I don't think 0.4dB will make any difference. The algorithm (not that I have detailed knowledge of it) should be able to cope with small amounts of imbalance as one of it's major "raisons de etre".
It's a fair bet that the Flexradios do not always have <0.3dB amplitude match between I/Q channels across all HF bands.
I have not used PSDR-IQ V1.12 for years. The V2.5.2 version gets >40dB most of the time. Only Christos can really answer how good relatively the 1.12 rejection algorithm is compared to V2.5.x.
Excellent to see a product I designed in good use. Time will tell if the Mitsubishi RD16HHF FETs outlast the Heathkit SB220 !!
--- In email@example.com, tom <loeblt@...> wrote:
> Hi Rob,
> Your amp is still humming away here and the 20 watts drives my SB220 to
> about 300 watts out.
> I saw some numbers on an old posting from Bob, g8voi, that stated that
> even a .4 db difference in the channels can degrade the image rejection
> to the low 30 db range, so I was a bit concerned. I am somewhere around
> 25 db rejection at the moment.
> I thought perhaps I could insert a small pad in one channel to equalize
> the audio between the SR and the PC audio card rather than tearing into
> the Softrock but that may wind up being the only option.
> As Christos suggested, the PSDR V12.20 EIA button seems to work very
> well for me at the moment.