Automower-fans

Morning to the Forum,
sorry if I have to write in English, my French is too rusty to be spoken,but I understand it.
So,the problem is the following.
My mower is dead, no light no nothing.
What happened.
Just for info: In October,I realized the substitution of the battery to a Li-Ion one following the schema
proposed by Poildecarotte, Thank you for that.
Then I charged the battery, with success and I placed the mower in the garage,waiting for springtime.
Last week I checked it for reinstalling on the charging station, but I realized that:
1) I forgot the switch on the panel to ON
2) switching on-off-on no sign of life
Checked the battery, was still good (20 volt)
checked the Voltages on the white connector on the MB
20 volt OK 5 volt OK 3.3 volt 0 ???
I cannot understand what happened
No sign of burning on the MB , no smell either.
Do you think I can check some more or I have to buy a new MB ?
Thanks Mauro

I think I have found the problem.
I had an ESP8266 connected to the serial port of the mower to send commands
(for opening the electric fence of a second lawn) and receiving status.
I found out that the Rx pin on ESP8266 (GPIO14 'cause I used software serial) shows 22 ohm to ground.
I do not know why, something weird.
I could imagine that can be the cause of a short on the serial interface of the automower or viceversa
a short on that caused the trouble on the ESP8266.
The result is that the 3.3 rail on the automower is killed and maybe something else.
That is very strange because I used a stepdown from 20 volt to power the esp8266.
Conclusion : this is the 4th Motherboard that goes belly-up. (no)(no)
But I need absolutely to control the mower via serial, otherwise it cannot move to the secondary lawn (??)(??)

Hello

Never connect an ESP 8266 directly on the AM serial. The ESP kernel sends spontaneously data on serial at startup or in case of crash. This data can cause motherboard corruption.

I have developed a firewall between ESP and motherboard to allow only controlled codes towards AM. You can contact me on PM if you want more information / hex code. Result will be published later on this site but...when I have time.

I have doubts that a 22ohms ground on serial would kill 3.3V line. 0v on a serial line happens all the time during communication. And 3.3V is only present when AM is on, which does not happen anymore with your AM, so you should look elsewhere.

Regards
A.

[i]Never connect an ESP 8266 directly on the AM serial. The ESP kernel sends spontaneously data on serial at startup or in case of crash. This data can cause motherboard corruption.
[/i]
In fact, just for that I used software serial.

I'm not sure, but the 22 ohm was on rx pin of the esp , suppose the AM send data and find a short ,don't know what can happen.
Just for that I was thinking of optocoupling the serial lines.

Your firewall would be wellcome,thanks
I'm sending you my email address on private message
Thanks
again
Mauro

As far as I know, Softwareserial cannot run reliably on an ESP8266 using wifi. You'll get garbage on serial....sometimes, especially when wifi transmits. It looks like it could work but it is unreliable.
This is related to interrupt management by ESP8266 kernel to manage wifi.

Generally speaking, ESP8266 is a great plateform, but do not forget that there is a kernel behind the scene, which can take the hand in unopportune moments.
That is why I used an ATTINY for AM communication: you can program it "bare bones", and manage all interactions up to the last bit. ESP8266 makes complex stuff and wifi / MQTT link, attiny makes a stupid but important thing: transfer and filter communication with AM. Nothing else, nothing less.

So your communication chain is : ESP Attiny AM ,with Attiny acting as serial interface/filter between the two.
Concerning softwareserial on esp8266 , I've performed quite a long test before connecting it to the AM, in practice I used an Arduino promini working at 3.3 volt
acting as a dummy AM (just to reply correctly to incoming commands).
The wifi communication between ESP and Raspberry was also OK .
Worked for abt one month with no glitches.
Frankly speaking I do not know what happened to the AM, for sure the 3.3V rail has gone.
Now I'm biased by this failure and I don't know how to proceed.
Probably I'm going back to radio communication that has never created problems of this kind,but had some occasional signal losses.

So your communication chain is : ESP Attiny AM ,with Attiny acting as serial interface/filter between the two.
Concerning softwareserial on esp8266 , I've performed quite a long test before connecting it to the AM, in practice I used an Arduino promini working at 3.3 volt
acting as a dummy AM (just to reply correctly to incoming commands).
The wifi communication between ESP and Raspberry was also OK .
Worked for abt one month with no glitches.
Frankly speaking I do not know what happened to the AM, for sure the 3.3V rail has gone.
Now I'm biased by this failure and I don't know how to proceed.
Probably I'm going back to radio communication that has never created problems of this kind,but had some occasional signal losses.

This is normal that you have 0V on 3.3V pin of white connector as long as AM is not on. So the cause is probably elsewhere.
Of course, if 3.3V rail is gone AM cannot go on.... chicken & egg problem.