search

r00li / Aibo-ERS-xxx-battery

A replacement battery for the Sony Aibo ERS-2xx/ERS-3xx/ERS-7
Other
18 stars 3 forks source link

Firmware Programming/Reading Issues #3

Open ahblair4 opened 11 months ago

ahblair4 commented 11 months ago

I'm finding that the MAX17320 software is halfway reading my device then giving up, displaying "NO SLAVE DEVICE" and refusing to program. It will initially return some data, then cut off the connection. Any ideas? image

r00li commented 11 months ago

Hi! The MAX software is a bit crappy sometimes. But I don't think I had any issues with this. Generally the "NO SLAVE DEVICE" means that it cannot connect to the fuel gauge chip if I remember correctly.

What are you using to program the chip? The official MAX development kit or FT232H interface? If using the FT232H, did you reprogram the chip as mentioned in the guide? What error do you get if you don't have your programming interface connected to the computer?

If the interface is working fine, and you checked your connections, then the other thing that you need to do is "jump start" the chip when you first connect it to the cells. Did you do this?

ahblair4 commented 11 months ago

Thanks for the reply!

I’m using an FT232H reprogrammed as shown in the guide, but upon checking the clock line with my scope, the clock seems to sporadically drop out, so I have another board on the way to hopefully remedy this.

With the FT232H unplugged, I see a ‘No USB adaptor’ error, so it appears everything is fine for communication between the FTDI and PC. I jump started the cells initially, and after I soldered in a new MAX IC. I’ll update you hopefully Saturday when the new interface arrives.

r00li commented 11 months ago

Interesting. Yeah, the clock dropping seems a bit weird. Let's see what you get with a new interface. Did you use something with an official FTDI chip (like the Adafruit board) or one of the Chinese ones? Generally it shouldn't matter, but FTDI software is a bit... fishy as we know 😄 .

That sounds good then. Could also be a dead MAX chip of course. I think last time I made a bigger batch of boards, 3 of the MAX chips simply failed to program.

Actually I now remember that I had a board with a similar issue once. It would connect, but then disconnect almost immediately. I think in the end I scrapped the complete board, because just replacing the MAX chip didn't help - most likely it was an issue on the PCB that I couldn't find.

ahblair4 commented 11 months ago

Appears it was something goofy in my soldering, all good and flashed now. I am, however, unable to see charge progress and consistently get the chime of death after reinserting several times. Is there a chance the firmware didn’t flash correctly even when it said it did?

EDIT: I've come back to my workstation and have it hooked back up. I went to reflash and was given an error stating the Permanent Fail needs to be resolved before programming. I don't have a permfail light on, but CHG FET and DIS FET are both lit up, so is this something to override?

I found where you had a PermFail ( https://electronics.stackexchange.com/questions/601411/max17320-going-into-permanent-fault-upon-restart ) with the same address nBattStatus 0x8000 and I assumed changing this value to 0 as you had would let me flash finally, but that did not stick and just burned up a write. image

r00li commented 11 months ago

Hmmm. Interesting. Disconnect the battery, restart the software then connect back. If you look at SBS tab you should see the battery reporting the manufacturer as SONY. If it does then it would mean that the programming was done correctly.

As for perm fail... the issue that I had was completely due to the MAX software being weird. Any time I would change the configuration it would set the perm-fail bit, which would then be programmed into the chip. Did you use the configuration file from this repo? If you didn't make any changes and programmed it directly, then it shouldn't be going into perm fail.

One thing that I am seeing here is that it is reporting quite low state of charge. Any chance you could charge up the cells using an external charger to something around 50%?

ahblair4 commented 11 months ago

If you look at SBS tab you should see the battery reporting the manufacturer as SONY. If it does then it would mean that the programming was done correctly.

Interesting, it does report correctly as Sony ERA-XXX. Is it possible that my battery voltage is too low to report on the charge station screen? My confusion stems from being able to run my ERS-7 on the station with my old battery that dies as soon as it's pulled off the charger, but doing the same with this pack only results in death tone.

EDIT: After checking the SBS tab, the pack reads "Sony" as the manufacturer, which indicates success. It should be noted that this is only after loading firmware into RAM as the PermFail blocks NVM flashing. The pack eventually took a charge.