With my recent experimentation with the WSJT protocols on 2m, I've run into a new problem. The FSK441 protocol is effectively 100% duty cycle while it's transmitting, and with the sequencing it's 30 seconds of transmit, 30 second of receive. This has the side effect of creating a very warm amplifier, to say the least. I'm currently using an old RF Concepts amp, 45 watts in/170 watts out.
To reduce the chances of melting your amplifier, it's recommended that you drive the amp at reduced power. That past week I've dropped my input power to about 20watts PEP. Even at that rate, the heat sink on the amp was very warm to the touch. I figured the easiest solution to solve this problem would be to put some fans on top of the heat sink to try and draw away the heat, create some constant airflow.
I went down to the work shop and rummaged through some old junk boxes and pulled out 3 working 12v fans that were pulled from an old PC (it occasionally pays to be a pack rat). So, now I just needed a mounting method. Over the course of the week I had time to think about it, and after a trip to the hardware store I came up with an overly complicated mounting bracket that would hold the fans down to the amp. I'm not sure where the quote came from, but i've always subscribed to the thought "Anything worth doing, is worth OVER-doing!".
I found I had some small aluminium strips that I had purchased at the local home center, it seemed to be the right material for the job, but I didn't quite have enough of it, so off to the store I went. Luckily for me they recently built a big box home center 2.0 miles from my house. Said company now extracts money out of me on a weekly basis.
So, with crude measurements, a few whacks of a hammer, some drilling and I have a basic bracket. I simply tied the lead wires together and crimped them into an Anderson Power-Pole connector, which is the standard low-voltage connector in my shop.
The three fans together draw about 350ma at13.8v. Ideally I'd like to find a small circuit that would detect the amp keying and would turn the fans on for a set period of time. Normal voice operations doesn't really require it, but extra thermal protection never hurts. For WSJT modes having the timer be longer than the typical sequence would effectively keep the fans on constantly, which is a good thing.
The whole assembly is currently sitting snugly on top of the amp, but it's not permanently attached. I think I can get away with not having to drill any holes into the heat sink since it's fairly tight.
Now, the dangerous thoughts are running through my mind. Since I've greatly increased the thermal dissipation, how much harder can I run my amp? Could I run at full power?
I don't currently have the ability to measure the surface of the heat sink, but I'm guessing I could probably run closer to 30w input, maybe 130w output. I think it might be time to look for an inexpensive IR thermal sensor, like I saw at show-n-tell at the last NEWS group meeting. I'm guessing Harbor Freight probably has something that will do the trick.