Old Version Works, latest crashes

[Jay-Ti]

Well I'm at a loss here,

I have a 4 wire Miami 54123 with a V1 Board.
Having installed v3.3.1 it ran reasonably fine apart from a few crashes every few days or so.
But installing the Latest version or the dev version (or even 3.4.0) it just keeps on crashing on the Hardware Watchdog.

I've narrowed it down to the CIO. Having it disconnected, The ESP boots and stays online for a long time.
But connecting the CIO makes it crash Instantly after booting. And does nothing when connecting it after it booted.
As far as I've seen the pinouts haven't changed, so since it worked on 3.3.1 i assume my pinout is correct for the latest versions and hasn't changed since it worked on 3.3.1. Especially since the DSP is working.

I have cleared the flash, I've recompiled and re-uploaded everything including a new file system, i got the same settings defined the V1 as the board.

Any ideas or help to get the latest Versions running would be Highly appreciated.

[visualapproach]

Turn off mqtt. Close all browsers. Update from usb (occasionally I get weird reset loops when updating via OTA)

[Jay-Ti]

Managed to get 3.5.0 working and pretty stable, can control from dsp and Webpage.
Using the same Hardware Setup, But updating To 4.1.1 and uploading the corresponding file system and restarting everything,choosing 54123 in the hardware setup and restarting again, i get nothing, 04 from the dsp, no control over webpage(cant even toggle god mode).

now, the working setup is cio rx to D3 and tx to D2 , Dsp rx to D6 and tx to D7.
Choosing 54123 for CIO and DSP and V1 board in Hardware config on 4.1.1 flips the DSP pins to rx 7 and tx 6.
But even using custom pinouts and changing them back with a D, or using the GPIO numbers(and power cycling everything) doesn't lead to anything working.

Am i missing something i haven't tried yet? Or should i just give up and embrace "never touch a running system"?

[visualapproach]

The pinout in v3.5 was the same as v4 for PCB v1. I tricked myself too because RX/TX is not in the same order on the CIO and DSP:
const int CIO_RX = D3;
const int CIO_TX = D2;
const int DSP_TX = D6;
const int DSP_RX = D7;

The god mode will exit as soon as an error is shown, so if you have E04 you need to reset the pump before god mode can work.
Internet says E04 is overheating error. Could it be overheated or should I see if the ESP reports F instead of C to the display maybe. What if you only use C as a test.

[Jay-Ti]

Using the pinout you just described nothing changes.
04 is not E04, 04 on the dsp Segment display is what you get when there is no communication to the DSP, at least from my two 54123 pumps.
So yeah, no communication between the ESP to the CIO nor the DSP.
Although, is the "Temperature Actual" Value always derived from communication with the cio? Because i got a value there, i just cant do anything else since God mode doesnt activate. Also cant change the Temperature Unit if theres no communication :P

[visualapproach]

I'm looking at the god mode issue. Temp is initialized to 25C but then derived from CIO messages

[visualapproach]

I have fixed a bug in previous beta version. It didn't update messages to the display if not in god mode. I also made it possible to enter god mode without good communication. Plz try update to beta version. (You can do it from the web ui)