Failure analysis of Marelli Electronic Throttle.

My previous post had on error. Marelli is not out of business, they are a large comglomerate that have many trademarks under their umbrella including Weber, Solex, Midas and others.

Summary:
The failure mode for the Volvo Electronic Throttle Module, made by Marelli and used on 1999, 2000 and 2001 70 series Volvos, is the wearing out of the thick film resistor on the two throttle butterfly position potentio-meters. It is not due to heat or due to a dirty throttle body. The time to failure is due to the accumulated movement of the throttle shaft, so in town stop and go driving with a lot of throttle shaft movement per mile will wear out the potentiometer film faster than a long distance driver with few throttle movements per mile.

First, let me introduce myself. As a Metallurgical Engineer I managed a failure analysis laboratory for Rockwell International for 20 plus years. I am very familiar with component failure analysis and have done or supervised thousands of analyses. In general, electronic components seldom fail; it is the mechanical parts in the assembly, such as, switches, potentiometers, variable capacitors, connectors and solder joints that fail. There is no need to be afraid of electronics, just the mechanical components.

Failure mode:
At the ends of the throttle module are black plastic caps. In the inside center of each is a ½ inch (approx.) bushing with a “D” shaped hole that fits over the end of the throttle shaft. Of course, this only turns less than ¼ turn, from closed to full open. Alastair, in a previous post, said there is a rheostat on each end. When I popped out a nickel sized spring retainer the bushing and rheostat were fully exposed. Alastair said he found the wiper, that was riding on a black resistor material, contacted with the ends of the wiper. On close examination I find that the wiper has two sets of four fingers riding on this resistance film. Again, as Alastair said, the wiper rides on the resistance film as if you were scratching your fingernail across a black board. They use redundant fingers so that if there is dirt on the film there will be several that are still making connection. My examination, at 10x magnification, found that the wiper is covered with the resistance film, enough to bridge the space between the fingers. On looking at the inside of the rheostat, the film it is grooved. Closer examination showed that the thick film resistance material was on a piece of flexible plastic film, like photo film, but probably Mylar. It looked like it could be pulled out so I tried. It came out and measures 7/16” x 2 5/8” with 2 strips of resistance material. The wiper has carved grooves deep enough that light shines through on the film (this was on the end that the external cord is clipped to.) At the other end is similar construction, though the resistance material has not yet been scraped away, though there are clear tracks and an accumulation resistance material at the end of the track, like you had raked a garden one way only. If I was a betting man, and since I have no way to do a full electrical test, I would place a big bet that this is the problem.

I consulted with a component engineer that worked with pots, switches, and other similar components. He verified my thoughts that there is no combination of materials for resistance film and wiper material and geometry that would work in this application. A metal wiper and a carbon film is the combination that is used on low cost radio volume controls, not a wiper that is going to see up to several thousand movements per mile. (Think – how many times does the engine power and thus the throttle position change per mile.) The only design for this throttle position sensor that would have live through 100,000,000 million movements per 100k miles (assuming a movement every 5 feet) must be some type of non-contact sensor.

Why two potentiometers? See http://www.auto-solve.com/etc.htm. This discusses the design of the throttle position sensors. There are two resistance signals from these potentiometers. The combined signal allows the ECM to calculate a mean voltage output from the two signals allowing the throttle butterfly position to be calculated with greater accuracy. Since it is critical to know the throttle position very accurately any degradation of the signals from the potentiometer would throw the ECM computer and its program into spasms.

Other components:
The circuit boards are made using DuPont ceramic thick film technology with gold, silver and silver-palladium interconnect. I found a good explanation of the materials of construction for this exact board at http://www.dupont.com/mcm/applic/H-78296.html. The advantage of the ceramic is low expansion with temperature extremes. The components are potted with a clear, elastic compound, probably a silicone (?). All components are visible. All connections from the boards to the outside connections are two aluminum wires that are welded to the board and appropriate feed-thru. Between the power board and the control circuit board are 11 connections that are single wires and 3 that are doubled, probably power and ground. All components, except resistors, are soldered down using surface mount technology, resistors are screened (painted) on and then trimmed to value by laser. Most passive components (resistors and capacitors) are connected by screened on traces. The active components (transistors, diodes and integrated circuits) are connected by welded wire bonds. Based on looking at thousands of circuit boards and thin film technology devices, I would rate these as excellent and very appropriate for the application.

The fix:
Replace the throttle module at the first sign of erratic running, rough idle, car can only go 20 mph in limp home mode. There has been a lot of talk about cleaning the throttle body of carbon and dust. Don't bother. Others said that the quality of gasoline or petrol made a difference, ain't so. Alastair and others opened these end caps and used circuit cleaner and achieved a temporary fix but the problem is still there, the resistance film is damaged and will continue to get worse.

Warranty:
There are several warranty out there. My son got one from Colorado First, a company under United Car Care. It specifically mentions what IS covered and the ETM is not mentioned. Volvo VIP warranty does cover it. I did find that the dealers cost on the module is about $338 and he marks it up to $535 plus labor plus software upgrade. Again, we all feel this is a safety issue, and a time bomb failure. It is not IF but WHEN.

Other comments:
Some had said that the replacement parts are Bosch. I called Bosch and the person I talked with said that was not so. I'll look at my son's car this afternoon and see if I can determine the manufacturer.

What we can do:
Call Volvo and register a complaint that this is a safety issue caused by a defective design. Tel: 1-800-458-1552
Email: customercare@volvocars.com

Call or email National Highway Traffic Safety Administration at (www.nhtsa.gov) or call 1-888-DASH-2-DOT.

Try to bargain with your dealer to give you the parts at cost.