I have received a PM from Ludwig @DH8WN sharing his experiences. I can also share it if he is fine with it?
Basically the compressor seams sensitive to stray RF. So a Commom Mode Choke in the antenna coax seams to be recommended.
Ludwig described that in some cases it was no problem i.e. with a 20m band vertical but an asymetrical dipole tuned with a ATU was causing the compressor to cause modulation problems.
He recommends the lowest compression setting to be enough.
When using an external amplifier the problem seams to increase.
We were discussing if a ferrite choke in the cable between 817 and compressor would make even more sense.
It shound be a relatively easy to replace that 8-pin RJ45 cable with another cable adding a ferrites. A normal network patch cable 1:1 should work fine.
So it seams like some tests are needed.
Ed we have a bad skip distance on 20 and 40m as it seams. But to EA it should be perfect.
Thanks Joe,
It’s good to hear that the issues, that I appear to have are being reported by others and work-arounds found.
in my case, replacing the RJ45 cable with a standard IT one with Ferrites fitted would not be so simple as I have modified that lead so that the connections are correct for the Xiegu - who use a different layout than Yaesu do on the FT-817/8. The RF ingress appears to happen even when running the x108G barefoot (but that’s up to 20 watts in any case), not only when the amplifier is running. Perhaps lining the plastic speech compressor’s case inside with silver foil and earthing that may help?
In short: all from the feeding line over the TRX until the Compressor acts as one side of a “dipole” and the Mic cable with the Op as the other side. The Compressor is located in the “middle” and is feeded with RF by this “dipole”.
I’ve decided to make a CMC between the feeding coax and the TRX (or the ATU). My Coax is DXW 174 with a diameter of only 2.9 mm. So I need only a small and light toroid for 4 or 5 turns of this cable.
For 4 or 5 turns of a IT-cable you need a larger toroid with more weight.
It works fine for 5 W. If I’m using my PA with 40 W mostly I have to switch of my Compressor.
If input and output leads are acting as RF capture devices (not a (resonant) dipole - except perhaps on 70cm?) perhaps some RF decoupling capacitors are needed in the design?? The supplied RJ45 cable is flat - so that will be difficult to put through a ferrite core. This unit has been around several years now - I’m surprised that FA haven’t updated the design to address the problems.
We seem tto have hi-jacked this thread. Perhaps a new thread just for the DYC-817 problems is needed?
a happy new year!
If we talk about a resonant dipole than it would be absolutely to short. I think, you mean resonant dipols. But so far as I know structures with two conductiv parts on bothe sides and a feeding point between may be called a dipole.
But further discussions better under the new topic.
Congratulations to all a Happy New Year 2020!
I was glad to find a gift from Santa Claus under my Christmas tree for my SOTA activations, these wonderful and very light headphones!
Dito.
(although I fear that the kit building will be above my skills. but who doesn’t try never wins).
Just in case the QRP labs elves run into issues, I am preparing to ask Santa to bring me a NanoVNA to do some decent antenna, traps and coil experimentation (and over-engineering).
The QCX Mini will probably be a lot easier than the normal QCX, since most of the SMD parts come factory soldered, including the OP amps; so much less scope for a bad path. Hans has done some really helpful videos showing how to debug your QCX once it is finished, and if all else fails, you can send it to him and he will fix it (by the sounds of it). A cheap component tester may be useful:
I also found my Sandford dissipative wattmeter to be indespensible:
SANTA & SOTA 2020 
