View Full Version : Drive-by-wire, drive you crazy

Valentine One Radar Detector

03-16-2010, 06:54 AM
It looks like the problem Toyota is having with "unwanted acceleration" is related somehow to the drive-by-wire throttle control used in models like the Prius hybrid.

With drive-by-wire, there is no physical connection between the gas pedal and the engine. Sensors tell a computer how far down you've pushed the pedal, and this information is translated into engine RPMs.

If Toyota had used a cable, not only would the problem probably not have occurred - and it would be simple to diagnose and easy to fix if it did.

A throttle cable physically connects the accelerator pedal to the engine's throttle. Pushing on the gas pedal increases tension on the cable, opening the throttle. Easing off the gas decreases tension, closing the throttle.

They've been in use for about 100 years now and the technology is pretty much perfected.

It's true a throttle cable can stick, causing the engine to race and the car to accelerate, if it happens to be in gear - just as is apparently happening with the drive-by-wire system in some Toyotas. But the problem is infinitely easier to find and fix - very much unlike the Toyota drive-by-wire system.

Either the cable's sticking, or it's not.

A quick physical inspection will determine this within minutes - and the fix is as easy as greasing (or replacing) the cable.

In and out in 15 minutes.

Now, the thing with drive-by-wire is there may be nothing obviously wrong. You can't see a sticking drive-by-wire because there are no moving components to see. Just sensors and electronics. You're stuck chasing down the proverbial ghost in the machine - who may not be cooperative. The problem could be intermittent or for all practical purposes, nonexistent. There are literally millions of "affected" Toyotas in circulation with the drive-by-wire system but only a small handful (a few dozen reported incidents - at the time of this writing, at least) of actual "problem cars."

How do you deal with a problem buried deep in the software that manifests rarely and sporadically and for no apparent reason? Toyota engineers reportedly haven't been able to get the cars to do the Unwanted Acceleration Waltz under laboratory conditions that would enable them to definitively nail down the source of the trouble.

It seems to "just happen" - and the why is currently unknown.

But until the why is known, the problem can't be isolated, let alone fixed. Toyota engineers are left looking clueless while Toyota owners get to enjoy the relaxing sensation of driving around in a car with a possible mind of its own that may suddenly decide that more speed is needed - perhaps at the worst possible moment, such as when a group of elementary school kids is walking across the road in front of you.

All because a simple, proven, effective means of controlling the engine - the throttle cable - was ditched in favor of a much more elaborate, computer-controlled means of doing exactly the same thing.

The question arises - why? Why replace something that works perfectly well with something that (apparently) doesn't work nearly as well - and which absolutely adds to the complexity (and thus, cost) of a new car?

No one seems to know.

I've heard that it may have something to do with emissions control, because drive-by-wire is both more precise, as well as easier to tie into the computer brain that runs the whole car, than our old friend the throttle cable. It may also be easier to calibrate during assembly at the factory. With plug-in electronic components, it's easier to assure that every single car coming off the line is set up exactly the same way. With an old- school cable, there's more leeway and maybe a need for minor adjustments. One car's throttle "feel" may be slightly different than another's.

But we're talking "improvements" (with drive-by-wire) that are likely so minor that the average driver would never notice them. Back in the '90s or the '80s or even the '70s, complaints about throttle cables were rare and no one (that I recall) ever seemed to notice that (for instance) one '89 Mustang's throttle tip-in was just slightly different than the next one on the lot's.

But we can't leave well enough alone, can we? We just have to have the latest gadget/technology, simply because it's possible. Nonexistent problems increasingly require elaborate - and expensive - "solutions" that no sane person would want, if given the choice.

This business may not kill Toyota but it's going to hurt it, badly. The automaker's main sales draw for years has been the safety and reliability of its vehicles and now that's out the window.

All because a perfectly functional component got tossed in favor of another bit of over-the-top technology that easily could have been done without.

04-14-2010, 02:43 PM
:) If you think Toyota has problems with their cars, Just think how it would have been if GM had gone through with their Technology that was purposed back in the 80's and 90's. I had an open classroom invitation to the local GM District Training Center in Tualatin Oregon. We are all familiar with the DIS (Distributor Less Ignition) System, which has worked out pretty Good. EFI (Electronic Fuel Injection), no major problems, so far that I have heard of. But how many have heard of a SLS, of a CSL, ESS, of a EBS? These were some of GM's ideas Twenty, Thirty Years ago.

CLS. NO CAMSHAFT folks. Electronic solenoids would open and closed the Valves in time with the Ignition System. The timing of the valves could be advanced or retarded as needed for acceleration or braking as needed.

SLS Starter Less System. Thats right, no Starter. The computer would know what Cylinder was next to fire, Would then open an intake valve, a squirt of Fuel would be injected into the firing chamber, close the valve, then the air fuel mix would be ignited by a spark from the spark plug. Thereby creating enough expansion in the cylinder to push the piston down for the next cylinder in line to fire. (I had a '77 Ford 300 ci, 6 banger that would do this every so often).

The next two, to me, would have been the scariest of all. ESS and the EBS.
Electronic Steering System, and Electronic Braking System. Just the thought of have nothing but a bunch of wires between my foot and the brakes is enough to scare the stripes off a Skunk and empty its' Sack at the same time. The Electronic Steering System was to be controlled by Sensors under the Steering mount, which was attached to the Dashboard where ever the designers placed it. Stepper Motors would then would move the wheels as to which direction and how much to move. Some what as a stepper motor changes the Air Flow in a fuel Injected System now.

The Electronic Braking System would have no, and I mean NO mechanical connection between the operator and the braking system, other than placing their foot on the pedal. No master cylinder, Brake Lines, Calipers, Wheel Cylinders, nothing. Sensors would be used to note how far the pedal was pressed and would apply braking to the wheels as needed. The Valves in the CLS would also be used to assist in braking. I'm sure that the other Manufactures had similar thoughts along this line as well. I'm glad they did not jump on this Band Wagon. All kinds of thoughts come to mind. I've always said that you can get a car to move, but if you can't steer or stop it, you have a major problem brewing. I can just see some Dad playing with his son with a Remote car, plane,or boat, and you're cruising through the neighborhood, and all of a sudden, you make a u-turn through somebody's yard. Do a burn out and park on some ones Porch. Explain that to a Cop, or your local Dealer. HA.:eek:

04-14-2010, 08:08 PM
I figure that DBW arrived >solely< as a cost savings.
How so?
Given that most new cars are equipped with cruise control, there was already an actuator in place that could open the throttle as needed, working in parallel with the cable.
Given that driver demand was a needed input for the ECU, and a throttle position sensor was
already developed for that purpose,
it was a simple leap (of faith in electronics) to move the throttle position sensor to the accelerator pedal, and remove the cable entirely.

I can think of at least two ways this could go wrong:
- Bugs in the ECU program, or the body computer program, or the program of whichever of the dozen plus small computers that might be present, or worse yet, bugs in the way they interact with each other.
- Connections, as in various chip sockets, connectors, wire crimps, etc., have not really kept pace with silicon, or maybe even software, in terms of reliability.
Wait; that gets worse, too.
Factory tests can detect infant mortality, but lapses in anti-static measures at the factory show up as blown transistors,. months down the road.
The tin plating used to retard corrosion of the brass in connectors can grow microscopic whiskers that eventually bridge adjacent pins.
Wait, that gets even worse. Our European friends have mandated RoHS construction for electronics, that basically forbids use of lead, that makes soldered connections secuire. So any electronics that might be sold in Europe will use mechanical connections or lead-free sorta-solder.
Which means that even if all the damn computer programs are perfect, they may produce erroneous outputs because they are seeing erroneous inputs, with various sensor connections being possibly open, or possibly shorted to each other or to something worse.
Producing computer systems that are sufficiently paranoid about their environment to have a chance of detecting a connection failure basically requires adding redundant components or redundant computers; basically the sort of thinking that begat the Space Shuttle seven identical computers doing the job of one.