Is the most of errors ( NACK , NO REPLAY , FAIL PDT) related to new library ?

[Reza]

hi all
for a time , i have some problem , i put several post about this and some people angry for my several post. i'm sorry about this , excuse me .
for a time i read topics and i see some people have same problem. error similar to NACK , NO REPLY , FAIL PDT AND etc.
i want share with you my tests and experiments .
i want in end of this topic solve problem.
i hope every body can help , share idea about this here .
until now , i test some issue and dont solve problem .

the problem is:
after load sketch if distance was far (little) between node and gateway so first error is for cannot comunication node to gateway (some error about this is NO REPLY - FAIL PDT . . . ). also if connect , the connection is weak and most of commands is lost and dont back ack (some error about this is NACK) .

i think this is related to new library . because before update(to v2.1) i use this device with this distance (didnt any change in locations of devices and nodes and gateway) . and i didnt have any problem. also i didnt use any repeater and for connection between nodes and gateway i had a connection directly. and all of command send with out any NACK . but now (after update) for 1/5 same distance i have errors .
friends ! i use same device (hardware) and same sketch that use before update.

my experiments :
for this issue i test some hardware and software. but can not any help.
for hardware i use some capacitor for radio (4.7uf - 100uf - 470uf)
also use adapter radio (regulator 5 to 3.3)
use 3 type of radio ( 2chip and 1 chip , 100m and 1km , with antena without antena)
use some arduino nano ( ch340 and other models)
use power switching 5v (1A - 2A - 10A)
for gateway use raspberrypi3 and serial gateway (with capacitor and adapter and all type of radio)
for sketch i use pa_level (for all models of radio) :
min for gateway - min for nodes
low for gateway - low for nodes
high - high
max- max
the important issue is my devices and sketch work well before update library so i think this problem is related to update. so i want test again with library v2.0 or 1.5 and after this i report here.

regards.

[gohan]

same problem here after recompiling with 2.1.1 i got same problem, after reverting back to 2.1.0 I got communication back to work.

[Reza]

@gohan thank you for experiment. I am back to 2.0.0 and i am testing . in 2 days . i think this is work better ! Error of NACK is very decreased. I am testing in future days

[gohan]

Have you noticed any difference between 2.0 and 2.1 library in terms of reliability and range?

[Reza]

@gohan
sorry. now i have see some problem. so i think this problem is not related to library.
i dont know where is problem. i want leave a wireless sensor and start a wiring network...

[m26872]

@Reza So. You have also compared v1.x with v2.x ?

[Reza]

@m26872 hi . no i want just found this problem . because beforeupdate i dont have this problems.all command sent well. before update also i use just a 4.7u capacitor. but now use 100 , 470 u also radio adaptor and regulator . but most of command can not send .i think this is related to new lib but i test and i see this problem is yet. i dont know what am i do so i want convert my network to a wiring network!

[mpp]

The advice in this (old) thread is to downgrade to 2.0.0, is that still valid?

[sundberg84]

@mpp - I dont know, but I have found the current dev. branch 2.2 quite good and some bug fixes are made there.

[mpp]

Are you still using delays between transmissions?

I never had these transmission errors when using the rfm69 library from lowpowerlab. With that lib my nodes would run for months on 3x LR44 batteries. With these delays the same node runs for 1 day...

I wish there was a hybrid solution between lowpowerlab (awesome low level transmission) and mysensors (awesome features/framework). :(

[sundberg84]

@mpp

delays between transmissions?

Not sure what you mean - Sometimes I use wait() to get a more stable reading before sending it.
RFM69 works great in MySensors - I have built a PCB which im testing at the moment with a standard pro mini, RFM69w and 2xAA and its looking good (it will run for months, maybe year!)