It looks like just about anything with a CAN Bus (Controller Area Network) in it “has issues” with any water leak into where bits of it live.
This mechanic complains a great deal about it, and other foibles of the “modern” electronics afflicted automobile. In particular, he tells the story of a FORD F-150 that had the tail light gasket leak, corroding some electronic doo-dads, and ended up with a $5600 repair bill.
That’s substantially the same as what happened to my Mercedes ML320 CDI (for “only” $2000…) and similarly rendered it useless until repaired (thought it could drive to the shop as the engine still worked; just no gauges, turn signals, lights,…)
From my reading of this, it looks like about anything after the model year 2000 is “risky” and you need to verify No CAM to know that water will not kill your car.
The video:
A list of some of the cars with a CAM in the early years:
http://www.auterraweb.com/aboutcan.html
I have no idea how complete the list is for early years, and late years are clearly missing. A better list is clearly needed, but how to find one?
As for me, I’ve likely got “enough cars” to not need to buy one for a very long time, or maybe forever. For anything I’m likely to do off road, my 2001 Subaru looks fine. The FORD Expedition I’ve got looks to have similar CAM bus structure, so hopefully I never have ‘water issues’ with it. IF I ever do, I’ll be looking to replace it with something older and simpler. I have 3 cars that are clearly pre-CAM, one of them being straight mechanical Diesel, so I think I’m OK. But do need to start the “fixing them up” stage (that is almost entirely AC leak issues… and paint).
On Matt’s Offroad Recovery, I’ve seen Toyota Tacoma and Jeeps that were totaled after going waste deep in water (saturating some electronics somewhere in the car so they would not run). Anyone wanting a Real Durable Survival Offroad truck / Overlander will want to avoid anything new; or from the last 20 years (or accept that one deep puddle and your car bricks…)
Oh Well.
I wonder if there will develop a market for simple, non-CAM afflicted, off road vehicles “going forward”? Maybe somebody in India, China, or Russia can make a new one…
Musings from the Chiefio 
How did you transport all your cars from California to Florida?
Nothing wrong with CAN buss which is merely RS485 with a specific protocol. Only 4 wires – positive, ground and A and B data lines.
The problem appears to be with cheaping out on the environmental protection of connectors and the little microcontrollers at each node.
I can’t see how using the old multi (many) wire looms is going to be any better, just more complex, heavier, expensive and with many more points of failure.
The Model-T may have marked the apogee of automobiles. The brakes were controlled by rods. No hydraulics. {/sarc}
Hey, E.M. – I pointed you to that Jeep that had none of those issues.
Hmmmm… wonder if it has sold, yet. [H.R. takes off to check]
Okay. I’m back.
That 1955 Jeep CJ 5 is still showing on the Whitmore Motor’s website. $8,500 and just shy of 51,000 miles. All the mechanicals are working, so they say.
Two were already in Florida (used when I would fly in for a contract job, usually a year to year & 1/2 long at a time). I drove one out to California (net -1 , total 1 in Florida) prepping for the house closing.
The one driven to Cal. was trailered back to Florida, towed by the ML 320 CDI (that I drove, back when the trailer lights worked before the SAM/CAN died; when I still planned to use it as tow vehicle to move my stuff). This was when we sold the house and loaded up all the “must travel with me, not put into storage” into the 3 cars. F.F. drove the other one. Net 3 added, total 4 in Florida.
One was driven from California to Florida by my Florida Friend after we flew in to California to load up a truck of stored stuff after we bought a house here many months later; as a way for him to get home. I drove the big UHaul truck to Florida. Net +1, total 5.
For the 2nd UHaul truck, we (FF & I) drove out, loaded the truck, and picked up one from my Mechanic and trailered it behind said Uhaul truck. He drove the other one back. Net 1+ to Florida. Total 6.
One, driven to California in the line above, had been bought in Florida and drove to California, then back, so net 1+ via purchase, total of 7 but it didn’t really need to be moved from California to Florida, other than I needed to run out there. Net +1, total 7. This was the replacement tow vehicle (when the trailer got stolen making it moot). It was too heavy to trailer back, and the UHaul was a better choice for the tow back.
So, over about a year, at various times I needed to be in Florida or California, I’d drive or fly as needed and generally try to arrange to drive a car from California back. Some times F.F. was helping close /move the California house, and would split off to visit friends and family with one car, then drive back. 2 were towed on UHaul trailers, as the opportunity presented.
I hope I got all that right… I wasn’t planning on remembering the details ;-)
You can get a lot moved with a couple of drivers over a year+ when you need to be on both coasts from time to time. Just needs 2 going in one car one way, splitting up on the way back; or flying in and driving back. Each of the “drive back” events also usually had all available space in the cars / trucks full of “stuff”.
So now, finally, all my cars are in Florida. Though I have 2 “tow vehicles” I don’t need as they were only bought to move my stuff, but with the cargo trailer stolen, I had to resort to UHaul trucks anyway. (UHaul trailers smaller than trucks by a lot, and the ML crapped out so one UHaul truck was before the FORD. Then after the stolen trailer I just gave up and got ‘er done with another large truck instead of a couple of FORD runs with rental trailers…)
Yeah, long annoying tale of woe and “adjusting the plan” as events arrived…
@Jim2:
No water pump nor oil pump either. Water via convection (tank in roof, tall radiator) and oil via crank splashing).
It’s been all down hill from there as ever more unreliable parts were added ;-)
@H.R.:
Unfortunately, I’m over supplied with cars by at least 2 (maybe 3 or 4) at the moment. So unless they would like a mostly working Mercedes ML in trade…
BTW, were I to get another car / truck, it would likely be a Jeep Cherokee XL from when they were highly reliable and capable off road. They now command a premium of a few thousand $$$ over others… but with 3 “trucks” already and 4 “all wheel / 4 wheel drive” vehicles, it’s hard to justify.
Actually, that’s a model A that had the rods. Apparently the T has a cable that actuates a brake in the transmission ???
@Eyire:
I agree with your assessment that it isn’t the CAN Bus per se, but the environmental protection. I’ll disagree on the notion that it is going to be just as bad for the older wiring systems (having lived through it twice with CAN Bus AND dozens of old cars without it).
Why?
As the guy in the video pointed out for the FORD F-150 tail light, and as I experienced in my Mercedes ML 320 CDI (twice):
Old simplex wiring:
Tail light fails. Clean connector and replace bulb. About $5
New Multiplex Wiring:
Whole car ends up non-functional from “damage referred from leak to other parts”. Cost $4000 to $5600 (and may exceed Blue Book on the car causing economic end of life).
Watch the video embedded above and / or read the link to my story about the Mercedes from Hell for more detail.
Essentially, the system is the exact opposite of “robust” to water intrusion. It is highly fragile, prone to catastrophic failures, and incredibly expensive to fix.
(IF they can fix it at all – my Mercedes, after two replacements of the Hind Brain SAM, still can not light up the trailer brake lights. The mechanic has done software upgrades, resets, etc. etc. And no, there is no way you can find a broken relay or anything as it is ALL in the Controller Module they call a SAM)
Now I can see a theoretical way to fix this. Pot all the Controller Modules in epoxy (with appropriate heat extraction designed in) and place them where water intrusion doesn’t damage connectors (Mercedes puts a whole pile of fusses near them with MORE, not less, potential corrosion points – something like 5 or 6 fuse boxes on that thing – that is NOT an improvement). So for both the FORD and Mercedes do NOT put them behind the tail light with failure prone gaskets and / or cracks from minor near trivial accidental impacts. Also, don’t route the sunroof drain through the same space, as the Mercedes does, for when it leaks, your car ends up incapacitated with a $2000+ repair bill instead of a $10 hose replacement…
THAT is the problem. What had been, and ought to be now, under $50 for sure and often under $20 fixes of VERY COMMON failures of rubber goods, becomes a car that does not run at all or is illegal to drive (I lost most lights, gauges, and signals) and has repair costs in the $THOUSANDS.
On Matt’s Off Road Recovery, he has pulled several Toyota Tacoma trucks and Jeeps out of water crossing where they got water to the floor of the interior and it KILLED THE TRUCK. Totaled. $10K+ kind of costs. That ought to be a simple “dry it out and replace fluids” instead of an Economic Kill from the cost of all the electronics. ( I know what it ought to cost as a friend put his jeep entirely under water once and I got to watch him fix it…)
Getting your truck wet ought not total it.
Now, yes, that is not the fault of the CAN Bus. It is the fault of the designers who put the CAN Bus parts where water can kill them, cause them to malfunction and spread that kill to other parts of the CAN Bus system of overpriced electronics and render a car a total loss from having a tail light gasket leak….
But I can entirely avoid it simply by not buying any car with a CAN Bus in it. So the presence of a CAN Bus is a very good indicator of cars that “die when damp” and cars that have “mystery symptoms for years that can not be diagnosed”.
My first hind brain failure took over 6 months to figure out… by 3 professional Mercedes Mechanics including the dealership. Only finally being figured out when I moved to Florida and the persistent rains killed it entirely. Then the replacement died and only then did they find the tail light lens gasket problem. Cost about $3000 to $4000 all told for 2 hind brains, tail light gasket, several diagnostic attempts, and some gratuitous repairs to things that were not broken but the Hind Brain said they were… not to mention all the “pain and suffering” to the owner. For what ought to have been a corroded tail light fixture and bulb, and the gasket found when it was replaced – as in my several older cars over many decades…
So I would add that in addition to more “environmental protection” (including full immersion…) they need better software to prevent what I would call “referred damage” and better diagnostics to tell you what is actually wrong when one of the controllers is telling lies from failure.
But that doesn’t exist, so simple avoidance of the whole issue is the best “fix”.
Do Note: Auto companies have NO incentive to provide that better system The present failure modes give them:
1) A lot more maintenance $Money in the repair shop.
2) A lot more cars sold from trivial water damage and other failures. (A minor quartering dent that can be ignored now totals the car…)
3) A lot more cars sold from excess Cost To Repair obsolescence. Note in the video the car that needed ALL ignition coils replaced for $600 of parts instead of just one when a single coil failed. Just because the CAM detected the one new one was slightly different resistance. Note my ML where IF I ever want to Try Again to get trailer brake lights (all the other trailer lights work right, and it wasn’t the trailer) I get to start with a $1000+ part and several more hours of “programming” by the tech… and what happens when that CAM module is no longer available from the manufacturer?
All to “save” a couple of pounds of copper wires? Stupid choice. Unless you are the dealer raking in the money…
So yes, theoretically you are right. In actual practice it is a money sucking hell hole when anything goes wrong or gets damaged.
Example 2018 FORD pickup failed tail light from water intrusion: $5600
The first 5 minutes is the story, then he goes into how he found everything and what all the fix involved.
It also outlines how a water leak becomes increasingly confusing as the CAM bus modules become confused, and now instead of a simple light bulb connector cleaning becomes complicated network connectors made of Unobtaneiam replacement.
You can avoid all that by buying an older simple pickup.
Well EM, all I can say is you obviously haven’t read about the CAN Bus system. Here’s the “facts”
CAN bus standard is commonly used in all vehicles due to its key benefits, such as:
Robustness
CAN bus standard is ideal for safety applications such as vehicles due to its durability and reliability. There are also 5 mechanisms to detect errors in the CAN protocol such as bit stuffing, bit monitoring, frame check, acknowledgment check, and cyclic redundancy check.
Low-cost
When the CAN protocol was created, its purpose was to enable fast communication between electronic devices and modules, while reducing errors, weight, wiring, and costs.
https://www.autopi.io/blog/can-bus-explained/
Yup, Jim2, I’ve not read about the “facts”, I’ve only lived the reality… My Bad. /snark;
Quoting EM:
“Now, yes, that is not the fault of the CAN Bus. It is the fault of the designers who put the CAN Bus parts where water can kill them”
Water (and the subsequent corrosion of electrical contacts) has always been a problem. Adding electronics and feature creep means adding more possible points of failure. But it is a bit unfair to pick out Canbus as the sole problem.
“But I can entirely avoid it simply by not buying any car with a CAN Bus in it.”
Those cars are aging out. Surely there are some manufacturers who can do proper designs? Or at least some better than others?
Note that NMEA2000 is based on Canbus and it is widely used in boats and is doing well.
Also, water by itself isn’t necessarily bad for electronics. Not including salt water. Or Coke, or other liquids. Or probably any “water” that gets splashed up from roads.
Interesting. The last car I bought was an older but still newer than 2000 PT Cruiser. My sister was driving it up Double Oak Mountain (near Birmingham) and it died on her. Talk about frightening with highway speed traffic on one side and the ‘cliff side’ of the road cut on the other. She demanded that we ditch that one. We tried to sell it to a dealer and they wouldn’t take it. We had to let the lien holder repossess it after making the deal on the car she bought. Yes, it was a new one and we’ve nearly paid its note off. We do not plan on ever buying a new car. We will renovate, to the extent we can, our collection of older ones when this one dies. It has the newer collection of electronic bells and whistles and doing maintenance on it ourselves is a pain; but my nephew is a pretty good auto guy and I’m not totally clueless, either. We’re well inland, but we do get substantial rainfall at times and have to drive through wet roads. No issues with the electronics to date; but we do know they’ll eventually fail.
One could frame the CAN bus problem as poor execution. Cheap electronics, cables, and connectors; used to save money. Whether execution or design, I don’t want a car that uses it unless there is some info that it is not only electronically robust, but also physically robust.
@Jim2:
I wonder if there will develop an “after market” for wire harnesses, connectors, and CAN Bus parts (SAMS, controllers, etc.) that are “water hardened”…
EMS – it would be good if such an aftermarket developed. It would probably take a good bit of knowledge, some of which might be proprietary, to replicate the functionality.
Alternatively, I wonder if it would be possible to encapsulate the existing modules with a non-corrosive, waterproofing agent. Silicone or polyurethane. Silicone is typically easier to remove if needed than PU.
Sell it while it works!