General question about the WIFI amps and pushing them, how can operators tell they are being pushed to hard? Am I correct in thinking that they then generate additional side-bands especially at the low audio frequencies as they hit compression as in this effect typically seen on the satellite:-
I guess it is not just the WIFI amps but any amp that can do this.
Or am I mistaken in what I am thinking?
Adrian
Any amplifier driven too hard (making it non linear) in the RF path will shown this very behaviour (generating 3rd order intermodulation or IM3 for short). You just see them in the example as the low audio are the most prominent contents.
For those transmitting some kind of waterfall text it is more easily seen as ghost images below and above the letters as the main letter itself is sent with constant levels across frequencies (compared to SSB voice).
Hi Adrian.
I use one of those amps on a regular basis and it needs about 170mW for full output.
I did a demo at a local club a while back and forgot to put an attenuator in the drive. When I tried to send some CW the trace on the screen, which should have been a single line looked more like the PSK beacon. On SSB it looked rather like the image in your post. Easy to fix though, just turn down the drive until it looks and sounds right. The really nice thing about all these SDRs is that you have your very own Spectrum Analyser to check the signals. Are you going to Milton Keynes tomorrow?
73 David
OK thanks guys for the feedback, there has been posts about letting operators know if they are putting out bad signals, but it is good to have confirmation of what I see and think is the cause.
What I would like to see if possible from the Amsat team is a General post/sticky of problems with signals and what could be the likely cause of the effect. Put it with the Operating guidelines so that operators can easily find it and look for themselves, it may or may not be of use, but may help.
Adrian
@G0MRF It's an 8 hour drive round trip for me so I doubt it.
During QSO's I observe my signal on the Goonhilly WebSDR. If I send a long duration CW-signal it drops down after a short while. I think because of overheating. After a break the signal is OK again. In SSB I have no problems with this booster. But I think for digital modes this booster is not suitable with small dishes.
73
Thomas, I observed the same with the "4Watt" version I borrowed. Even during SSB. After mounting it on a big heatsink I had no problems.
Since end of July I'm using the SG-Labs transverter, that is working very well.
Regards, Mike
Thomas, I observed the same with the "4Watt" version I borrowed. Even during SSB. After mounting it on a big heatsink I had no problems.
Hallo Mike
The "4Watt" version has no heatsink, but the "EP-AB003" has one. See
http://www.szedup.com/products…-expansion/wi-fi-booster/
But for continious signal operation the heatsink may be too small. Thanks for your observation.
Yes Thomas, but the 4W version needs one!
Yes Thomas, but the 4W version needs one!
OK, I measured the temperature outside. It rose 4 degrees after a hour of operation. But it was a "Handgelenk mal Pi" measurement with a little thermometer. 
Right. The AB003 has the same MMICs inside but a much bigger case and heatsink. It's a good idea to switch it into standby during receiving.
Right. The AB003 has the same MMICs inside but a much bigger case and heatsink. It's a good idea to switch it into standby during receiving.
Do you mean not to connect pin 2 and 3 (modification) ? I leaved it unmodified, ha 
.
I mounted a small heatsink with cooling fins for TO-3 transistors with cable ties to the bottom. The cable ties are routed between two cooling fins thru the holes. I think it helps.
Hi,
some mods here: https://twitter.com/sp5gdm/status/1098834735512592384 and here https://www.twipu.com/SP5GDM/tweet/1123142445527707648
Today I played for awhile with the Chinese version of "4Watt" amplifier.
Here are some of my takes:
1. The 4W and 8W have same RF Circuit.. same transistors same performance.
2. The linear power output for SSB is around 1 watt.
3. Modifications by my priority:
- Short circuit the logic IC to disable the Tx/Rx switching.
- Bypass the output IC switch result by almost 1dB in output power.
- Bypassing the input divider and IC switch by direct connection of the input connector to the PA input after the switch and two cuts in the print to disconnect the unnecessary circuit. This result in additional 3-4 dB in gain which decrease the required output power and gain of the required pre-amp.
What about the built in Attenuator...
according to specs the end amp, has 20dB gain..
it seems that an attenuator is before the power splitter in the input stage..
did anybody try to remove it ?
JP
