Inverter Firmware 5.32.R

[Yahha777]

I guess this was bound to happen. With this software there is no way to exchange throttle signal between inverters. Simply due to the lack of a running counter that can be mapped to CAN.

This suggests that the board from a Prius (Yaris) with two controllers can no longer be controlled as before, or I read the instructions upside down:https://openinverter.org/wiki/CAN_commu ... _CAN_-_new!

[johu]

No, not with this version. With the next release I will add a cyclic counter that satisfies the control message requirement. CRC checking can be turned off on the receiving side.
For now you can stick with version 5.27.R on the receiving side to retain the functionality

[Yahha777]

The "pot" and "pot2" parameters are not saved on both controllers (that is, Rx and Tx) after a power failure. Before the power is turned off, everything works and the slave controller receives messages identical to the master. Is this a feature of the firmware? Or am I doing something wrong? I haven't tried it since 27.
MG2
MG1

[johu]

Indeed 5.32 removes any mapping from pot and pot2, even TX mapping. That should be changed. So 5.27 on both sides. You need to hit Save to Flash to make mappings permanent

[Yahha777]

That should be changed. So 5.27 on both sides. You need to hit Save to Flash to make mappings permanent

Yes - everything works with the previous firmware. The settings are saved.

MG 2 sends

MG 1 receives

But nothing is displayed in the CAN map. But this seems to be a problem in the web interface. In the old (gray) mode, all CAN data is visible.

[wacomg]

Hello !
I have this setup :
Prius Gen2 Inverter
Prius Gen3 MG2 , engine is a bit loaded with another one, which is connected to another Gen2 inverter (switched off )
60v LiFePo battery
board made with prius-gen2-v1.1.zip from johu's github

firmware 5.32
defaults loaded, resolver shows angle, pot calibrated ( i use single pot)

pinswap 5 : MG2 didn't move at all with any syncofs settings
pinswap 7 : i'm able to find both syncofs

first syncofs: "enter a small value for "manualiq" and set "manualid" to 0.1, the motor" spins "smooth in both directions"
second syncofs: motor spins only in one direction

i've chosen first .

Now try to use pot and direction swith . Start pressed, direction , pot .. on small pot values it rotates , but after some point it tries to overspeed . After reading forum i found something about field weakening . Then i played with fwcurmax parameter and realised that motor works correctly with 0 > fwcurmax > -80. But in this case i can't get more than 80A

current with any values fmax,ocurlim,throtcur

Seems like fwcurmax limits max current

what to do next ?

[VoltaTheo]

Hi there !

Currently working with a Tesla SDU board (v8) with 5.34 firmware, and we are having some issues with can message reception.

We understand that since 5.32, it is not possible to remap any of the critical inputs. This means all pot, directions, start, etc. should be "pre-mapped" on control ID (63 for us). We are sending messages from our VCU (verified with a PCAN on the bus), but they are not being received/interpreted by the inverter at all.

While testing, we are still able to map the values for a brief moment, and in this case we are indeed able to receive them fine. But of course, they are then deleted since we are not supposed to remap them.

Also, in this situation, as soon as we switch the trottle mode to can or can dual, all is fine, until we do the brake check and put it in start, we get a cantimeout message. Not sure what this is due to.
After this message, we still receive our pot values, but potnom does not move, so no motor action of course.

Any idea of where these problems could come from ?
Thanks a lot !

[johu]

Can you post your code that generates the CAN message and a dump of a few messages?

[alexbeatle]

Can I just send start command via CAN and leave the rest of the data as 0 or other data must match the current state of the vehicle, e.g. also send the current pot values, forward/reverse status, etc.?

Planning to auto-start the drive per the following.

Code: Select all

  //START CTRL
  if (swIgnition && Drive_Status != Drive_Stat_Invalid && Drive_OpMode != Drive_OPMODE_Invalid && Drive_OpMode == Drive_OPMODE_Off && Drive_Status == Drive_Stat_WaitStart && IBST_driverBrakeApply!=IBST_Invalid && IBST_driverBrakeApply==DRIVER_APPLYING_BRAKES && IBST_iBoosterStatus!=IBST_Invalid && IBST_internalState!=IBST_Invalid && IBST_internalState==LOCAL_BRAKE_REQUEST) 

    {
          //send start command via CAN
          Drive_Cmd_Start=true;
    }
  else {Drive_Cmd_Start=false;}

[johu]

If you haven't configured CAN throttle or CAN cruise control you can leave these items at 0.
Also canio is logically ORed to the physical inputs.

[VoltaTheo]

Can you post your code that generates the CAN message and a dump of a few messages?

Hi Johu, sorry we didn't respond earlier. We had some tight deadlines, so decided to abandon CAN pedal signal for a while.
Coming back to it now. Did some more tests and situation is still similar.

Having a bit of trouble with attaching photos etc, but here is a link to a word document showing : throttle settings, can message sent through a Pcan, reception of the pot value, but no movement in the potnom value. And this cantimeout message which we are not sure how to get more info on.

https://estaca-my.sharepoint.com/:w:/g/ ... Q?e=YyenOX

If you need any more info, I will do my best to send it asap this time.
Many thanks