Hyundai IoniQ 2020 HV battery pack

[cgalpin]

Hello Royboy, any luck by checking my EB11 wiring sheet?

I am unable to get any life out of mine either.

At least looking at the Ioniq 5 "BMU extension connector" external wires, my pin 13 is brown, and my 11 is brown and 10 is yellow (10 and 11 twisted together). According to the Hyundai docs, the BMU has 4 connectors, two with 24 pins, two with 20.

the pinout (Kona connector number EB11) is:

Code: Select all

           |---------------|
/---------------------------------------\
| 11 10 |   9  8  *  *  5  4  3 |  2  1 |
|       |  --------    -------- |       |
|  * 21 |   *  *  *  *  *  * 14 | 13 12 |
|       |  --------    -------- |       |
| 33  * |  31 30 29 28 27 26  * |  *  * |
\---------------------------------------/

EDIT: ASCII art seems to only work in some browsers, so here's a screenshot of the ASCII art!
ascii_art_plug.png

(This might be mirrored left-to-right for you, I wrote it down based on the harness connector. My car is still mostly intact so I haven't verified it.)

Here are my (unverified) notes on the external connector pinout:

• 1 & 2 - Memory Power 12V (powered all times, 15A BMS fuse)
• 12 - "IG3" relay power 12V (on when vehicle is "on", 10A fuse)
• 10 & 11 - P-CAN High & Low (powertrain CAN, main control channel for BMS)

Inside the battery pack, most of these connections run directly from this external connector to the main BMU module connector.

BTW, if you're in EU or some other places then you can pay 15 Euro to Hyundai Global Service Way then you can access the exact schematic diagrams, connector pinouts, etc for the Ioniq including schematics for inside the battery pack. The above is from notes I took while looking there.

Looking from the outside, at the back of the BMS extension connector, the pins are 180 degrees opposite your diagram for the Ioniq 5. At least I think so anyway. I connected the pins in the list above, as well as all black wires to ground (also tied the battery chassis to battery ground), and tried different bus speeds, with and without termination, but don't get CAN traffic. I also tried all the other twisted pairs of wires coming out the connector just in case too.

I am in the US and signed up for https://www.hyundaitechinfo.com/default.aspx ($40 for a week) and cannot find a schematic or pinout for this connector ( they refer to it as the "BMU extension connector" ), only the connectors on the internal BMS. My next step is to open the battery and trace those out the the external connector.

[projectgus]

Interesting, cgalpin. The Ioniq 5 has the 600V traction pack and related HV systems, doesn't it? So it's possible it doesn't share a legacy with the Kona, earlier Ioniq, etc. Although at the same time I'd be surprised if it's completely different.

For the Kona pack, if all the 12V is attached and the CAN is terminated properly then I think it starts sending messages pretty much immediately.

[maciek16c]

cgalpin, does it draw any current from 12V? What CAN interface do you use? Ioniq 5 uses CAN-FD so frames could be sent but treated as errors and not shown

[cgalpin]

Ugh, I am not getting notifications even thought I am subscribed to this thread.

@projectgus this is a 800v architecture, yes although I believe the nominal voltage is considerably lower. The battery I just opened has 30 modules, which i believe means its a 72.6kWh, and is currently at 647.0V.

@maciek16c I will try measure if it's drawing any current tonight. I am using a n EVTV CANDue with SavvyCAN and was primarily using it's regular CAN, but did try the CAN-FD connections as well, although not as thoroughly. What speeds do you suggest trying?

I have pulled the cover and am tracing the wires from the outer connector to the BMU, and the pins 10&11 in the schematic above go to G-CAN high/low. The hyundai tech site pinouts for the BMU show a G-CAN and an M-CAN - i am not sure which is which.

Here is the relevent section for the pinouts. My plan is to verify pins 1,23, on connector C for +12v, the right CAN wires, and also the grounds (although @vin says it uses chassis ground)

Battery Control System (Standard Type) - BMU pinout.pdf

(193.01 KiB) Downloaded 115 times

[projectgus]

@projectgus this is a 800v architecture, yes although I believe the nominal voltage is considerably lower. The battery I just opened has 30 modules, which i believe means its a 72.6kWh, and is currently at 647.0V.

Right, lotta volts! I think there is a decent chance that it was a fresh design rather than an iteration of what's in the earlier <100S systems. The BMU pinout looks similar in some ways (like the RXNH/RXPH naming for the CMU connections, IG3 for the "vehicle on" power), and different in others.

Are you seeing 12V on both the "IG3 Power" and "Battery power" pins of the BMU connector C?

I don't know what "SDU_ Wake Up" on the same connector means, and if that's an input or an output, but maybe that's a wakeup signal for the BMU as well?

[maciek16c]

SDC could be solar dcdc converter. I think 500k / 2000k is often used.

[cgalpin]

cgalpin, does it draw any current from 12V? What CAN interface do you use? Ioniq 5 uses CAN-FD so frames could be sent but treated as errors and not shown

Does 0.14A sound like enough of a draw to power the BMU?

I am going to double check everything tonight, and see if i can trace more wires, but last night i traced the two canbus pairs (M-CAN and G-CAN) as well as the three power wires and two of the grounds. They were what i had originally other than an extra black wire i had been grounding which i could not trace and the additional canbus pair. No better results though.

SDC could be solar dcdc converter. I think 500k / 2000k is often used.

no solar panels on these cars, and could not trace that pin to the external connector so don't think thats it. But it does feel like it needs something else . I did try hooking up th two battery power wires first, then the IG3 power later, but no difference.

[maciek16c]

0.14 A sounds like BMU is doing something. Are you sure your can interface is working correctly? If it has two buses you can do loopback test to verify it. You can also measure ac voltage on M-CAN and G-CAN, if it sends anything you will see ~0.3 - 0.6V. Maybe some CAN frame is needed, for example ignition on 0x523 [0x60, 0, 0x60, 0, 0, 0, 0, 0]

[cgalpin]

I will look into a loopback test, and try send a message. I am not sure if my can testing setup is working correctly. I will test something I have used it on before successfully to make sure.

I should have mentioned that I did do a resistance check on the to CAN busses.

G-CAN H/L is 123ohm with nothing connected so that tells me the BMS is terminated and I need 120ohm on the candue side (which I have and verified it brings the overall resistance down to 60ohm)

M-CAN H/L is 7.43kOhm with nothing connected which i don't know what that means

Thanks a lot for the help

[projectgus]

M-CAN H/L is 7.43kOhm with nothing connected which i don't know what that means

Probably means there's no termination on the BMU node of the bus, so you should terminate with 60 ohms on your CANdue node.

[cgalpin]

Thanks. I believe i have a problem with my canbus board, and am trying to verify/fix that before trying again. I am hoping thats the problem and not me

[elco]

Thanks projectgus,

My next step is to find the CAN messages the operates the contractors in the contractor box, and I'm guessing that there are specific ID from the inverter which initiates that. I've had a look at the GitHub, but not sure if the info in there will help me to achieve that, will need to figure out to use Torque first lol. But the Input is much appreciated as always.

Hi,

I have also a kona 64 kWh battery pack, I am trying to see if I get the contactors to work via BMS of the battery pack. (followed the EB11 pinout)
- Is their someone that managed to get the positive/negative and precharge contactors to work via the original BMS?
I am wondering if the BMS just needs to be in satisfied operation mode to automatically get these contactors to open (makes most sence to me)
Or if you really need to send command towards the BMS to get the contactors to open.

I checked with carscanner pro, I did not find BMS DTC but can be that Carscanner pro cannot find it etc.

Very interrested to hear about your experiences,

Elco

[cgalpin]

I got a new can board and verified it's working on a known setup - a gen1 nissan leaf pack I have.

But still no luck getting any messages from the ioniq 5 bms. I have traced all the wires on connector C of my BMU other than the "SDU_ Wake Up", looped all the interlocks, and tried both CAN pairs with and without termination, but nothing. The best I have gotten is seeing car scanner send messages to try figure out how to connect to it (without any luck).

@vin are you still out there?

[maciek16c]

@elco I have Hyundai Santa-Fe PHEV battery working on zombieverter VCU.

[Bill]

Hello everyone
I'm working on ionic 5 battery too. I could get the pen out for the BMS connector. I've done the connections to a diagzone connector. Pin 1 power. Pin 12 IGN power. 10 and 11 canbus. And i used chases ground. Using diagzone on Android i couldn't detect the bms. I've attached the pinout and the canbus diagram.
Ive used 120ohm on the canbus as well. But it didn't work.
Cheers

[projectgus]

@johu/other mods - is it possible to move the Ioniq 5 posts out into another thread, please? It seems increasingly like it's going to be a totally different control scheme, and it's going to be easier for everyone to follow if the discussions sit in separate threads.

[projectgus]

@elco was private messaging me about the Kona battery pack and brought up the crash SRS signal (pin 13, white and black.) I'd been ignoring this signal but it seems likely that it might need to be valid in order for the BMS to close contactors.

(@maciek16c I don't know if the Santa Fe PHEV has this same input? If you got yours working without it then maybe it's not needed after all.)

I have a SRSCM (also known as ACU - Airbag Control Unit) here so I powered it up on the bench with nothing else connected and looked at the "Crash Signal" output:

It's 50Hz 80% duty cycle PWM signal. The output from the SRSCM is open drain and needs to be pulled up at the receiver (I expect the battery pack will pull this up, but it might not as this signal is also wired to the BCM.) EDIT: On my Kona pack this signal is pulled up to ~12V at the battery end.

80% may be a "crash" duty cycle, as the ECU has nothing else connected to it in this test. However, digging around a bit more I found this Korean patent issued to GIT Auto (who supply Hyundai) which includes a diagram showing a high duty cycle 50Hz PWM signal (could be 80%) that switches to a 20% PWM signal for 1 second in order to signal a crash:

(Google Translate of the diagram.)

So 80% might mean "no crash". Or the BMS might be happy to not see any signal here...

Have also posted the CAN log for the SRSCM bench test.

[maciek16c]

Santa Fe has the same SRS input on pin 13. It wasn't needed to close contactors but maybe bigger (full EV) battery needs it?

[EV_Builder]

Becarefull with conclusions it might fail after time too.
"fooling" a system with unknown data is tricky in addition of the fact that you don't know the implementation of the sw / hw.

Saying that it's good to read knowledgeable people are joining efforts.

[MarekN]

Thanks. I believe i have a problem with my canbus board, and am trying to verify/fix that before trying again. I am hoping thats the problem and not me

I'm wondering if you have managed to solve the problems and established CAN communication or have some other findings?

I have also a kia/hyundai 77,44 kWh battery pack. I faced same issues as described in this thread.

I found that the wire cord socket (BF-11 also called as "BMU extension connector") descibes the10,11 pins as G-CAN FD (HIGH) and G-CAN FD (LOW) respectively. I suspect that indeed CAN-FD capable device may be need for BMS communication. I have read that "newer vehicles use a CAN-FD communication protocol". In addition I watched few viedeos of the PowerBASTRO where person has used G-SCAN with "Hyundai / Kia ACU Adapter 029 - G1ZDDPA029" and "Hyundai / Kia CFCI Module - G1DDDME001" to connect to Hyundai IONIQ 5 / KIA EV6 batteries to sucessfully communicate with BMS. Unfortunatelly I have neither CAN-FD device nor mentioned adapters that I could use for testing.

[cgalpin]

Unfortunately I have not. I have a Candue board which is supposed to support CAN-FD and that canbus port works on regular canbus testing with a leaf battery, but I am nt sure if I am using it correctly for CAN-FD use. I will research.

Life has gotten in the way, but my plan was to plug in as much of the other components like the motor, ICCU, charge port etc (and getting them at least 12v power) in the hopes that being connected to them might wake the battery up.

[maciek16c]

@cgalpin There could be problem with Savvycan CanFd support. It has to be enabled in connection window but it's only shown when supported canfd interface is selected (using qtserialbus so peak can and few others):
https://github.com/collin80/SavvyCAN/bl ... 19-L401C19

I have Canable v2 and savvycan sees only frames <= 8 Bytes long because it isn't possible to enable canfd in connection window.
Cangaroo shows all frames correctly
I used Toyota Yaris IV iBooster for testing

[cgalpin]

Yes I believe you are correct. In fact when I got my new CANDue I did notice a difference in serial log output using the CAN-FD connectors and Collin even explained it could be due to CAN-FD messages but I guess I didn't understand . What I now believe was happening was it was seeing CAN-FD messages but it can't handle them and they never made it to the UI!

I am still not clear on the status of CAN-FD support in SavvyCAN but i'll try find out.

But i'll try the canable as well. The Canable v2 appears to be sold out at openlightlabs.com so I guess I'll take my chances on getting a knockoff from aliexpress

Thanks a lot!

[maciek16c]

SavvyCAN partially supports can-fd but there are some functions missing and when i try using frame analyzer it crashes. I had to load log from other application to see 32 byte frames. Canable works for me, but when using slcan firmware timestamps in savvycan are missing.

[cgalpin]

Just noticed this. Sorry @projectgus for the continued derailment of this thread. I could start a new one for the Ioniq 5 battery and continue there if you'd like, but can't move the existing messages over myself.

@johu/other mods - is it possible to move the Ioniq 5 posts out into another thread, please? It seems increasingly like it's going to be a totally different control scheme, and it's going to be easier for everyone to follow if the discussions sit in separate threads.