New Battery: Fisker Ocean 113kWh FM29AWDHVB1

[Korndi]

Hi!

I have a Fisker Ocean battery pack with
these specs (from the label):

Model: FM29AWDHVB1
Nominal Voltage: 379.4V
Nominal Capacity: 298Ah
Nominal Energy: 113.1 kWh
Weight: 588 kg
Cell chemistry: NMC (CATL)
Year: 2023, only 5.000km

Goal: Use as stationary ESS connected
to Sofar HYD 20KTL-3PH inverter.

I can provide CAN logs once I have
the LilyGo T-2CAN hardware.

Is support for this battery planned?
Happy to help with testing!

[dalathegreat]

Any battery can be added, with enough time and resources! Wiki has some info on what adding a new battery entails. https://github.com/dalathegreat/Battery-Emulator/wiki/Data-needed-for-new-battery-integration

Even just taking pictures of all connectors, LV, HV etc. is useful, so we can start to build the Wiki page for the battery!

[Korndi]

Hi!

Thank you for the quick response
and the wiki link!

Quick update on my situation:

The battery pack (FM29AWDHVB1) is
currently still at the seller's
location in Hall in Tirol, Austria.

I am currently in the process of:

1. Purchasing the battery pack
2. Organizing transport to my
   location ( Upper Austria)
3. Ordering the LilyGo T-2CAN
   hardware

Once the battery arrives at my
place, I will:

• Take detailed photos of ALL
  connectors (HV + LV + CAN)
• Record CAN data with LilyGo T-2CAN
• Post everything here

This will probably take 2-4 weeks.

Thank you for your patience -
I am very motivated to get this
working and will contribute all
data back to the project!

Best regards from Austria 🇦🇹

[Korndi]

Battery has arrived!
Here are photos of all connectors.
Still waiting for LilyGo T-2CAN
for CAN data logging."

[Korndi]

Hi! I have Fisker Ocean FM29AWDHVB1
powered with 12V KL30 + KL15.
CAN pair shows 2.5V differential.
LilyGo T-2CAN only shows TX, no RX.

Tried:

• Crystal 8 Mhz (250kbps)
• Crystal 16 Mhz (500kbps)
• All 3 twisted pairs

Is CATL NMC using different baudrate?
Any hints for Fisker/CATL CAN protocol?

[Korndi]

"Important finding!

I believe the CATL BMS requires
VCU CAN messages to start sending data.
Without VCU messages the BMS listens
but sends nothing (explains TX only, no RX)

Questions:

1. What VCU CAN ID/message does
   CATL NMC typically expect?
2. Is there a 'listen only' mode
   in Battery-Emulator to capture
   BMS responses?
3. Should I try sending a generic
   VCU heartbeat message first?

Battery: Fisker Ocean FM29AWDHVB1
KL30 + KL15 confirmed working (246mA)
CAN pair: Blue/Violet showing 2.5V"

[Korndi]

"MAJOR UPDATE - Fisker Ocean working!

Hardware: LilyGo T-2CAN
Connection: CAN B (Native)
Crystal: 16 MHz (500kbps)
Battery set to: BYD Atto 3

Results:
✅ CAN RX working on CAN B!
✅ Pack voltage: 370V (correct!)
✅ SOH: 99%
✅ Capacity: 113.1kWh

CAN IDs confirmed:
0x100, 0x103, 0x110, 0x180,
0x181, 0x182, 0x310

Decoded:

• SOC = 71% (0x110 byte 7 = 0x47)
• Temp max = 34°C (0x310 byte 4)
• Temp min = 25°C (0x310 byte 5)
• Cell voltages 3.62-3.84V

Full CAN log attached!
Who can help implement
Fisker Ocean battery support?"

[Korndi]

🎉 MAJOR UPDATE - CAN Communication

finally working!

Hi everyone, big progress to report!

After weeks of trying, I finally got
full CAN communication with my
Fisker Ocean 113kWh CATL NMC pack.

---

The KEY finding that solved everything:

Battery MUST connect to CAN B
(CANHB/CANLB), NOT CAN A!

With CAN A → zero RX data
With CAN B → full BMS data! ✅

Also: Crystal must be 16MHz (= 500kbps)

---

Confirmed wiring (20-pin connector):

• KL30: Brown/Red wire (permanent 12V)
• KL15: Brown/Blue wire (246mA draw)
• GND: Black/White (chassis)
• CAN H: Blue wire → CANHB
• CAN L: Violet wire → CANLB

---

Active BMS CAN IDs confirmed:

• 0x100 → Cell voltages group 1
• 0x103 → Cell voltages group 2
• 0x110 → Cell voltages group 3 + SOC
• 0x180 → Temperatures group 1
• 0x181 → Temperatures group 2
• 0x182 → Temperatures group 3
• 0x310 → Main BMS status

---

Decoded so far:

SOC → 0x110 Byte 7
0x47 = 71% ✅

Cell voltages → 0x110
Byte 1 = cell index (0x00-0x0F)
Byte 0 + 3600 = mV
Example: 0x7D+3600 = 3725mV ✅
All cells: 3620-3843mV ✅

Temperatures → 0x310
Byte 4 = max temp (0x22 = 34°C)
Byte 5 = min temp (0x19 = 25°C)

Pack voltage → 370.0V ✅
SOH → 99% ✅

---

Best protocol tested: Geely Geometry C

Results with Geely Geometry C:
✅ Pack voltage: 370.0V correct
✅ Cell voltages: 3690/3700mV correct
✅ Cell delta: 10mV correct
✅ SOH: 99%
❌ SOC: reads 0% (wrong byte)
❌ Temperature: 0°C (not decoded)

---

Important: BMS needs VCU messages!

The CATL BMS stays silent without
VCU heartbeat messages.
TX 0x12D keeps the BMS talking.
Geely Geometry C protocol sends
the right VCU messages automatically!

---

What I need help with:

1. Correct SOC byte mapping for Fisker
2. Temperature byte mapping
3. Full 104-cell voltage decoding
4. Eventually: native Fisker Ocean support

Full CAN log attached below.
Happy to test any firmware suggestions!

Pack info:

• Fisker Ocean One 2023, 5.000km
• Model: FM29AWDHVB1
• 104S CATL NMC, 113.1kWh
• Stored 1.5 years, SOH 99%

[Korndi]

canlog_01-37-35.txt

[dalathegreat]

Hi @Korndi , nice work! Please join the Discord and let's continue chatting there!

[dvbhannes]

Great work. I still have that to do with an Aiways U 5 HV battery.

[Simply-Johann]

• Ist dein log kürzer als nur 1 ms?
• Wenn TX 0x12D am laufen hält, warum ist es nicht im log?
• Warum denn so viele TX (Transmitts), wenn 12D für das "Laufen" genügend wäre?
• ID 310 scheint im byte6 einen Multiplexor zu haben (03 bis 06), aber das log ist vorne und hinten abgeschnitten, Sorry...
  Du benutzt wohl mir unbekannte Software, die "zu viel transmittet"...
  Ich würde die professionelle Hardware PCAN-USB von PEAK vorschlagen und metodisch vorzugehen.
  Auch den Kopf, direkt beim Einschalten mitloggen, wie in meinem Beispiel
  Ansonsten, Gute Arbeit!

TRC-Kopf.txt

[Korndi]

Hey Johann, danke dir fürs genaue Drüberschauen! Du hast den Nagel auf den Kopf getroffen.
Das ganze TX-Chaos in meinem Log kam daher, dass ich auf dem LilyGO das fertige Geely-Emulator-Profil von Dala am Laufen hatte. Das Ding hat quasi ein komplettes Auto simuliert und den Bus vollgespammt, weshalb es das eigentliche BMS total übertönt hat. Und ja, das Log war definitiv viel zu kurz, sorry dafür!
Einen teuren PEAK-Adapter werde ich mir für das Projekt auf der Werkbank aber erstmal nicht holen, der LilyGO muss das eigentlich auch schaffen, wenn ich ihn richtig einstelle.
Mein Plan für die nächsten Tage: Ich stelle den LilyGO auf SavvyCAN um und mache ihn zum reinen, stummen Zuhörer (Listen-only). Dann starte ich die Aufzeichnung am PC, bevor ich die 12V Zündung (Klemme 15) auf das BMS gebe. Dann sollten wir den kompletten "Kopf" vom Hochfahren absolut sauber und ohne fremdes TX-Rauschen im Kasten haben.
Ich lade das neue, saubere Logfile dann direkt hier hoch. Danke für den Tipp mit ID 310, das schaue ich mir dann im neuen Log auch gleich genauer an!