on each side to a common connection which goes to each of the three areas along the outside (on both sides). All the scratches on the silver were there when I opened it, not sure why they are there. Also the silver center trace is connected to others by a silver "paint", not sure why they were not a continuous connection between resistive areas.
You can see where each leg consisting of the resistive material has been hand-trimmed with a grinding tool, to achieve the desired total attenuation. The trimming is very inexact and rough per each individual leg. Someone is basically looking at a meter and going back and forth until the total att. is what is wanted. Maybe this design is used for all the models in this power range, trimming each one by hand. Not exactly an efficient model for mass-production but warm bodies are cheap where it's made.
Comments?