1985 VW DOKA syncro (full leaf stack, zombieverter)

[Bratitude]

in April I officially moved into a proper shop space with @quentinsgarge, whom builds and restores vw vanagons/syncros

we where looking at the engine bay of his VW DOKA truck, and it looked like a leaf stack might just fit...

-1985 VW DOKA crew cab
-syncro drive train (4x4)
-2014 leaf stack (motor, inverter, pdm)
-mg water heater(work in progress)
-brat industires(me) VW-leaf adapter kit
-zombiverter VCU
-tesla gen2 ibooster
-epas steering column
-33in tires
-big disk brake upgrades
-chademo (work in progress)

testing with an old gen2 volt battery

setting up 16 modules from an ioniq 5 for 38kWh with the oem MAX17841 boards
which fit nicely under the rear passenger seat bench.

the leaf motor is mated to the syncro gearbox with a direct connection to the inputshaft, no clutch/flywheel.

the locking diffs are 5.43
2nd gear is 1.82 (final drive 9.8826)
3rd gear is 1.125 (final drive 6.10875)

these are decent ratios for bearing speeds in the gearbox.
the truck is a very nice drive, lots of power, good braking and steering, with supple suspension.

zombieverter is running 2.30
current prams:

doka-regen-2nd-gr.json

(2.13 KiB) Downloaded 11 times

currently having issues with the pdm not requesting any charge current. the zombie goes into ac charge mode upon plug insertion of a type 1 granny cable.
yet to test with a level 2

wiring from leaf stack is: perm 12v, switched ign, CAN H/L.


can logs of plugging in granny evse:

zombie-leafstack-evse-plugin.csv

(858 Bytes) Downloaded 20 times

can logs of key on:

zombie-leafstack-key-on.csv

(857 Bytes) Downloaded 16 times

[tom91]

Not a propper log, you are overwriting ids. You need a time based log.

The 0x390 from charger shows 0x8C which is not a value I have seen yet for a plausible lead. I would expect it to be 0xn8 or 0xn4.

So if you can get me a log with time based ids then we can check before plug in and plug in.

Also what are you doing for BMS/UDC measurement? As the values look a bit weird in you log.

[danjulio]

setting up 16 modules from an ioniq 5 for 38kWh with the oem MAX17841 boards
which fit nicely under the rear passenger seat bench.

Can you say what BMS you're using with the OEM CMUs?

[Bratitude]

Not a propper log, you are overwriting ids. You need a time based log.

The 0x390 from charger shows 0x8C which is not a value I have seen yet for a plausible lead. I would expect it to be 0xn8 or 0xn4.

So if you can get me a log with time based ids then we can check before plug in and plug in.

Also what are you doing for BMS/UDC measurement? As the values look a bit weird in you log.

UDC is just from the leaf inverter. no BMS setup at the moment

whoops sorry here you go:

doka-leafstack-evse-plugin.csv

(1.34 MiB) Downloaded 20 times

doka-leafstack-keyon.csv

(856.34 KiB) Downloaded 17 times

went over to a level 2 evse and the zombie would not go into charge mode. back the shop it dose with the granny cable.

Can you say what BMS you're using with the OEM CMUs?

will be using https://wdrautomatisering.nl/shop/gatew ... 1-gateway/

[tom91]

back the shop it dose with the granny cable.

Does it charge in this situation?

Please also log this.

[Bratitude]

Does it charge in this situation?

Please also log this.

that is what i logged, on the granny cable it dose not charge, just goes into charge mode but dose not draw any current. after a bit the contactors open, and it no longer is in charge mode.
heres a proper full log of the granny being plugged in and it timing out:

doka-leafstack-evesgranny-plugin.csv

(1.64 MiB) Downloaded 19 times

i also have a few other year pdms on the shelf, which i can try.

[ianlighting]

and it looked like a leaf stack might just fit...

Nice!

Just cos it’s not 100% from the photos - were you able to keep the Leaf stack in 1 piece as it is in a Leaf? Not separated like I had to do in my T2?

Do you think that engine bay fit would be the same for all T25 (uk name) / T3 (rest of world name) models?

[tom91]

So what the log shows. As soon as the charger reports available power the commanded power drops to 0 0x64 (d100) from the Zombie.

Code: Select all

uint16_t calcBMSpwr =
        (Vbatt *
         Param::GetInt(
             Param::BMS_ChargeLim)); // BMS charge current limit but needs to be
                                     // power for most AC charger types.

    uint8_t OBCpwrSP =
        (MIN(Param::GetInt(Param::Pwrspnt), calcBMSpwr) / 100) + 0x64;

    if (opmode == MOD_CHARGE &&
        Param::GetInt(Param::Chgctrl) == ChargeControl::Enable) {
      // clamp min and max values
      if (OBCpwrSP > 0xA0) {
        OBCpwrSP = 0xA0;
      }

      // if measured vbatt is less than setpoint got to max power from web ui
      if (Vbatt < VbattSP) {
        OBCpwr = OBCpwrSP;
      }

      // decrement charger power if volt setpoint is reached
      if (Vbatt >= VbattSP) {
        OBCpwr--;
      }

      if (OBCpwrSP < 0x64) {
        OBCpwrSP = 0x64;
      }

So I would need to what your BMS_ChargeLim does and Pwrspnt during this time. Can you try bumping your Pwrspnt to something more (like 5000) just for testing? Possiblly something fall over in the code.

[tom91]

Never mind I found the reason.

Your 0x390 behaves different to any log I have seen so far.

Byte 5 goes 0x85

This is really strange, can you check your wiring of the charge port and provide more information on this.

Maybe I need to do more cleaning in the understanding of the byte 5 content.

Code: Select all

    PlugStat = bytes[5] & 0x0F;

    if (PlugStat == 0x08)
      PPStat = true; // plug inserted 32A
    else if (PlugStat == 0x04)
      PPStat = true; // plug inserted 16A
    else
      PPStat = false; // plug not inserted

    Param::SetInt(Param::PlugDet, PPStat);

[Bratitude]

thanks tom, i have a few other year pdm's on the shelf, so swaped over one labeled 292C05SA1B

there is no "time out" when the granny evse is pluged in. zombie fires up and sits in charge mode and stays there

heres a new log with the new pdm:

doke-leafstack-evsegranny-plugin-pdm-292C05SA1B.csv

(1.34 MiB) Downloaded 29 times

this seems to be a little more inline with what to expect, when comparing with the pervious log of the other pdm

when the evse is pluged in "OBC_charge_status" switch's between "finished" and "charging or interrupted"

ill pull the harness and go over it in more detail, perhaps there's a discrepancy

[tom91]

There still is wierd behaviour in the 0x390 signal. This is likely a charge port wiring issue.

The charger never starts to report an available charging power as you would expect (and is the behaviour in others running with a zombie and leaf logs)

[Bratitude]

Pulled the full harness off.

Apparently i cant even follow my on guidelines.
I had pin 19 (green) (quick charge permit signal) connected to 12v from the pdm.

https://openinverter.org/wiki/images/f/ ... ctions.png It

is active when pulled high.

With that removed, and everything hooked back up, still no charge current upon plugging in the evse. Going to take some logs tomorrow and then swap the original pdm back in.

[tom91]

Check all your wiring. Key one is grounds as the signals are oscillating where they should be stable for the Proximity and control pilot.

[Bratitude]

I noticed some other strange behaviour outside of the ev system (I did not wire the rest of the truck) and found what looks to be some ground loops, shop partner is now tearing down the truck and going through everything.

Meanwhile I will setup the ev system isolated and test from there

[Bratitude]

I spent some time and made another iteration of my pre-wired zombie interface:

With a bunch of the trucks electrical (outside of the ev system) cleaned up, still no luck on charging. Still coming down to charge statues flag on the bus.

pulled the j1772 plug and right off the bat I noticed it was one from a 2021 leaf. (This is a 2016 stack) the resistance across pp is the same… but installed a 2016 outlet…. And it works!

We are now charging!

It was the outlet used causing the error.

I’ll dig deeper to understand the differences between the two

[Bratitude]

Visual differences in the 2021 and 2016 j1772 plug

“Top” is 2021, “bottom” is 2016

Notice the difference in pp resistor plugs (which are hidden under the rubber boot normally)

[marcexec]

Different resistors in sockets? Or is that a plug?

[tom91]

pulled the j1772 plug and right off the bat I noticed it was one from a 2021 leaf. (This is a 2016 stack) the resistance across pp is the same… but installed a 2016 outlet…. And it works!

The Proximity and signal are colour switched?

[Bratitude]

truck is running and charging good! still adjusting regen and throttle setting etc, there's not enough grip with the rear tires currently.... iv have gotten lockup skidding via regen. the trucks center coupling is welded so its currently just selectable 4x4, but its getting reverted back to the viscous coupling, which has been donated from visco werks https://vw-kern.at/visco_en.php for this truck as a test for a ev drivetrains!

awd will be a game changer in having high regen ramps!

the torque derating in reverse is just to little, i have gotten the truck and have had to change gears to get it to backout of inclined spots. have not looked to deeply if this is the leafs hardcoded limits, or a result of the zombie.

just came across a good deal on a standard range+ tesla model 3 battery, part number 1104428-00-M, so it should have the "short" modules inside.

which stacked in a 17x17x55in box will fit in the gear tunnel of the truck. This will move the weight even lower than under the rear seat, and outside of the cab.

the yard let me take a log.

wrecker-m3-vc.csv

(204.18 KiB) Downloaded 3 times

wrecker-m3-cp.csv

(1.26 MiB) Downloaded 9 times

using https://github.com/joshwardell/model3dbc dbc file on both the charge port and vehicle can, i was looking for the following info from:
-0x3D2 total charge and discharge power history
-0x352 soc and nominal pack charge
-0x332 max and min cells
-0x401 cell voltages

being bricks (parallel cell groups) start at count 0, brick 96 will be blank (0v)...took me a sec to notice..

43kWh so around 82% soh roughly
cell delta of about 63mv

okay not to bad of a pack for the price!

[cajamatt]

just came across a good deal on a standard range+ tesla model 3 battery, part number 1104428-00-M, so it should have the "short" modules inside.

which stacked in a 17x17x55in box will fit in the gear tunnel of the truck. This will move the weight even lower than under the rear seat, and outside of the cab.

not a fan of the ioniq modules??

[Bratitude]

not a fan of the ioniq modules??

Using 16 ioniq modules in the Datsun 720, so only have 13 undamaged modules left.
we want 200km range out of the doka so 38kWh wont cut it.

[Scrappyjoe]

Awesome to see this brat. I'm hot on your heels with a syncro with front diff replaced with outlander RDU. A word of caution. Treat the syncro gearbox gently - it shouldn't be exposed to over 200Nm, 180Nm is better, and keep your throtramp gentle. It's a tough gearbox, but it doesn't handle more torque than that and live long, and rebuilds are crazy expensive.

[Bratitude]

Awesome to see this brat. I'm hot on your heels with a syncro with front diff replaced with outlander RDU. A word of caution. Treat the syncro gearbox gently - it shouldn't be exposed to over 200Nm, 180Nm is better, and keep your throtramp gentle. It's a tough gearbox, but it doesn't handle more torque than that and live long, and rebuilds are crazy expensive.

Cheers! I have an outlander mocked up in a front subframe too they just fit with the diff centered with out hacking anything up! my drove shaft adapters sure make the axles easy too

Quentin built the gearbox, it’s been reinforced. Driving 33 sure dosnt help in terms of being kind to it, but engine ≠ ev when it comes to gearbox health. constant pre load and smooth delivery vs rapid and frequent gear changes…