No problem with the pics. I have a microscope.
Hi Roger!
That "bug" is not mine! I think Andy is not doing double buffering. The trick is to decouple asynchronous processes, one provides the data, another one gets it for own processing. Here we speak about receiving through the air from a TM1000 or receiver and on the other hand about cyclic updating of displays from them, and evalution of alarm conditions and min/max on it.
The more sensors deliver data, the more issues can be observed in the display handling of a transmitter. Min/max are fed from display items.
(In standard, S32 represents 6 sensors.)
I just implemented the new "Text" sensor, interactively usable with SRXL (Rx 4649T) because S32 is getting the channel data as well (--> stick programming).
Forget about it!
- "Text" via Xbus (TM1000, ARnnnnT): The whole data handling in the transmitter screws up.
- "SRXL": Works so far but often crazy subdisplays.
Reason: This "Text" sensor consists of 9 lines. Each line needs an own data packet (16 bytes), thus it is the same as 9 additional sensors. Together w/ the other 6 sensors S32 represents - that is 15 sensors in total.
(It's anyway too slow updating the text screen via "SRXL" (not genuine, not standard conform), thus bad for interactive application.)
On top: Xbus does not use CRC (checksum).
As well: I²C (Xbus) is not very interference-resistent because it is single-ended (signal/ground) as almost all RC crap (except of DJI NAZA: CANbus, differential-ended).