Jackass Award - GM Pickups Fuel Pump Control Module

2009 Chevy Silverado plow truck (with apologies for the CCTV-screencap image quality)

First, a bit of a back story. Electronic fuel injection was just hitting the mainstream when I began working on cars for pay (rather than to avoid paying...) just over 20 years ago. There were several different strategies involved in feeding the systems with fuel, all of which involved at least one electric pump. To prevent the pump from running unnecessarily after a stall, a crash, or just sitting with the key on, the pumps were usually switched on by either the engine computer (through a relay), an oil pressure switch, or both (early GM systems especially).

Most had a single in-tank fuel pump that delivered a constant feed of fuel at a pressure governed by a regulator (usually throttle-body or injector-rail mounted and vacuum controlled) that bled any excess fuel back to the tank through a return line.

Others - most notably the Europeans - would use a low pressure feed pump in or near the tank to supply a second, high pressure pump that ultimately accomplished the same function as the single-pump setup.

The odd vehicle had the ability to vary pump speed, usually through a dropping resistor, or via high frequency voltage toggling (known as pulse-width modulation). This was done primarily to reduce pump noise at low speed and idle, when demand was low and ambient noise less likely to disguise the ruckus.

As technology has introduced more precise injection control, a large number of vehicles have gone to what are called "returnless" systems, where there's no longer a second fuel line coming back from the engine compartment to return excess fuel supply. A single line simply supplies fuel at the required pressure to the injector rail. This eliminates some parts, but most importantly, prevents the fuel from being warmed by engine heat prior to its return to the tank, which apparently offers benefits in emission reductions and possibly even power production.

Often there's a fixed pressure regulator in the tank or a nearby filter/regulator assembly to make this single line system possible. Others vary the fuel pump's power supply to control its output and therefore pressure, usually with feedback from a fuel-rail pressure sensor. This is where our Jackass Award story begins.

An important note: GM is not alone in using the basic design I'm about to discuss. Ford trucks are well know for fuel pump control module failures, for example. But some questionable engineering choices do make the one used on GM's recent model year full-sized pickups - the example featured, the first of these I've encountered, is a 2009 model - Award-worthy.

This seemed like a good idea to someone...

The above photo shows where GM chose to locate the Silverado's (and the identical Sierra's) FPCM (Fuel Pressure Control Module), just above the spare tire beneath the box at the rear of the truck. Ford is equally guilty of mounting the fuel pressure control module where corrosion will eat it. They were doing it well before GM decided to follow suit. Hey, if it didn't work for Ford...

To diagnose this thing, most of the trouble code diagnostic "trees" require you to unplug this connector. The lid swings down to unlatch it - impossible, by about an inch, with the spare tire in place. Ever lowered the spare on a modern pickup?

Rube Goldberg would be proud

I have. Fortunately, this truck , in spite of being used for plowing snow and landscaping, is well-kept, clean, and not a big muddy ball of corrosion, so the process was only slightly aggravating: find and extract the toolkit (often buried or missing), unlock the spare tire lock (good seize-in-place potential), assemble the correct sequence of crank segments (instructions? Who needs 'em?), spend a few minutes trying to get the tool to align and function in the winch (you'd think the built-in guide would make that a first-attempt thing. It doesn't.), crank the spare down.


Finally able to disconnect the connector, we verify that the module has failed and needs to be replaced. Surprise! Our local dealer has one in stock.

On the vehicle, this view is only possible with a mirror, a boroscope, or a cameraphone jammed up against the floor of the box. Nice corrosion.

If you look closely at the spare tire-view photo, you can see that the module is fastened to the bracket it's mounted to from above. A Jackass Award qualifier. Extra Jackass points though because the fasteners in question are rounded-head Torx bolts with fine threads - seen from the top, after mild, fruitless digging out with a pick - in the above photo. The three of those are not coming out without a level of personal attention that's all but impossible in-situ.

Another cameraphone-aided view. Only my hand could see this otherwise. Thank goodness for ratcheting wrenches.

The FPCM is mounted to a large bracket that also holds the TBCM (Trailer Brake Control Module) and another small electrical component. It bolts to the left frame rail and the spare tire winch mount. Fortunately, its three fasteners have conventional hex-heads, though, in true Jackass form, they are also top-mounted, and the winch-side one is conveniently and for no apparent reason located directly beneath the pickup bed's reinforcement beam. There will be no using air tools or even a ratchet on those. Nice. (At least this particular truck was mud and corrosion free.)

Voila!

I think I may have peed...

Once extricated from its bracket, the FPCM peed on our bench. Actually, I was glad to see this, because condemning electrical components can be stressful because you usually can't see anything wrong. Water leaking out? I'm feeling pretty comfortable with my diagnosis! It may need more, but it definitely needs this.

Note that the silver side seen above is the top. The plastic tub forms the bottom, making a decent water-retaining bowl. Obviously the thin layer of sealant wasn't enough to keep the water out (or in). You surely wouldn't want to mount it the other way, and give it a fighting chance.

Carefully opened afterward with a 24 oz ball-pein hammer, water can be seen inside. Think those IC chips like water?

The new one has a thick bead of sealant oozing out of it. It got new bolts too, though they didn't come with it.

So we get our new module, mount it to the bracket, fish the bracket's top-mounted fasteners back in and ratchet-wrench them tight, successfully clear the trouble codes, crank the truck, and get...  ...a brief fire-up followed half a second later by a stall and refusal to restart.

Recheck the codes to see what else is wrong, and see..

That's not good.

... that the FPCM needs to be programmed. It's a several hundred dollar box of rocks without the proper software. The common-failing Ford truck fuel pump modules don't need this. They're even available from the aftermarket, plug'n play. Not this one; the final, Jackass Award-clinching move.

Shouldn't have been a surprise, actually, because in this generation of GM pickup, even the power window switches have to be programmed when replaced. Got an identical, same year truck, identically optioned, and want to temporarily swap the switch to put a window up when the original switch breaks on a -25ºC day? Won't work - not programmed to that truck, as one of our tow-truck fleet operators discovered, to his extreme pleasure.

Money, money, money...

At this point there are a couple of options.

- Tow the truck to a dealer (what GM would clearly like you to do, though they'd probably be happier still if their dealer had diagnosed and replaced the module in the first place).

- Tow the truck to a shop that has GM programming capability (which they have to pay to subscribe to - GM is still making a smiley face).

- Program it yourself. Assuming you have the several thousand dollars worth of equipment and/or software to do it, of course. You will also have to pay GM a subscription fee to access the download. $55US gives you 2 days of access to that content, for example, so GM still wins.

-------------------------------------------------------

For designing a vehicle that even requires a stand-alone module to operate the fuel pump, then making that module vulnerable to moisture intrusion and mounting where such intrusion is virtually guaranteed (particularly in a vehicle type that frequently gets operated in exaggerated conditions), then making replacement of the module physically difficult by fastening it with corrosion-prone fasteners mounted on the backside of a bracket sandwiched beneath a pickup box (rather than just attaching it from the underside), and ultimately requiring that said module then be programmed for the simple task of running a fuel pump - General Motors, I'm forced to hand you a Jackass Award.

When Engineers Get Bored (or, another small part of why GM went bankrupt...)

2009 Pontiac G5 and 2006 Chevy Cobalt

On its own, engineering a vehicle is no easy task. Every component has to meet conflicting goals of being able to fulfill its purpose (whether it's an exterior part that simply has to look good and not weather fade in six months, or a suspension part that has to survive a decade or more of repetitive salt immersion and continual structural loading) while costing a minimum to produce, and increasingly, it has to do it while being as light as possible while remaining durable enough that it won't fail during its anticipated lifetime. I get this.

Typically, each and every part in an automobile, from the lowliest little clip, to major components like an engine block or body panel, has a part number, and those parts have to be cataloged, inventoried, warehoused, shipped, and carried by their respective dealerships. Imagine the costs involved just in that alone.

So redundancies would seem to be a costly, wasteful proposition, right?

Please note the two cars seen in the photo above. The foreground car is a 2006 Chevrolet Cobalt sedan. In the background is a 2009 Pontiac G5 coupe. Both of these cars are essentially identical, save for the body style (which just happens to vary in these two cars), and some minor trim and fascia parts. Both are built on the same version of GM's corporate platform known as Delta (first sold here as the Saturn ION), very probably in the same facility.

With so much commonality, you would expect that both of them would use identical parts, except where items specific to the two door/four door body and brand-specific differences came into play, right? That would just make sense.

Look a little closer. Notice anything?

One of these things is just like the other (but not)...

These two cars happen to feature an identical 16 inch wheel design (odd, as one is a Pontiac, and one a Chevrolet), fitted with the same 205/55/16 tires size, but they're actually not identical parts; count the wheel-nuts. They use two different bolt patterns. Which means that these cars will have, at the minimum, two different wheel hub/bearing assemblies - at each end, as front and rear are also different - and two different brake rotors (front) or drums (rear). Not to mention the wheels, which are not simply the same wheel with extra holes, as the back side of the casting is unique to each configuration.

Now, in fairness, this 2006 Cobalt has a marginally less powerful version of the same 2.2 litre four cylinder engine used in the 2009 G5 (which may actually have slightly larger rotors), but we're talking less than 10 hp and 5 lb-ft of torque, and in 2006, both four and five-bolt wheels were available in the Cobalt line using the same size brakes. (Later Cobalt SS/Sport and G5 GT models had yet again a different brake set-up, with even larger four-wheel discs.)

So it begs the questions: Who thought that it would be a good idea to engineer and produce two otherwise identical wheels with differing bolt patterns, and - this is the big one - why in hell would you spend the time and resources to create, produce, integrate into the production process, and stock two completely different sets of wheel-end components to meet the same engineering needs in a single vehicle line?

All non-supercharged IONs, even those with the Delta platform's "big" 2.4 litre engine, used four-bolt wheels, so I fail to see any engineering justification, other than perhaps to keep some engineers busy.

On its own, this pointless expenditure would be a drop in the bucket, but enough single drops together can break a dam, and there's little doubt in my mind that this and countless other questionable decisions contributed to GM's financial woes leading into the late 2000's recession and subsequent bailouts. While I'm just as certain that GM is far from being the only company to suffer from this kind of thing, these two cars illustrate the problem brilliantly.

Miscellaneous Ramblings - Suzuki Canada's Golden Opportunity

Suzuki of America recently announced that it will no longer sell passenger vehicles, focusing instead on its more successful motorcycle, recreational vehicle, and outboard motor product lines.

Immediately afterwards, Suzuki Canada released a statement asserting that it would continue with business as usual in Canada. (See the Star's coverage of that news here.)

Given some of the relative crap that Suzuki sold in the States (see below) - which did not precisely mirror what was sold here - and Americans' general disdain for smaller vehicles in general, Suzuki USA's weak sales aren't really all that surprising.

Part of the blame lies with General Motors. GM used to have a good-sized stake in Suzuki, and numerous Suzuki products were sold in both the U.S. and Canada wearing various GM badges, particularly from the mid-eighties to the late nineties. The Chevy Sprint? A Suzuki Forsa. Geo Metro? Suzuki Swift. Chevy/GMC/Geo Tracker? Suzuki (Grand) Vitara.

GM's CAMI assembly plant in Ingersoll, Ontario was originally a joint-venture, building Metros and Trackers, eventually returning the favour with an expanded version of the Chevy Equinox being sold as the Suzuki XL7. Well, sold by the dozen, anyway.

The Suzuki/GM relationship soured somewhat when General Motors bought ailing Korean automaker Daewoo in 2001, giving the General ready access to a newly redesigned generation of compact and subcompact cars built in a relatively low labour-cost country. In the U.S., doubtless with GM's influence, Suzuki ended up shilling several Daewoo products as their own; versions of the Daewoo Lacetti, sold in Canada as the Chevy Optra and Optra5, became the Suzuki Forenza and Reno, while our Chevy Epica (actually Daewoo's flagship, the Leganza) was their Verona.

Worse yet - in an insult to everyone - the Chevy Aveo (sold in both countries; it was the four door version of the Daewoo Kalos) became Canada's Suzuki Swift+. I'm guessing the "+" was to differentiate it from the vastly superior "nonplussed" Swift sold in the rest of the world.

(From the "where did that come from?" department, Suzuki also briefly sold a version of the Tennessee-made Nissan Frontier as the Equator. Available on both sides of the 49th, it was a good truck, but not an obvious fit for Suzuki's product portfolio. Confused buyers stayed away in droves.)

European-spec Suzuki Swift Sport

GM's financial troubles eventually resulted in GM selling most of its stake in Suzuki, which has since gone on to form a co-operative agreement with Volkswagen. Maybe VW was flattered by the startling resemblance between Suzuki's Kizashi and the 2006 Jetta.

While the loss of car and truck sales in the U.S. may be detrimental to Suzuki on the whole (or maybe not, if it was losing money doing it, as this pull-out suggests), this could well prove to be a golden opportunity for Suzuki Canada. Previously, Suzuki's American operations would have had an overpowering influence on product-planning decisions. Now Suzuki Canada is effectively master of its own destiny, at least within the confines of what its Japanese parent allows. Rather than try and move upmarket, as they did with the coolly received (but actually quite respectable) Kizashi, they should play to their strengths.

The Canadian market is much more Euro-centric, particularly in Quebec. We like smaller cars than our southern cousins. Suzuki's specialty is building small cars profitably, to the point where they build or license small cars globally for brands as diverse as Fiat, Subaru, and Nissan. They make small, we like small - sounds like a good match.

European-spec Suzuki Swift (special edition model)

I'd suggest that Suzuki Canada's first move be to bring in the real Swift, and do it, well, swiftly. There's a four door version of the two door hatch seen in the above photos that's nearly as good-looking, and from all accounts that I've read, it's a really nice car to drive too.

So c'mon, Suzuki Canada, bring on the Swift. And don't forget to market it.

(You might want to try offering all-wheel drive in the sedan version of the SX4, instead of just in the hatch, while you're at it - and again, this time actually promote the feature, Subaru-style. I'm just saying, I think that would work...)

To find out more about the Swift, and to see some of the models Suzuki offers in the rest of the world, check out their global website here.

Jackass Award - GM's Theta-platform Front Hubs

2005 Chevrolet Equinox

This is a 2005 Chevy Equinox, one of a trio of first-generation Theta platform mid-size crossovers that General Motors sold in North America, the others being the Saturn Vue and Pontiac Torrent. The Spring Hill, Tennessee-made Vue came first, introduced as a 2002 model, with the Ingersoll, Ontario-made Equinox and Torrent following in 2004.

Seen below is the front hub assembly from the Equinox pictured above; a Torrent would be identical, a Vue similar. Incorporated into the hub is the front wheel bearing, which is a wear item, and a common failure part in many vehicles, though I can safely say that it's a particular weak spot in numerous GM vehicles. When the bearing portion of this part fails, it can result in excessive play, noise, or both. This one was noisy.

Front Hub Assembly - Before

So why is this worthy of a Jackass Award? Well, it's not for the failure rate, which, to be fair, may or may not be statistically significant - my sample area includes a higher than normal percentage of GM vehicles, after all. Not for using a hub design, which while more expensive to replace, is far, far easier to change than the press-fit wheel bearings frequently found in such applications.

No, the Jackass Award goes to the engineering team for this part, who obviously felt that it was not only a good idea to fit a steel hub into an aluminum steering knuckle (steel and aluminum don't play well together, especially in the presence of salt), but to do so with NASA-level tolerances that made this part a tight fit even before the swelling effects of corrosion became a factor. The bolts are what actually hold the hub in place, and there are several other vehicles where there is clearance or even cutouts around their hubs, so this one-step-short-of-press-fit design was unnecessary.

The end result is a hub that is remarkably reluctant to part ways with the knuckle. So reluctant, in fact, that it normally has to be beaten to a pulp to be removed. (The rear hubs on these vehicles don't exactly fall out either.)

Forget using a slide hammer here, as that will just pull the flange out of the bearing. The most effective extraction method, short of removing the knuckle/upright and putting it in a press (which no sane mechanic would want to do), is to baseball-style swing a Bloody Big Hammer at the flange until the hub and knuckle declare defeat, or the mechanic passes out. You can see what the hub looks like after 15 minutes of intermittent hammer swinging in the photo below.

I did mention the aluminum knuckle, right? Be careful not to hit that with the BBH - it's not really forgiving if you hit it in the wrong places. It's not unusual for the sheetmetal dust shield to suffer during this process, either. If it isn't rotten already. Don't worry, the dealer stocks it; you won't be the first to kill one.

Extracted Hub - After 15 minutes of on-and-off pounding, it's feeling a bit out of shape...

You'd think that a vehicle designed in North America and built in Canada would take corrosion into account. As you can see below, corrosion is definitely a factor here. (The machined face of the knuckle has already been cleaned in this photo.)

See the corrosion?

Oddly, the nearly identical component set in a similarly-sized Chevy Uplander minivan does not suffer from the same problem. Yes, it fails regularly, and it does use an aluminum knuckle with a steel bearing, but it normally comes apart with little fuss.

For designing a serviceable part in a manner that makes it far more difficult to actually service than it ever needed to be, while simultaneously failing to take into consideration the real-world environment that this product would face in its primary market, please accept a Jackass Award. A great example of Bonehead Engineering.

...and here's what the new unit looks like. I couldn't leave you hanging.

Jackass Award - Proof That a Mechanic Slept with an Engineer's Wife...

2008 Saturn Vue 3.6 L V6, as seen from below - front is at the top. (See the filter?)

Understandably, service access is not always the top priority when designing a new vehicle; cost and relative ease of production often trump it. I get that.

I also sympathize with the engineers somewhat as they don't always have the big picture, having to design a door hinge or a heater or whatever without specific knowledge of what's around it beyond basic dimensions. This same component - this engine, in this case - also often has to be able to fit into multiple vehicles. That's reality. I get that too.

However, what I don't get, and can't deal with, is stupidity when it comes to stuff that has to be serviced regularly, and oil changes are about the most regular service any vehicle will ever require. An oil filter isn't like a hydraulic valve lifter that will probably last the lifetime of the car. It will be replaced at a pre-determined interval, as many as 50 times in the 250,000+ kilometer life expectancy of the average private vehicle. We're talking several times a year.

So why would anyone put the oil filter in a location that almost guarantees that it will make a mess of the A/C compressor, subframe, and probably the technician? It's an awkward extraction at best.

More critically, why would anybody put it within millimetres of a catalytic converter that's designed to get several hundred degrees of temperature within a minute or so of the vehicle running, so as to virtually ensure that whatever poor sap has to change it will receive severe burns?

You can't tell me that this filter's location didn't receive some amount of consideration, yet there it is. Someone deserves a career of designing glovebox hinges for this one.

It's easy to criticise: what could they have done differently? At the minimum, they could have angled the converter away from the filter - there's clearly space to do so. They could also have made it possible to access it from above (it isn't), or have provided an opening or door in the air deflector to get at it from in front of the subframe (there's almost space for that with this same engine in the larger Lambda models {GMC Acadia, etc.}). Not as preferable, but it could also have been integrated into the oil pan as was the case on the earlier 3.5 litre DOHC Oldsmobile V6.

While GM is by no means alone in doing things like this (everyone does them to some extent), something this dumb feels like it could only be malicious. Proof that a mechanic slept with an engineer's wife...