Replacing the MiniTiouner

[M0LNG]

It sounds like perhaps the Zero 4K is using the Broadcom SoC for the CPU and other elements, but with an separate NIM including a Si2166 demodulator. It seems like an odd choice, but perhaps the built-in tuners & demodulators on the SoC have some shortcomings which the NIM doesn't. If you feel like taking your unit apart it would be great to identify the NIM in use.

The tuner shouldn't care about the symbol rate; its job is just to apply some gain and mix the IF signal from the LNB down to baseband; all we really care about is that it has the necessary frequency range.

It's the demodulator which needs to be able to operate correctly with a low symbol rate. The main problems with demodulators seem to be (a) their narrowest baseband LPF filter setting may be too wide, allowing interference from adjacent signals, and (b) their automatic frequency correction range may be too wide, meaning that the receiver may jump to a signal other than the one requested.

The AFC range should in principle be the easiest issue to fix, since it should not be constrained in any way by the hardware. But in practice changing this value on the Si2166 did not seem to have the desired effect.

The LPF bandwidth is a more fundamental issue, as the available range will depend on the filter hardware present on the chip. But that said, the filter bandwidth should not matter too much in the case where adjacent signals are at similar amplitude, which is normally the case on QO-100.