Change RxBw to proper value for 90KHz TX-dev.

[mvdswaluw]

0x94A2 means 134KHz RX-bandwidth. With +/- 90 KHz FSK deviation 200 KHz is a better choice. Else modulation is lost in the RX-filter.

[JohnOH]

Will the two configurations interwork?

[mvdswaluw]

Please define "interwork"? Do you mean TX deviation should be matched to RX bandwidth and vice versa? If so, the answer is yes.

[JohnOH]

Will a node configured in the old style be able to successfully TX & RX with a node configured in the proposed new style. In essence do I to reprogram all my existing nodes to maintain connectivity?

[jcw]

I think there's some confusion about how the RFM69 interprets that b/w setting - I'm not 100% sure, but it might be that the 125 KHz config is interpreted as ±125 KHz, i.e. 250 KHz swing. Maybe there is some way we could check this?

[jcw]

reprogram all my existing nodes to maintain connectivity?

It's an RX-only setting, and if Milan is correct, it would merely increase RX sensitivity slightly.

[JohnOH]

I have made the change and see nothing unwelcome.
Is there an equivalent change to the rf69 to consider?
Perhaps the comment could be updated to 200kHz?

[mvdswaluw]

Only reception is degraded, but it will work if nodes are close to each other. Thankfully the filters will let the +/- 90 KHz pass, but attenuated. I found this out because some of my nodes were a bit off frequency and de-tuning the RF12demo resulted in loss of reception quicker than I expected. So I started looking into the RF-settings. This one was the first obvious.