I'm new here, so I apologize in advance if I'm not clear. I'll do my best to clarify your questions as soon as possible.

For a professional project, my team and I purchased the rear seats of a 2014 Mercedes-Benz S-Class S550 Long (W222) to study the ventilation system. Here are all the informations I got about them: the fans are made by ebm-papst, the connector to the wire harness has a brown wire and a yellow/red wire, the brown wire spliting into 2 white wires and the yellow/red wire spliting into two white/red wires for the two fans. The part ID from Mercedes is A099 906 51 01, and from ebmpapst it's 9971705083. I would like to understand how they work in practice and whether it is possible to power them independently to equip them in a seat in another car. This is what one of the two fans looks like once removed from its plastic housing in the original Mercedes seat:



The part ID are this time: Mercedes: A222 838 01 00, ebmpapst: 6038197003.

I assumed that the white wire is the GND and that the red/white wire is the wire to power the fan, just like a regular PWM-based DC fan.

Here is also a photo of the motor and its electronic board once dismantled:



Given the electronics, the structure of the motor, and the plastic with a magnet in which it sits, I assume that it is a BLDC motor and that it was custom-made, as I cannot find an equivalent with this size and bearing on the ebm-papst website.

Since there are only two wires, communication via LIN is ruled out (or maybe it is DC-LIN?), so my first instinct was to power them with my stabilized laboratory power supply. I first tried applying a stable 12V DC power supply (standard in automobiles), but no relevant consumption was recorded on the ammeter (1 mA), and the fan did not spin. I also tested with a PWM (12V, 100Hz, sweep between 20% and 80% duty cycle), but still nothing. There was a slight variation in intensity between 1mA and 4mA depending on the duty cycle, but the fan did not spin. The department responsible for purchasing the seats had the seller certify that the fans and the seat electronics in general were functional.

NOTE: we sadly do NOT have the calculator neither a complete S-Class to make measurements on an oscilloscope for example, only the seat and its harness. If you don't know what to do with your W222 S-class, measurements are greatly appreciated lol.

So I set about reverse engineering the board that controls the fan, and here is the electronic diagram I managed to obtain (note: The resistance values are measured directly on the board without desoldering the component itself, as my hot air desoldering station is out of order at the time of writing this post. Notable voltage values have also been added, the rest being either unreadable with the multimeter or 0V):



I couldn't find the exact reference associated with the “828SAP” marking, but I spotted the Microchip logo, so I initially assumed it was a BLDC motor driver. However, after analyzing it, I couldn't find any Microchip drivers with the SOIC-8 package, with Vcc opposite GND and which uses all its pins (most SOIC-8 devices and drivers I found have N.C. pins). so I assumed it was a microcontroller to manage the signals needed for the fan to work properly. That's why, on the diagram, I assumed that “828SAP” was an ATTiny402 (corresponding package and pin position/usage, very popular in the automotive industry), but it's possible that this isn't the case.

EDIT: After further checks on this forum, it would appear that Mercedes-Benz/Daimler group very oftenly uses PIC12 or PIC16, and the “E/MS” marking on the component strongly suggests to me that it is a PIC12LF1572-E/MS.

I'm running out of ideas. All components, such as the Microchip chip and the Hall effect sensor, receive voltage for their operation and when a magnet is brought close to the Hall effect sensor, it reacts correctly. The system just seems to be “asleep.” Either my electronic circuit is completely wrong and I've been on the wrong track from the start, or the microcontroller is waiting for a specific option or signal to wake up. I don't know how to run these fans. Any help, analysis of the diagram or even lucky guesses from experience would be appreciated. Thank you in advance!

Note: the disassembled fan was used only for reverse engineering the board; the tests were performed on one of the other fans that was assembled and functional.

 

First Post...WOW...that's all I have to say.

 

Welcome .
I feel like in a physics class , WOW
I like to see tech things as well

 

Ha! Well said..
I assume the you are a former Marine..
I’m a former Submariner from the late ‘60’s
This guy just left a Jarhead and Bubblehead in the dust!
He lost me at “professional project”
Be well SemperFi
SSBN629

 

In addition to 1:1 cars my other interest is slot cars, digital slot car racing in particular where they use PWM motor control, in addition to other functions. The PCB in the housing is a basic decoder; what you are missing is the encoder. You can build a basic board using an Arduino or Raspberry Pi for your encoder.

 

How would you go about that? I have an Arduino Uno but I'm sorry I don't exactly get what needs to be "encoded" or "decoded" in a PWM? Is it some form of Power Line Communication where +12V is merged with the speed at which the fan should spin perhaps and that's what needs to be encoded?

 

Good Question. It's kind of based on the microchip. Is it the VC-828 or something else? You said it's a E/S 828SAP but I can't reference that chip. You might have to buy a power seat module to figure it out.

 

Yeah, manufacturers don't like people like me looking inside their electronics, so it's just some marking for temperature operability and year/week of making. Based on the context, the Microchip symbol and the SOIC-8 format, I strongly believe it's a PIC12F1572, marking E/MS are matching, chip format too, quite used in automotive applications for stepper motors and blowers, so I think that's the one, but the firmware is (of course) locked by code protection...
I would love to buy a seat module to get to the bottom of this but the department responsible for what's bought is not a fan of buying something just to probe it two times and never use it again... Maybe someone could tell what are the signals or just the wiring diagram to know what type of signal to send and explore from there.

 

Maybe one of the gusy who do code on cars could help. @BenzNinja is the most active. Hopefully he can help. But your question is not one we run into here all that often. I saw you posted the question on the EE forums, lets hope someone there worked on a similar project.

 

Made some progress on the fan, if you wanna check it out it's here.

TL;DR: the fan needs DC-LIN, it spins then stops, this cycle repeats forever, and I can't tell how to send properly the data or if the program I wrote actually sends the data correctly.

 

You're talking above my knowledge level at this point. Looks like fun stuff.
Going back to your original post where you said "....I would like to understand how they work in practice and whether it is possible to power them independently to equip them in a seat in another car...." Would it be easier to find a small dc motor to fit in the same housing that can power the fans? Referencing the part number, it loks like that's an axial flux motor, those are still pretty pricey.

I found a datasheet for a similar fan from EBT on Mouser and Grainger, that might tell you what you need to know: https://www.grainger.com/product/5AG...waAlzmEALw_wcB

https://www.mouser.com/ProductDetail...8aAog0EALw_wcB

 

Fun it is indeed! I was so excited when the fan started to spin, one step at the time.

Regarding alternatives, the reason why we wanted to use Mercedes-Benz seat fans is because they are extremely compact (60x60x25mm) and provide sufficient airflow, they're like perfectly tailored for what we are trying to achieve.

To be fair, we planned on giving up because we thought it was proprietary protocol given the encrypted chip name, and go with the alternative here. But now that I managed to make the fan spin, we are doubling down on the orignal Mercedes-Benz fans idea more than ever, It spins, not correctly, not continously, but it spins and that's enough for us to keep seeking how to control those fans.

Note: looking at the price of the alternative, I doubt I'll be able to convince the purchasing department to spend $701.2 ($175.30 x 4) to replicate the Mercedes-Benz seat fan configuration. If a seat control module that's like $120 second-hand was already asking for the moon, I let you guess how it will go...

 

Quick update regarding the fans: they now operate continuously and predictably! A huge thank you to @Waw642 for his assistance, for his research in one of his post, and for his guidance.

I've updated the post on the EE forum here. TL;DR: the fans use the SENT protocol, or at least something similar to SENT, not DC-LIN or PWM. There are similarities but also differences compared to the standard protocol...

The best thing would be to be able to send it a PWM command to set its speed rather than being limited to the three speeds on the Mercedes control panel, but I think the quantization is hard-coded into the microcontroller. I'll keep digging into it.

 

Reviving this thread just in case someone could help...

To get you all up to speed where I'm at so far:
The seat unit (Seat Assy A222 920 65 21, ZGS 02) I was working on at the start of the thread was the "older" one, and on that one I successfully drove both Backrest (A099 906 51 01, contains 2 fans wired in parallel) and Cushion (A099 906 53 01, contains 4 fans wired in parallel) using a 10-bit frame (Sync High + Direction + Speed) sent 3 times then hold High. Both units responded identically. The PHY layer was single-wire, Active High, Pulse-Width Encoded (Logic 0 ≈ 900µs, Logic 1 ≈ 1800µs), similar to SENT.

To get a second pair of seat cushion and backrest ventilation motors, I had to order them genuine from Mercedes as the previous seller I got the old seat unit sold the other one. Upon recieving them, the part numbers were different even though I gave the same VIN for the order: WDDUG8CB4EA048908. Now those are:
- Cushion: A222 906 36 01 (Contains 4 fans wired in parallel)
- Backrest: A099 906 82 00 (Contains 2 fans wired in parallel)
No change in the topology and both are still controlled by 2 wires on the harness, so naively I thought "they'll be the same, why would they change anything to an already robust and complicated command for such a small and simple feature right?"...

How wrong I was :')

The color on the wiring and the manufacturer of the ventilation motors are different.
On the "old" version, both the seat cushion and the seat backrest ventilation motors were made by ebm-papst, on the harness there's a Mercedes P/N and an ebm-papst P/N, and the colors on the wire were yellow with red trace and brown.
On the newer version, the seat cushion ventilation motors are still made by ebm-papst, there's still a Mercedes P/N and an ebm-papst P/N and the wiring color is still the same, yellow with red trace and brown, BUT on the seat backrest, the ventilation motors are made by Y.S. Tech, P/N FD126025HB-D(2I7E), and on the wiring harness there's a Mercedes P/N and an Elektrosil P/N (EM-30.8404 17/06/25 A1497). The wiring colors are also different, this time it's a yellow wire and a brown wire. You can see both of them attached down below:




My guess is that when I ordered the parts genuine from Mercedes, they gave me the "post facelift" version of the components, and they changed the way the waveform is sent. The electrical topology probably did not change, but the firmware inside the seat control unit running those ventilation motors must've changed.

On those "newer" unit (the one I'm currently characterizing), I painfully successfully reverse-engineered the Cushion driver using a 13-bit frame: [Header 3b] [Speed 4b] [Echo 4b] [Tail 2b]. However, while the bottom 2 fans run perfectly, the top 2 fans exhibit visible RPM lag/stalling at low speeds despite receiving the same command (since they are wired in parallel). I suspect the ZGS 08 protocol might support a "Split Zone" or "Boost" header that I haven't fully identified to balance the load.

Regarding the Backrest, this unit is completely unresponsive to the Cushion protocol (13-bit). It appears to ignore the standard header entirely.

Something worth adding is that when I try to plug both the seat cushion and the seat backrest to the same command (so the command is going on the yellow with red trace and the yellow wire, both brown wires are tied to the same common GND), the seat backrest immediately collapses the command line down to 0V with random spikes every 3ms that you can see below:



Unplugging the seat backrest ventilation motors from the data line and the seat cushion ventilation motors happily start again to respond to the command and spin.

If anyone have a >2018 W222 or a C217 and a logic analyzer or oscilloscope, could you please provide a trace or a capture of the "Wake-Up" / "Init" sequence from a working ZGS 08 module? That would be so awesome!

I am trying to determine if the Backrest requires a distinct Device ID / Header (e.g., addressing Zone C/D) or a different initialization bitmask compared to the Cushion.

It's been a month I've been trying anything to make those new ventilation motors actuate, so any kind of clue or new eyeview are welcome! If you need more information feel free to ask me directly.

Thanks.

 

Have you looked at the later model Mercedes Cabriolets with the "Air Scarf" fans & heaters installed in the headrests? Not sure if the fans they use are thin enough for what you are trying to achieve.

 

Yeah I took a look at them and they look interesting indeed. The issue now is that since I managed to make the "ZGS 02" ventilation motors work and presented it with my team to the client, he now wants to get the genuine and recent ones from Mercedes. However when you request the seat cushion and seat ventilation backrest ventilation motors genuine from Mercedes, they send you the "ZGS 08" version, and I'm back square one

At this point either I take the "easy" route and convince the rest of the team that I can't guess anymore what is the exact frame needed, tell them I can provide it for the ZGS 02 version, build the signal converter that would take a simple 12V 100Hz PWM like any other kind of fan in the automotive world and actuate accordingly the Mercedes ventilation motors, but not for the ZGS 08 ones. Or I take the "harder" route and keep looking anywhere I could (and I mean REALLY anywhere, even on Russian/German forums or China retrofitters lol). I've been at it for more than 2 months and still have not a single valuable result on the backrest ventilation motors, and the seat cushion ventilation motors actuate but there's still that issue of RPM lag on the upper parts and some of them sometimes don't actuate at all...

But I still have hope, maybe that waveform I'm looking for is somewhere on the Internet, maybe someone out there is tackling the same issue to build a signal converter from that proprietary waveform to a more "normal" one to control any kind of fan from the OEM door card commands and we'll connect, maybe I'll have a lucky guess and that would unlock the rest of the path forward, maybe... but for now I'm in a dead end. I won't give up though, call that being stubborn but I'll keep searching and get to the bottom of this that's for sure, if you bang your head against the wall enough the wall might end up breaking before your head haha (I'm turning clinically insane please send help).