apache-labs.com

PS works beautifully on 80, 60, 40 and 17 meters, but it always goes crazy on 20 meters. It doesn't matter whether I'm transmitting into an antenna or a dummy load, or whether I'm sampling from the built-in internal sampler or an external sampler. It also doesn't seem to matter what my PS settings are.

Here's the PS results obtained on 20 meters, which don't look anything at all like the behavior on any of the other bands:



Compare that to 17 meters:



And 40 meters:



To be honest, when PS first "locks on" to the signal on 20 meters, for the first second or so it does function as one would expect, generating amplitude and correction curves that appear virtually identical to the curves on the other bands. However, within a couple of seconds the curves rapidly deteriorate into the ugly distorted curves shown in the 20 meter pic above. Also, the splatter reduction starts out very good, but just as quickly degenerates into a quite ugly transmitted signal.

What I do as a workaround is to simply use Single Cal which (almost always) does lock onto a good set of correction curves that look just like the curves on the other bands, and the splatter reduction is also very good (very similar to that on the other bands).

So PS appears to be functioning as designed for the first second or so, which suggests to me that it's not a hardware problem (but I could be wrong).

For what it's worth, here are my PS settings:



For the record, my system has always behaved as described above.

Anyone else seeing anything like this? Anyone know what could be causing it, or how to fix the problem?

Thanks in advance,
Mark

What equipment is in between the ANAN and the dummy load when you get the 20M results that you posted. And were those results into a dummy load?

Scott, I think you've solved my problem! (btw, I've replaced the links in the OP with actual screenshots for clarity)

For reference, my dummy load is an MFJ Versaload cantenna and my amp is an AL-80B.

To answer your question, the Anan's ANT1 has always been connected to the AL-80B RF-IN jack thru a very short (18 inch) length of RG142, and the amp's RF-OUT feeds into my MFJ-989D tuner. The dummy load has always been connected to the COAX 2 jack on the tuner (my wire doublet is connected to the tuner's COAX 1 jack).

Your question made me realize that in order to remove any and all variables for testing purposes, I need the connections to be as follows :

1) For barefoot testing of PS thru the Anan's internal sampler, the dummy load needs to be connected directly to the Anan's ANT1.

2) For high power testing of PS thru an external sampler, the Anan's ANT1 needs to be connected to the RF-IN of the amplifier and the dummy load needs to be connected to the RF-OUT of the amplifier.

And here are the results:

This is 50 watts barefoot thru the built-in sampler:



And here's 750 watts thru the external sampler:



As you can see, both graphs show much cleaner curves, and the splatter on transmit is very clean in both cases.

So the only conclusion I can draw is that the 989D tuner is somehow corrupting the PS algorithm in Thetis, because if I take the tuner out of the picture then PS behaves as it should.

That means I need to find a solution to the tuner corrupting PS, but at least I know that my radio is working correctly (which is a relief)!

Thank you Scott, you are truly amazing!

Mark