Hello,
I suspect that the 10.7 MHz Output is generated using the baseband IQ stream and an up-converter since the receiver uses direct sampling. The +/- 5 MHz is likely the guaranteed bandwidth due to possible falloff when you get close to bandpass filter edges.
This may give a clue as to the USB IQ sample rate and bandwidth.
I honestly had never tried to use the 10.7 MHz IF on HF since anything I was going to put on the IF would natively do HF. So I had never considered it. I had only done VHF / UHF with an SDR on the IF of the 8600. And I had never done any measurements, just eyeballed the responses. So I had to grab the trusty 8640B and see what the results were.
As a side note, on HF I found the S meter to be very close to calibrated, typically within 1 dB of correct from -10 dBm to below -100 dBm. And it passed on the correct level to the 10.7 MHz IF. If a tuned signal at 20000 kHz was -60 dBm at the R8600 antenna input it was about -60 dBm at the 10.7 MHz IF, +/- 1 dB or less. I was suitably pleased.
On HF I did my measurements around 15000 kHz, I only did them near the center of the HF portion of the radio. Maybe latter I will run test from one end of the sub 30 MHz band to the other.
I used a 15000 kHz signal at -60 dBm as the starting point, and left the R8600 tuned to 15000 kHz the whole time, adjusting the sig gen output to find the band flatness and response. I found the 3 dB bandwidth to be just about 8.7 MHz wide, but it was asymmetrical.
With a -60 dBm signal from the sig gen and the R8600 tuned to 15000 kHz the IF was 3 dB down, at -63 dBm, at 12320 kHz and at 21035 kHz So tuned to 15000 kHz the IF signal was down 3 dB at -2.68 MHz and again at +6.035 MHz.
But that was at the -3 dB points, how about the -10 dB points? On the low side the -10 dB point was at 11110 kHz, and on the high side at 22630 kHz. About 11.52 MHz of width at the -10 dB points and still asymmetrical.
Then I ran the whole thing again at 155.0 MHz out of the sig gen. The R8600 S meter was not quite as accurate, about 1.5 to 2 dB off, but the IF output was still pretty much on the money. A -60 dBm signal at 155.000 MHz out of the sig gen still yielded -60 dBm at 10.7 MHz, even though the R8600 S meter showed -61.8 dBm.
On the high side the -3 dB point was at 161.275 MHz, and the -10 dB point was at 162.5 MHz. On the low side the -3dB point was at 148.55 MHz, and the -10 dB point was at 147.35 MHz. So the 3 dB bandwidth of the IF with the R8600 tuned to 155.0 MHz is about 12.725 MHz, and the 10 dB bandwidth is about 15.15 MHz.
Although wider, notice how much faster the VHF/UHF IF bandwidth fell off.
I don't know about the HF IF, but I bet the VHF/UHF 10.7 MHz IF is not related to the IQ stream at all. That spectrum has to be down converted so the direct sampling (on HF) AD can handle it, I suspect the same down conversion is providing the IF on VHF/UHF.
T!
