is there a firmware and nano gateway code to reliable function of LoPy4?

[jg_spitfire]

Hi, I have an issue with the downlinks from my nano gateway, this is a project I had stopped but now I am on it again, the last year near march/april y bought the lopy4 and the downlinks were ok, I lost 1 of 10 downlinks sent, but unfortunately I lost all my files and when I retake this project I thought the best was to update the firmware and code of the nano gateway, but now the downlinks works in a random way, sometimes I sent 4 or 5 and everything ok but sometimes I lost 5/10 downlinks sent, the gateway says "timestamp error", does anybody has his nano gateway working OK?, If yes, could you tell me which firmware and code version are you using?

thanks

[robert-hh]

@jg_spitfire The timestamp error happens if the downlink message arrives too late at the gateway. The reasons for that are manifold. Just one of them could be the LoPy4. If you have Wireshark or the like, you could narrow down, if it's the LoPy4 or something in you link to the internet and the TTN gateway. Please check also if you are using the TTN router for your region.

[jg_spitfire]

@robert-hh Hi, do you know why the lopy4 stops sending the push and pull ack?

[robert-hh]

@jg_spitfire Actually no. Do you see any error messages, or does it simply stop? What are the most recent output messages when it stops?
I had seen that in the past with my LoPy base nanogateway - also stopping w/o any obvious reasons. That one is stable since long now. It only stops on mains power failures, which are pretty rare. It runs form a dump 5V wall-wart type power supply, not connected to any computer. The firmware is the basic (no PyBytes) home-built firmware version 1.20.2.rc6.

[jg_spitfire]

@robert-hh this is what I got, a big package that I have not seen before then my uplinks but no ACK push or pulls between my uplinks, after all, what are those ACKs?, thanks

[robert-hh]

@jg_spitfire The xxPy detects random packages from the noise. Their properties like length vary a lor, and rssi is typically close to the detection level, like <-120db at 125kHz bandwidth. The nanogateway does not check the CRC and forwards these packages. I do not know whether this is welcome by the server. But I could never observe a bad side effect. Other gateways do check the CRC and do not forward those packages.
Which version of the nanogateway software are you using? Is it the one from the micropython-lib?
If it stops, does it stop forever or do it revive after a while, like after 40 or 80 minutes? I agree, that is a long time. But that is the ticks_cpu() overrun period.
pull ack and push ack are messages of the Semtech packet forwarder protocol. See https://github.com/Lora-net/packet_forwarder/blob/master/PROTOCOL.TXT

[jg_spitfire]

@robert-hh Hi, I am using the firmware version you told me and now works, and the nanogateway version I have been using is the provided by pycom