My Toyota Prius 2010 caught fire

Now here's an interesting thing I had never stumbled on until about ten minutes ago...

It seems there was a TSB out for 2010 and 2011, issue was low-beam headlights (both sides) burning out on much shorter intervals than expected. T-SB-0204-12, 3prongpaul linked to a copy here.

The gist of the TSB is to add to the length of the original headlight harness by splicing in what looks like a meter or so of extra wire, so that there will be more voltage drop along the wire and a lower operating voltage at the bulb.

It can be inferred that the original problem was too heavy a choice of wire for the headlamp harness, resulting in too little voltage drop, and the bulbs running above their nominal 12 V spec and somewhere closer to the car's power supply voltage of 13.8 to 14. It might sound funny to call that a problem (in fact, I've seen posts where modders deliberately make the wire heavier than stock to boost the voltage at the bulb), but physics comes into play and says light output increases as voltage to about the third power (exactly the second power for total output, but more of the output shifts into the visible, hence the exponent of 3.1 to 3.4 on visible light output) ... but at the same time longevity goes down as voltage to about the thirteenth power! (Source: Don Klipstein at the heading "Why making bulbs last longer...").

So running a 12 volt bulb at 14 may give you 65% more light (woo hoo!) but shortens the bulb life by more than a factor of seven! so the engineers aren't really being idiots with their skinny wire ... they're trying to hit the sweet spot where the voltage that reaches the bulb is enough for adequate light without eating bulbs crazy fast.

It seems like Toyota, for 2010 and 2011, missed the sweet spot, and missed it on the high side, resulting in complaints about how quickly the bulbs burned out. And to correct it, they issued a TSB that requires a tech to pull off the bumper, disassemble, cut, and splice the headlight wiring on both sides, making a loop down in the bumper area to use up the added wire length, zip tie it safely into place, and put everything back together.

If we knew which of the right-fender-area fires happened on vehicles with or without that TSB applied, I wonder if there would be a correlation. I also wonder which way it would go ... maybe more of the un-TSB'd cars because of the lights running hotter? Or more of the TSB'd ones because of workmanship issues in how carefully the technician did the splicing and repair?

-Chap

 

Good detective work the Gen3 folks are lucky to get a new member

 

That would be a plausible theory, except that it conflicts with most of the previous explanations I've seen, including several posts in the thread you linked. According to previously prevailing wisdom, the voltage was too low with the original 2010-11 skinny-wire harness, making it impossible for the halogen in the halogen bulb to do its thing returning escaped filament molecules to the filament. That's obviously opposite to what would happen with a conventional non-halogen incandescent light operated at low voltage.

I don't know whether my Prius, bought used at 41,129 miles, has had the TSB modification, or whether it still has the original bulbs. I've had no headlight problems in what I'd estimate to be 150 hours of headlight-on time so far. I turn them off manually before shutdown.

 

Chapman, nice theory alright, but some of the fires happened in cars sitting hours after driving.

BTW, my prius had no TSB for bulbs and still OEM bulbs at over 70,000 miles.

 

I did read those several posts in that thread and the prevailing wisdom about too skinny and too low. Then I read the post #43 by Dark_matter_doesnt, who had actually read the TSB to see that the instructions were to add extra length in series with the original wire, therefore further dropping voltage, not raising it. Then to double-check, I read the TSB myself and found it to say just what Dark_matter_doesnt said it said. Then so you could double-check if you wanted, I linked to the TSB from my post here above.

-Chap

Well, so I'm not claiming to have a definitive explanation, and I'm not sure I've seen anything else yet in this thread that would qualify ... the thread seems more at the stage of development where there are some indications there might be a problem, and some room for collecting information that might be relevant, of which this seemed like it could be one more piece. A service bulletin that would have caused certain 2010s, but not others, to have wiring cut and spliced by technicians at the typical dealer range of skill levels, near what seems to be the locations of several of the fire origins, seemed like maybe it shouldn't go unmentioned in this thread.

Can it directly explain incidents that seemed to begin in a parked car with the lights presumably off? Probably not. Will this thread ever contain a theory that explains every reported incident? My hopes are less ambitious. But at this stage it seems ok to continue adding relevant observations to the thread. Without looking more closely at the harness routing, I don't know whether the work on this particular circuit would be near enough to other wiring, hoses, etc., to play some part in a cascading failure; not really an uncommon way for things to fail.

-Chap

 

Yes, I wanted to double-check, even though dealing with balky googledocs is a tedious pain. When I eventually succeeded in rounding up that TSB from the link ... Wow, you and Dark_matter are absolutely correct, and the conventional wisdom about the original wiring resistance being too high apparently is a lotta boloney! It never completely made sense to me, but, not being an expert on the chemistry behind halogen lighting, I grudgingly swallowed it. We still have people in PriusChat propagating that dubious theory, even advocating adding new heavier wiring and relays to minimize voltage drop to the headlights.

I wonder what Toyota implies by the word "some" in that TSB's introductory "Some 2010 – 2011 model year Prius vehicles may exhibit a condition ... ." Some owners didn't complain, or didn't use headlights enough to expose the condition during the warranty period? Or only some 2010-2011 cars had lower than intended wiring resistance, meaning larger than specified wire gauge? Or some inverters expose the lights to higher voltage than others?

Using a few wild-guess approximations, if we want to drop 0.5 volts at 5 amperes, that requires a resistance of 0.5/5=0.1 ohms. That equals the resistance of about 3 feet of AWG#25 (~0.45mm diameter) copper wire, depending on temperature. 0.5V × 5A= 2.5W dissapated by that 3 feet of added wire. It shouldn't become hot enough to cause a fire, assuming appropriate wiring insulation, reasonable ventilation, and good crimp splices. With sloppy high-resistance splices ...

An easier, safer, roughly equivalent solution might be to use "long-life" bulbs and leave original wiring alone.

 

"Some" is terminology used to minimize customer worry and complaints. It's a way to address a problem without revealing the extent.

Concerning the retrofit wiring on "some" Prii, it's still possible to botch the job so that only a few strands of copper carry the current, and leads to localized heating at that weak point.

I don't think there is an unusual amount of Prius fires compared to any other model, but I have no data to support my anecdotal opinion. I'm certainly not loosing sleep over it; not to minimize the loss some have experienced due to fire.

 

After reading the TS-B, my opinion is it lacks information as to the cause and resolve. Taking existing smaller gauge wire in a harness and splicing in larger gauge 3' before the load makes no sense for reducing the resistance as it would be minimal and likely offset with the splice. The only logic I see is adding a service loop or strain relief to lengthen the wires in the existing routing to prevent them from chaffing or being banjo'ed. Re-pinning the wires to the connector could also eliminate "suspect" bad crimps. Sounds like the Lawyers buried this one fast for those affected what ever the case.

 

short of measuring the wire resistance before and after TSB, we won't know the answer. But, it they were serious about increasing resistance, they would put a power resistor in series. GM did that as a recall in the beginning of 2000's because GM's ingenious electric engeeneers specified 12V bulbs for DRL and somehow they were not aware the car alternator outputs 13.6V.
i think this prius wiring problem was more of the intermittent contact somehow damaging the bulbs.
i'll closely examine mine to see if there are signs of problems.

 

If the added wire's material, gauge, and length are known, calculating its added resistance (and hence the added voltage drop) is easy. A separate resistor might be more cumbersome to mount safely, as well as being a more conspicuous oddity.

Voltage spikes associated with use of the automatic timed headlight shut-off feature have been rumored to be a factor.

My ScanGauge shows the "12V" system automatically jumps to about 14.5V when lights are on. I've yet to see a good explanation.

 

That is normal alternator behavior for most other vehicle types. When a load threshold is reached, such as turning on headlights, the alternator boosts output voltage. Perhaps Toyota replicated this behavior with the DC/DC converter that maintains the 12v system.

 

Agreed that it makes no sense for that purpose, but all that tells us is that that isn't the purpose. There is nothing in the TSB that says they wanted to reduce the resistance; that idea came up in various speculations by posters who might not have read it closely, but (just as you point out) it can't be supported by what the TSB says.

The most supportable inference from what the TSB actually says is, they wanted to increase the resistance ... because that is what any addition of more wire (regardless of gauge) in series with what's already there would do.

-Chap

As you mention it, I just happen to be in a good position to do that, having just taken delivery of a 2010 that is reported to have the eats_low_beams property. I'll go ahead and order that repair wire and post how it measures out.

I don't know if I can second guess them; engineering decisions are often trading off a lot of different factors. A meter of wire spreads out the dissipated heat over more area than the equivalent component resistor, for example.

Anyway, will know more when my wire comes in.

-Chap

 

Exactly right! So who gets credit for creating the internet rumor that the resistance needs to be reduced?

Yes, and won't require special mounting provisions different from other parts of the wiring harness. It's not going to be much heat, though. Measuring such small resistances accurately with an ordinary two-lead ohmmeter is tricky, as you probably know.

Please report the effective length and conductor diameter or gauge of the wire. Thanks!

 

For all I know, it could turn out to be specialty resistance wire or something.

I had a Bronco II that was built that way: the harness circuits for the taillights had sections of calibrated resistance wire, and the taillight-out warning circuit measured the voltage drop between the ends to confirm the lights worked. Just looked like wire, but if you ever made the mistake of replacing it with plain wire, the taillight-out warnings would always be on. (For once, I didn't learn that lesson by experience ... it was explained in the manual.

-Chap

 

I'm still betting on the quality of the pin/s crimp/s, wire strain (broken stands) and/or chaffing. Higher resistance causes higher heat and so does a dead short. This thread is about fires right?

 

Yes. I edited my previous post to say one needs to know the material, as well as gauge and length to compute the resistance of a lead. Incidentally, heating appliances sometimes use silicone-insulated nickel-plated iron electrical wiring in hot locations.

I'm betting Chap and dark_matter are correct. There's no other reasonable inference from the TSB (which, as others have noted, is not as easy to understand as it should be). Faulty crimps and broken strands cause unexpected heat to be concentrated in a small space, which can cause fires. Adding a lead in series causes added heat distributed over the entire length of the wire, assuming the splices are good. It's not a lot of heat anyway---likely under 5 watts.

Yes, this headlight issue should be in the Maintenance etc. department, but it's not, and it's important. In fact there has been a recent thread there with posts still espousing the now-discredited assumption that resistance must be reduced.

 

Right ... I only intended to mention the headlight TSB here in enough detail to contribute to the brainstorming about fires ... but, aside from the chance that there could be a connection, I did not mean to suggest that there surely is one. My thought is that the best way to evaluate the possibility here is just to learn about the TSB issue in a way focused on getting its details as accurate as possible, and not try to fit it or not fit it to a fire narrative, until it's first just understood on its own terms.

In the grand scheme of things, it might not even deserve as much attention in this thread as the fact that, apparently, it's been pretty common for people of widely varying skill levels to slap in random HID kits from eBay ... as far as I see in this thread, we haven't got information about how many reported fender-area fire incidents had that in their history ... or whether that happens any more often to 2010s

-Chap

Here's the repair "wire" (for one side! my local dealer wants $35 each, online it's about a third less).

It's a roughly 55 cm length of plastic wire loom that seems to have a much longer length of wire looped up and packed inside it. Both ends of the wire come out the same end of the loom. One is prefinished with a terminal for the headlight bulb, and the other end is stripped so you can crimp it onto the old end of the headlight wire (after cutting the old terminal off and stripping it).


The loom is all very tidily taped up, but open at the far end so we can peek in. It looks like the wire loops at least five times (it's white with a black stripe):


I measured its resistance a couple ways ... first directly, after zeroing the meter leads, I read 0.12 Ω. Then, thinking to get conditions closer to running a headlight, I put it on a constant current source (2 A, biggest I've got, sorry) and measured the voltage drop: 0.24 V, so that confirms the direct reading.

The spec for a standard H11 bulb seems to be 55 watts at 12 volts, so it oughtta draw about 4.6 A instead of my 2, so the voltage drop proportionally would be 0.55 volts. Apparently that's how much Toyota wanted to shave off the voltage at the bulb. (If the relationship between bulb longevity and thirteenth power of voltage holds, shaving the supply from 12.55 to 12 ought to extend the bulb life by about 80%.)

This is a big handmade power resistor for a dissipation of about 2.5 watts. I wouldn't expect that in itself to be any kind of hazard. I was thinking more about the possible opportunity for workmanship issues in the process of going into that cramped space behind the bumper and cuttin' and splicin' on stuff.

-Chap

 

The normal operating voltage of my 2011 is about 14.4v, the 12.x volts is only off the wimpy auxiliary battery.

 

Have you ever used backprobes to take the voltage at the headlight bulb under normal operating conditions?

Remember wires are governed by Ohm's law, and the ones to the headlights are skinny, and they're skinny on purpose.

-Chap

 

Chapman, great investigation! I'll look closely at my prius this weekend and will report the voltage reading at the bulb (no TSB), so we can compare with yours (after the TSB fix).