working on my 1.3 GHz PowerCube. It consists of 4 XRF286
transistors couple together using splitters and combiners.
It is based on a French design and the PCBs came from Germany.
The PCBs are mounted on heatsinks and the heatsinks are then
mounted back to back. The idea is that air is then blown over both
of them through the slot in the middle.
All fine in practice but the two PA boards don't have the same gain.
One board has 2.5 dB less gain than the other. However as they go
into compression the gain difference drops of course. I am not sure
exactly what the problem is as they were both built to be identical.
It is pretty obvious that it is a matching problem however.
If I was going to do this again I would probably obtain PCBs
from W6PQL as he uses a more suitable material than the FR4
used in the boards I have. Unfortunately he was out of stock at the
time and the amount of work I put into these amplifiers means I
wouldn't want to re-do them.
The good amplifier produces just over 100 watts output and the
bad amplifier about 70 watts. W6PQL reckons these amplifiers
are capable of about 150 Watts output each. So I have some work
yet to do.
As I said at the very start of this blog things do not always go
smoothly.
Why do I need all this power ? Well my local ATV repeater
GB3VR is not digital so I have to use GB3IV which is marginal
from my location. Hopefully one day GB3VR will add a digital
input on 23 cms (even if it remains analogue on the output).
Until then it will have to be high power, large antennas and tall
masts.
Follow up:
By carefully moving the input capacitors and a minor improvement
to the grounding I have managed to get the gain of the two pallets
to within 1 dB, that will have to do for the moment. For digital TV
amplifiers have to be run conservatively. At the moment I am
using 500 mA of bias on each device, I may increase that to
improve the linearity. The amplifiers certainly keep the unheated
shack nice and warm.
No comments:
Post a Comment