Car Wash Caused Hybrid System Warning

[rachaelseven]

Originally Posted by JasonPro

I've been through a touchless wash several times now without an issue. I think you're on to something suspecting the fuse box getting wet.

I hope that's all it was. Honestly, I was quite shocked how much water was able to get under the hood and into the fuse box. I know the hood isn't fully sealed, but I didn't expect that much to get under there.

[hobbit]

Did the carwash include stuff to spray upward against the bottom
of the car, hard enough to get up on top of the transaxle? Water
could easily land on things like the MG temp sensor connectors on
the transaxle, but everything underhood is *supposed* to be sealed
with gaskets around the connector bodies and where the wires come in.
.
The MG1 [aka "generator"] resolver and temp-sensor connector appears
to be the same grey one I mention as useful to disconnect in the
block-heater page, right on top of the housing. [Oddly, the resolver
and temp-sensor connectors for MG2 are separated front and back, go
figure..]
.
I spray out my underhood area every spring to get rid of salt and
crud accumulation, mostly avoiding the visible connectors when doing
so, and that makes it easy for me to generally keep my engine
compartment cleaner than just about any other owner's I've seen.
But after I'm done I sponge out areas like the top of the engine
where water tends to accumulate in the head-cover [gen2] around the
ignitors and various connectors and anywhere else I find pooling or
remaining water that I can reach, and take the car for a good highway
run afterward so the heat and airflow dries things out the rest of
the way.
.
_H*

[jayvee]

In this case, the "high" appears to mean a "high" logic level, not high temperature. It means the "circuit" level is high, or a short to B+ (+5V of the ECU).

Some attachments from the following TIS document:

HYBRID / BATTERY CONTROL: HYBRID CONTROL SYSTEM: P0A2C-247,P0A2D-249: Drive Motor "A" Temperature Sensor Circuit Low (2010 Prius)

What's a bit weird is that water going into the connector should only be able to bring the level low, not high, according to the circuit drawing in that document. One possibility is the water spray loosened (even temporarily) the connector or the pin at the connector, briefly making it an open connection. That would do it.

[rachaelseven]

Did the carwash include stuff to spray upward against the bottom
of the car, hard enough to get up on top of the transaxle? Water
could easily land on things like the MG temp sensor connectors on
the transaxle, but everything underhood is *supposed* to be sealed
with gaskets around the connector bodies and where the wires come in.

The carwash did indeed have some upward sprays designed to hit the rockers and tires. Doesn't seem likely they'd get up past all the under-panels with enough force to reach the top of the transaxle though. From what I could see, the entry path was the sides of the hood and/or the cowl area. If it was the resolver and temp sensor, that would tend to indicate a problem, since as you pointed out, those connectors are supposed to be sealed and a leak there would be likely to cause a problem down the line.

In this case, the "high" appears to mean a "high" logic level, not high temperature. It means the "circuit" level is high, or a short to B+ (+5V of the ECU).

Excellent research, Jay, thank you. I did see the 249 information code, so that jibes with what you put up there. I didn't follow the full diagnostic procedure, but it appears that I would have dead-ended at 'intermittent problem', based on the information I had. Open seems pretty unlikely, though - water spray breaking the connection is a long shot, if you'll forgive me for saying so. More likely, we'd be looking at the 'short to +B' option, which appears to trigger all the same responses, as far as I can tell from the diagnostic procedure.

[jayvee]

Originally Posted by rachaelseven

Open seems pretty unlikely, though - water spray breaking the connection is a long shot, if you'll forgive me for saying so. More likely, we'd be looking at the 'short to +B' option, which appears to trigger all the same responses, as far as I can tell from the diagnostic procedure.

Yes, that's what I thought, too, until I looked at the circuit more closely (attached in last post.) (OK, this maybe is more technical than we need to get, but.....) The only way that the circuit can "short to B+" (5V) is if the ground path is removed, which means that the circuit path through the sensor itself is gone (open, or at a very very high resistance). Water can only short it to ground, not short it to 5V, since if the water is touching "ground" (which it would have been), the bottom of the pull-up resistor on the ECU side ("MMT") would have been at ground potential, not 5V. But if the line is open, there is no current flow through the pull-up resistor, making the MMT line at 5V as well, which is the input to the ECU. And you would have to have that line at 5V to get the 249 code.

I agree that it's hard to imagine - but I've seen water and vibration do some very weird stuff to connections - and all it would have taken is a millisecond, since these signals are frequently read thousands of times a second by the processor. I've seen stranger. If it happens again, you might want to check that connector for integrity, including the wire-to-pin connection.

I've had mine through a similar wash (with undercarriage spray) a few times, with no problems......yet......Hopefully you won't see it again.

[rachaelseven]

Originally Posted by jayvee

Yes, that's what I thought, too, until I looked at the circuit more closely (attached in last post.) (OK, this maybe is more technical than we need to get, but.....)...

Almost certainly more technical than we *needed* to get, but who says we can't have a little geek fun

Your analysis of the circuit seems right - I'll admit a bit of weakness there, as I'm a mechanical engineer and prefer to deal with macroscopic things rather than electrons - but nevertheless, your logic appears sound from what I recall of my electronics classes. So I guess the question is then, if we want to , is there another way that signal could have been thrown? Noise on the bus maybe that just happened to trigger that particular error by chance? The only water incursion point outside the ordinary was the fuse box, so I'm really looking for the explanation that comes from my (stupidly) forgetting to close that up. Otherwise, I'll have to lie awake at night in fear of carwashes

[Mike Dimmick]

Originally Posted by jayvee

In this case, the "high" appears to mean a "high" logic level, not high temperature. It means the "circuit" level is high, or a short to B+ (+5V of the ECU).

Some attachments from the following TIS document:

HYBRID / BATTERY CONTROL: HYBRID CONTROL SYSTEM: P0A2C-247,P0A2D-249: Drive Motor "A" Temperature Sensor Circuit Low (2010 Prius)

I was looking at the Gen 2 documentation for this code yesterday and I think they do mean a short to aux battery voltage - in step 2 of the diagnosis procedure, if the voltage seen at the sensor end of the wire harness is between 9 and 14V, the manual tells you to 'check for +B short'. The ECU does supply 5V to the sensor, and the sensor has a separate ground return wire, rather than using chassis ground as many other sensors and actuators do.

[jayvee]

Originally Posted by Mike Dimmick

I was looking at the Gen 2 documentation for this code yesterday and I think they do mean a short to aux battery voltage - in step 2 of the diagnosis procedure, if the voltage seen at the sensor end of the wire harness is between 9 and 14V, the manual tells you to 'check for +B short'. The ECU does supply 5V to the sensor, and the sensor has a separate ground return wire, rather than using chassis ground as many other sensors and actuators do.

Thanks Mike,

(I promise, this is my last "horse beating" )

I agree, they probably mean a short to 12V; but in either condition for a 249 code (short to B+/12V, or open), the voltage at the MMT line would be at 5V or higher. But what would be the 12V path for this, in the case of a water short? If 12V was really being applied to the MMT line by way of a water short, it seems it would have to be at the ECU connections (not at the sensor), otherwise the chassis that the water was in contact with near the sensor would be energized to 12V! If water were present to short from some 12V line at the ECU, I would think a lot worse things would be happening and a lot more codes generated. The only other possibility I could see is a harness short if it is bundled with other wires and there were two breaches in the insulation (one for a 12V line, one for the MMT line).

The A21 connector at the ECU is inside the vehicle, on the right side of the instrument panel, obviously nowhere near the fuse box. So I don't think water intrusion at the ECU connector is the explanation. Noise on the line? Certainly is a possibility - those car washes probably generate lots of EMI. And most of the wire harnesses are not shielded (including this one).

I still think the most likely scenario is a short duration "open" on this line. But hey, I've certainly been wrong before!!

Rachael, you've discovered an interesting problem, to say the least!

[rachaelseven]

Originally Posted by jayvee

Rachael, you've discovered an interesting problem, to say the least!

Well I sure didn't mean to! And when it was pointed out to me that I'd recently been in the fuse box, I nearly aborted the post without even putting it up. But I figured maybe we could use a good mystery

So what if the chassis actually was energized to 12V just for a nanosecond? What if the water in the fuse box, where there is obviously plenty of 12V including the jumper cable point, connected 12V to the frame... not for long, a millisecond or less. The heat instantly vaporizes the water, breaking the short. But in that instant, the ground reference is gone or there is the effect of a 12V noise spike on ground. Maybe?

[maledyris]

Originally Posted by rachaelseven

It's the only tested and confirmed option. The $300 ISO/CAN version of the Mongoose cable (also made by Drew Tech and available on eBay) is a very high probability of success (Techstream recognizes it's drivers, at least), but without either testing one or pressing Drew for assurances, we won't know for sure.

I've been trying to figure this out too. I don't think it will work. The problem is that the $300 cable lacks the ability to adjust voltages on specific pins. This is required for reprogramming as far as I understand (documentation is tricky to come by). I read the j2534 spec and it's pretty straightforward. Originally I thought I would buy a reasonably priced cable and make the dll interface myself. Sadly, I no longer think this is possible with a cable that lacks these capabilities...