Front Camera Fault Errors - Issue fixed! DIY How-to

[dtbaker61]

A similar issue started out intermittent with my 2021 and it has been constant for thousands of miles. I get pre-collission assist not available and front camera fault service required. Cruise control doesn't work either.

They told me the module is failing and there might be an issue with the wiring. I can't tell if they're bullshitting me on the module or not. I haven't tried to take anything apart.

it's the connectors themselves. very delicate, and if you had your windsheild replaced, or the glue re-done with the recall, that probably started the problem.

just jam some foam between camera and connector to keep it from vibrating.... works 99.9 percent of the time for me. only blinks if I hit a rough bridge seam at 80mph

Sponsored

 

[ERIC8585]

it's the connectors themselves. very delicate, and if you had your windsheild replaced, or the glue re-done with the recall, that probably started the problem.

just jam some foam between camera and connector to keep it from vibrating.... works 99.9 percent of the time for me. only blinks if I hit a rough bridge seam at 80mph

Cool, I'll try taking the cover off to look. Never had the windshield replaced nor the recall.

 

[dtbaker61]

Cool, I'll try taking the cover off to look. Never had the windshield replaced nor the recall.

try this:
https://www.macheforum.com/site/thr...x-w-foam-99-effective-takes-10-minutes.54268/

 

[someone22]

This might already go offtopic a bit, but more into the technical stuff. I've managed to source the connectors and cable, made myself a fitting cable but with hand-twisted connections. This did not work as GMSL2 (which I assume is the protocol) with PoC works at over 2GHz, not something you can twist or crimp in your home shop.
Studying GMSL2 and PoC circuitry, it seems that at least there should be some voltage to be seen with a basic scope. I also live in an area where automotive is the bread-and-butter, and managed to get one senior engineer to explain what should happen He argues that there should be a constant DC offset even with a non-functioning camera. I did not see this in my tests; I've already ordered a new factory cable which should arrive in 3-4 weeks.

As an explanation, to the best of my knowledge: the IPMA and front camera communicate over a single coax. Since power is also applied, it's called PoC. The protocol is likely GMSL2 (but it could be 1 or 2). The IPMA provides power to the camera while also measuring the cable impedance/shorts/opens. Once the IPMA can read the serial of the camera (the camera is paired), communication continues. Since there is a single coax cable, the IPMA applies some power (voltage) to the cable and waits for the camera response, I assume either 5V or 12V. The data "rides" over the DC offset. There are some power management ICs that deal with PoC and can sense the trasmission line and decide whether to go ahead or not. I have only seen datasheets from TI, but haven't opened the IPMA to see which chips they actually use.

To add another layer, even if high-speed Fakra is used, it seems there are different plugs jacks and crimping standards. I have an "identical" 4-coax connector but the center pin is slightly recessed, which means connection is unreliable. Speaking to GMSL / camera devs, the RF integrity is very important, so only the Ford-specced cable should be used. The physical differences are invisible to the naked eye, like 0.1mm shorter or longer or wider.

Not good news, but I wanted to somewhat summarize my two weeks of intel gathering. I spoke to a few people developing the ADAS systems, the suppliers, to people in the fabs making the cables. It's unfortunate that a cable that's made within a 0.2-3h drive has to travel 3-4 weeks to get back home.

This should be another topic, I think, but it solidifies a bit the knowledge for the ADAS camera.

 

[ERIC8585]

try this:
https://www.macheforum.com/site/thr...x-w-foam-99-effective-takes-10-minutes.54268/

Thanks, I'll take a look. The dealer seems to be trying to do the opposite of what the TSB states (check/replace wiring), which would end up costing me double if the module is fine.

 

[ERIC8585]

try this:
https://www.macheforum.com/site/thr...x-w-foam-99-effective-takes-10-minutes.54268/

Wasn't loose at the connector.

 

[sukhoi_584th]

This might already go offtopic a bit, but more into the technical stuff. I've managed to source the connectors and cable, made myself a fitting cable but with hand-twisted connections. This did not work as GMSL2 (which I assume is the protocol) with PoC works at over 2GHz, not something you can twist or crimp in your home shop.
Studying GMSL2 and PoC circuitry, it seems that at least there should be some voltage to be seen with a basic scope. I also live in an area where automotive is the bread-and-butter, and managed to get one senior engineer to explain what should happen He argues that there should be a constant DC offset even with a non-functioning camera. I did not see this in my tests; I've already ordered a new factory cable which should arrive in 3-4 weeks.

As an explanation, to the best of my knowledge: the IPMA and front camera communicate over a single coax. Since power is also applied, it's called PoC. The protocol is likely GMSL2 (but it could be 1 or 2). The IPMA provides power to the camera while also measuring the cable impedance/shorts/opens. Once the IPMA can read the serial of the camera (the camera is paired), communication continues. Since there is a single coax cable, the IPMA applies some power (voltage) to the cable and waits for the camera response, I assume either 5V or 12V. The data "rides" over the DC offset. There are some power management ICs that deal with PoC and can sense the trasmission line and decide whether to go ahead or not. I have only seen datasheets from TI, but haven't opened the IPMA to see which chips they actually use.

To add another layer, even if high-speed Fakra is used, it seems there are different plugs jacks and crimping standards. I have an "identical" 4-coax connector but the center pin is slightly recessed, which means connection is unreliable. Speaking to GMSL / camera devs, the RF integrity is very important, so only the Ford-specced cable should be used. The physical differences are invisible to the naked eye, like 0.1mm shorter or longer or wider.

Not good news, but I wanted to somewhat summarize my two weeks of intel gathering. I spoke to a few people developing the ADAS systems, the suppliers, to people in the fabs making the cables. It's unfortunate that a cable that's made within a 0.2-3h drive has to travel 3-4 weeks to get back home.

This should be another topic, I think, but it solidifies a bit the knowledge for the ADAS camera.

They couldn't just use HDMI with PoE?

 

[someone22]

Maybe HDMI is unsuitable for automotive environments, no idea. I have not try messing around more with it, but I couldn't see the DC offset at all. Wondering if resetting the IPMA through ForScan is the way to reset the errors or if the Ford software has to be used instead.

 

[ERIC8585]

They couldn't just use HDMI with PoE?

Hdmi probably costs too much anyway

Sponsored