Toyota accelerations revisited—hanging by a (tin) whisker

1/10/2012 4:56 PM EST

Faults like tin whiskers are exactly the reason why the pedal sensor has two pots. The values of the two pots are compared against an expected profile and if there is any discrepancy, the sensor is deemed to be faulty and power is limited to "limp home" mode.

By the way, more modern sensors use hall effect sensing which is not prone to mechanical damage like pot wipers.

1/11/2012 5:15 AM EST

It's becoming increasingly obvious that the RoHS push for lead-free solder and platings was extremely poorly thought out.

As well as the tin whisker problem, lead-free solder requires higher temperature soldering (putting more stress on components), and solder joints are much more likely to suffer mechanical fatigue.

Just another example of environmentalists creating huge problems in the "real world".

1/11/2012 6:39 AM EST

Please don't forget that apart from needing almost twice the energy in production (I thought carbon emissions were to be avoided) lead free PCB's generally don't survive rework/repair due to copper migration from the PCB surface to the tin. This results in more scrap/waste. Also I have read that lead doesn't leach out of solder in land fill (one of the excuses for RoHS) and there has been no noticeable increase in lead levels of people performing materials recovery (the other excuse for RoHS) That said, many of the other substances banned under RoHS legislation are a problem and the world is better without them, eg. Chromium VI and Cadmium.

1/11/2012 7:32 AM EST

No doubt Prof. Gilbert was correct and tin whiskers could be one of the causes for rare cases of sudden acceleration.
http://www.thetruthaboutcars.com/2010/02/dr-gilbert-explains-his-research-into-toyota-electronics/
excerpt
"What I have done is, I have shown that in the fault detection strategy of the Toyota systems, there is a window of opportunity where [an error] could occur and not be detected."

also see discussions in:
http://townhall-talk.edmunds.com/WebX/.f20003b/2011

1/13/2012 8:38 AM EST

Lead-free solder enforcement is a very poor thought decision. It should be limited to low reliability short-life toys.
The reasons are: the reliability and the fact that lead going properly to landfill is not a big problem. The harm is in compounds created, when incinerated.
So, with the toys likely to be thrown to furnace, the lead elimination is OK.
But not at all with properly recycled automotive electronics.
cheers
tomkawal

2/15/2012 2:06 PM EST

So just exactly how much lead are these regulations actually keeping out of landfills? The only numbers that I have heard implied that there were about six pounds of lead in each PC that was scrapped. That sort of damaged the credibility of the claimers, at least as far as my belief in their veracity. I would ask the same questions about lead leaching out of landfills from electronics. My guess is that we see a lot more lead from demolished building materials that contained lead paint, which was in a more soluble form to start with. Besides all of that, putting a deposit on electronics the same as on deposit bottles, would have been a far less damaging way to keep them out of landfills. A $50 deposit on every cell phone and Ipod and similar devices would be much better for keeping that half gram of lead from reaching anyplace except a recycler. Am I the only one who has thought up this idea? Yes, of course the logistics might be a bother, but we could keep a higher level of reliability in all our electronic toys.

3/13/2012 4:31 PM EDT

Forget scapegoats like "tin whiskers" and "RoHS"

A similar thing occurs with plain old copper circuit board etching, debris under conformal coatings, and etc. because the 'tin whisker" is NOT the problem, stray CURRENTS are.

I've etched PC boards with similar copper failures from weak etchant, improper PC board layout, improper etchant flow..., although those are MUCH easier to spot with a plain old magnifying glass.

This is about several things:

1.) making things TOO SMALL... the 1,9 mm length of a whisker is irrelevant when:

a.) the gap between conductors is 7 mm instead of 2.

b.) when circuit impedances are driven too high, as I assume in this case. Similarly to the argument of "physical dimension' is "electrical dimension" or circuit impedance. Sure its fun to make a high impedance, electrically efficient circuit, but when that leads to circuit impedances so high that a few loose pC of charge wrecks a car, we gotta take a few steps back from this nonsense of "make it smaller and consume less current."

Bob Pease wrote in ED Mag, mid 1990s, about an op amp circuit that went into un-explained oscillations because of stray charges? A Teflon part was generating static charges of pC that were wrecking the op amp circuits input bias.

With lower impedance circuits, such stray currents become irrelevant.

3/13/2012 4:32 PM EDT

part Deux...



What difference does it make if a sensor consumes 3 mA instead of 3 uA in an electrical system passing 100A?

2.) Separation between system engineering and manufacturing. Electronics manufacturing is part science and part black magic, especially when dimensions get really small, the present topic presented as proof. This isn't the first time an engineer has been burned by not understanding whats INSIDE components.

Case in point, a circuit failure due to 200 pF stray capacitance in a relay. The stray path was from the COIL to CONTACTS...200 pF = dead short at RF frequencies caused by the potting compound.

3.) RoHS is a convenient scape goat. If memory serves from the early 1990s, whiskers aren't limited to Sn. Doesn't this occur as a result of charge transport?

4.) who designed such a CRITICAL control system without a fail safe in acceleration rate? Not "pedal system" but in "overall vehicle accel rate."

No dS/dT limits in the fuel system? No rate discriminator to say that "no one could have depressed the pedal THAT quickly?"

Kickdown switch, anyone?

3/13/2012 5:22 PM EDT

..and WHY is DoT contacting NASA? This is too easy for any good electronics manufacturing company to diagnose.

Maybe its because we lack those in the US?

These "whiskers" look more like **solder fragments**, especially as I recall photos from the original IPC 'tin whisker' research. Solder fragments often have extremely sharp and straight edges. This is separate from "solder balls" which is another prevalent failure in high density electronic boards.

Sorry, I dont have all day to tear this "report apart" but I dont see any credibility, especially after 20+ years working with circuit board assemblies, this one blows it for me - I dont see any tests showing the "whisker" is "tin."

See p.12 - the careful qualification, by testing, of "2 uM of pure tin"

Wheres the metallurgical testing of the WHISKER, which is the ENTIRE POINT? The SINGLE most critical factor has been omitted.

When the initial premise is "this is a tin whisker" the FIRST thing to ask is "how do we know its tin?"

It is not acceptable to assume the whisker is tin simply BECAUSE THE COATING IS TIN.

Those are STAMPED TERMINALS with RAZOR SHARP EDGES, that are later COATED with Sn. The coating doesnt always adhere to the surface. Those "whiskers" look more, in my experience, like ridges of tin coating that have broken off and fallen across traces.

P. 8:

"Although the vehicle would operate, we didnt consider it to be driveable"

Not credible, wheres SAE? SAE/NHTSA/FMVSS regs.... contains all the expertise needed to deal with this question, and a real definition of "driveablity"

Contradiction on p. 16. Tin whiskers ARE NOT "conductive debris". They are generated from different sources. I cited above where sharp, narrow shards of tin come from.

Photo is not of a tin whisker, its a piece of debris that fallen across the traces and broken.

This looks like a presentation by a grade school class, and is far beneath what I expect from NASA.

3/13/2012 5:26 PM EDT

more..

The original tin whisker research was by/thru IPC back in the late 80s, early 90s, and revolved around GROWTH of whiskers across microscopic traces etched in ICs (integrated circuits)

They are not visible except by SEM, the supposed "whiskers" are, if one knows WHERE to look for them at, and the edges of stamped IC pins and connector terminals are the first place to look.

Tin whiskers GREW as a result of current flow (charge movement) and chemisty, and as I recall, were not nicely defined sharp, straight edges, they grew somewhat as crystals (think "snowflake" as opposed to "edge of a ruler")

As I recall, they grew under ACTIVE conditions (circuit operation) and not "being in storage" where air currents can blow metal fragments around.

Further, as I recall, real whiskers did not start INSIDE the edge, but grew FROM the edge, the photos above, then, do not show 'whiskers" but "fragments" that have fallen across traces.

This is old news to electronics manufacturing.

3/14/2012 11:20 AM EDT

Go back to using lead!

For those of you suggesting how dangerous lead is to the environment: Really? How many people should die as a result of electronic failures due to the lack of lead? How many have already died? The original reason given to pull lead from electronics was the suggestion that lead leaches out of electronics in landfills and gets into the ground water. More recent testing indicates that was wrong (even in areas with the worst acid rain on the planet). Therefore, as long as you don't eat it, lead is not nearly the problem in electronic assemblies that most folks were led to believe.

Go back to using lead and require all electronics to be recycled - oh yeah, that's right, even with lead free electronics; we are still required to recycle.