Zweiter Frühling für ION-Antrieb --- Sparta, Batavus, Koga ...

[void-spark]

Seems like a checksum failure?

00:30:07.669 -> ESP-ROM:esp32c3-api1-20210207
00:30:07.669 -> ESP-ROM:esp32c3-api1-20210207
00:30:07.669 -> Build:Feb 7 2021
00:30:07.669 -> rst:0x15 (USB_UART_CHIP_RESET),boot:0x8 (SPI_FAST_FLASH_BOOT)
00:30:07.669 -> Saved PC:0x40047ed2
00:30:07.669 -> SPIWP:0xee
00:30:07.669 -> modeIO, clock div:1
00:30:07.669 -> load:0x3fcd5810,len:0x438
00:30:07.669 -> load:0x403cc710,len:0x91c
00:30:07.670 -> load:0x403ce710,len:0x25b0
00:30:07.670 -> Checksum failure. Calculated 0x08 stored 0x49
00:30:07.670 -> ets_main.c 333
00:30:08.741 -> ESP-ROM:esp32c3-api1-20210207
00:30:08.741 -> Build:Feb 7 2021
00:30:08.741 -> rst:0x7 (TG0WDT_SYS_RST),boot:0x8 (SPI_FAST_FLASH_BOOT)
00:30:08.741 -> Saved PC:0x40047ed2
00:30:08.741 -> SPIWP:0xee

I wouldn't worry about checksum just yet, that might be normal, I'll need to check on one of my chips
Other than that the log looks normal, and it probably stops there because tx/rx get switched.

I'd just try to use it as is, see if it behaves, or reflash if you really doubt it, reflashing doesn't normally hurt
Or if you have more, try another and see if it behaves the same

At the top of the c3 config file I have how to connect it:
# Pin# from schematic(GPIO#) - Name on (ESP-01) breakout board - Name on T-01C3 board
# 1(--) - GND - GND
# 2(02) - GP2 - IO2 > Optional ADC use for battery level
# 3(09) - GP0 - IO9 > Motor on/off (active low)
# 4(20) - RXD - RX > Bus send
# 5(21) - TXD - TX > Light on/off (active high)
# 6(08) - CHPD - IO8 > Bus receive
# 7(--) - RST - RST
# 8(--) - VCC - 3.3V

You can hook up just the RX of whatever serial adapter you're using to RXD/RX and set it at 9600 baud to see if it is trying to send commands to the bus
(Yes, RX on the board is reconfigured to -send- )

 

[gpu7990]

I make today first step for speedup the motor, but still far away from result.
first job to do was to make the motor spin like I'm driving it and see the maximum speed. In this case, the motor is installed in a vise shaft. It turned out to be difficult to do this, I brought the wires from the connector of torque sensor to the outside of the motor, and disconnected the torque sensor on the shaft from the board. Outside, I made a half-bridge load cell emulator, and finally I saw a spinning motor that thinks the pedal is always pressed.
The speed of my test subject is exactly 30 km per hour. Now I can try to make changes to the board or try something with hall sensors (I don’t know yet what exactly, since there is a magnetic ring of complex design)

 

[void-spark]

I make today first step for speedup the motor, but still far away from result.
first job to do was to make the motor spin like I'm driving it and see the maximum speed. In this case, the motor is installed in a vise shaft. It turned out to be difficult to do this, I brought the wires from the connector of torque sensor to the outside of the motor, and disconnected the torque sensor on the shaft from the board. Outside, I made a half-bridge load cell emulator, and finally I saw a spinning motor that thinks the pedal is always pressed.
The speed of my test subject is exactly 30 km per hour. Now I can try to make changes to the board or try something with hall sensors (I don’t know yet what exactly, since there is a magnetic ring of complex design)

I'm impressed

 

[void-spark]

Before anyone thinks I got distracted again, well, I did, sorta
I've been working on a decoder for the log files: Decode PUT DATA, more filters · void-spark/[email protected]
I have to get confirmation still if it works (It's for windows, and should be a single big install).
The idea is people can decode their own log files, instead of having to post them here and me having to reply with a decoded one
Also, it's a lot better then my old decoder, it has a GUI, allows filtering, understands most datatypes for get/put data (as far as I guessed correctly ), and put data should be correctly decoded too now.

I have been revisiting some logs too, the one in the screenshot is from the parking one from here: Zweiter Frühling für ION-Antrieb --- Sparta, Batavus, Koga ...

Sadly the boost log from that page is I think mostly incomplete, it might have disconnected during the interesting part, but this gives a decent picture I think
I'm pretty sure the '36' command gives the throttle setting, you don't see it in the screen, but it goes up later on when the motor starts reporting it's speed is not 0, and it's being sent continuously

It still seems likely the display and motor need to be preprogrammed to support it though
Sadly what also misses is a log of powering up this system, there might be some special commands after power on that enable this behavior

Out of interest, who's still working on projects? Just to get an idea of what's going on

 

[gpu7990]

Hi I tried the latest version of cu2 and as far as I understand it has some bugs.
Test with xhp motor - does not help properly, runs at low power and continuously (when I press the pedal, the motor starts spinning and always runs)

Also, when the help is turned off after 5-10 minutes, the display starts blinking, and I cannot turn it on, it is necessary to reboot the entire system.

Also, the assistance modes do not always switch, do not respond to pressing a button

 

[void-spark]

Hi I tried the latest version of cu2 and as far as I understand it has some bugs.
Test with xhp motor - does not help properly, runs at low power and continuously (when I press the pedal, the motor starts spinning and always runs)

That sounds like lack of calibration, can you do the calibration steps?

Also, when the help is turned off after 5-10 minutes, the display starts blinking, and I cannot turn it on, it is necessary to reboot the entire system.

Would have to see logs for this one

Also, the assistance modes do not always switch, do not respond to pressing a button

Same here

 

[Mike747]

Out of interest, who's still working on projects? Just to get an idea of what's going on

Well, for something completely different :

The right batt (PMU4) will not work (the leds blink when charger connected and no signal on bow-bus).
The one on the left (connected) does charge and communicate with motor/display. When I select support, E023 appears (expected!) and the leds on the battery then blink (not sure that is supposed to happen). I need to have the motor regestrated to the batt in order to investigate further.

It does work with both CU4 (picture) and CU3 displays! I can swap them (together with holder and button unit) out.
Observations:
*CU3 works like expected: push "arrow" or "light" to switch on, light switch also turn lights on and off, goes to standby by itself.
*CU4 works slightly different: you can actually turn on and off yourself with the power button. No more button to turn the lights on and off (like expected).
*I am really surprised that they both function like they should (well for this little bit) on the same batt.

A lot to investigate further: switching out CU3 and CU4 button-units, logging, PMU3 battery in the mix, etc.