Perhaps the title of this post is wishful thinking, but I wanted to give an account of what actually happened since I was involved with supplying the kits for both cars. It seems some people felt the need to throw G-Power under the bus a while back in order to settle a personal dispute with other forum members. Since the misinformation being spread around the boards has gotten out of control, I would like to give a clear account of what happened with both of G-Power's US test cars and clear up any remaining confusion about the reliability of G-Power's E9X M3 SK2 supercharger kits. Please note, this has nothing to do with the SK1 systems or any SK2 systems running outside of the US.
I’ve been an importer of ASA supercharger products since 2002 and have a close working relationship with the owners of ASA/G-Power (these companies merged not too long ago). I don't sell or develop products that I wouldn't install on my own personal cars, hence I've always tested ASA's (and now G-Power's) products on my own cars (whether it was an E46 330, E46 M3, E63 M6 or E92 M3). I do this to verify the quality of the products I develop, import and distribute. I can assure you that the hardware and software of G-Power's M3
production kits are not an issue at all. If there was something inherently wrong with the product, every engine would have given up the ghost by now. Clearly, that’s not the case since over 100 kits have been installed on E9X M3s around the world over the past 2 years (most of them in Europe, the UK and the UAE). Below is a pic of my current setup.
With respect to the E9X S65 SK2 supercharger system that ASA/G-Power developed for the US market, I purchased an E92 M3 back in January 09 to test it out and get an idea of the quality, reliability and overall performance (I’m also running their bi-kompressor system on my M6 and have not had any issues with it whatsoever). As of now, the M3 has about 9,000 miles on the clock with the SK2 kit and put down ~ 480WHP on 91 AKI and ~ 545WHP on 100 AKI @ 0.5bar of boost (about 7.5psi). G-Power’s new 91 AKI program has yielded additional gains of 20-30WHP in Germany and should be good for 500WHP+ on a Dynojet (I’ll be able to verify this once I flash the map I currently have with their latest 91 AKI file). I know 9K miles doesn’t sound like much, but these haven’t been “babied” miles in the least. So far, I have not experienced any reliability issues nor an engine failure. The car has been running flawlessly and recently did a LA => Frisco => LA test run without a hitch (roughly 800 mile roundtrip).
Now to the 2 cars that experienced engine failures, as many of you know, the first car belongs to Josh Shokri (tightie) and was boosted in April 09. Josh received a large discount on his SK2 kit and knew full well that his car was going to be one of the first "test mules". Part of the deal included him receiving various hardware and software upgrades throughout the testing process (including larger superchargers, different intake systems and software updates). The second car belonged to Jon Martin and was boosted in May. He also understood his car was a "tester" and would be used to evaluate the kit under different fuel grades, different compressor sizes and different software programs with more aggressive timing tables (he was also fully sponsored by G-Power). I know Josh has voiced his displeasure at times about the service he received; all I will say is there were a lot of people involved and communication did break down at times. Some of his qualms are valid and I don’t intend to say that everything was always handled as it should have been, but at the end of the day these were test mules and both cars were fully repaired on G-Power’s tab (that should say something about G-Power as a company).
As with any new product, there are always going to be "bugs" that need to be worked out, especially when dealing with different types of fuel. In the US, the main problem is fuel quality and octane availability. It's no secret that America is known for having some of the worst fuel in the world. Whether employed by BMW or a tuner like G-Power, this always creates challenges for engineers when developing high-compression engines and forced induction applications. G-Power does have access to 95 RON (91 AKI) in Europe, but they still wanted to see how different the actual fuel quality was and how that would affect the programming and boost settings they have calibrated for European cars vs. US cars.
Essentially, they wanted to see what the highest limits were on the lowest common denominator of US fuel (91 AKI). They now have their answer. While nobody wants to experience an engine failure, this is sometimes the only way to know where the actual limit lies. When BMW blows an engine at the test track or even at the race track, it doesn't mean they don't know what they're doing. You can’t fault a company for simply pushing the limits and standing behind their product in the process. If anything, they should be commended for it, especially since the customers involved had their engines repaired at no cost to them.
As any engineer will tell you, failures are a necessary part of the development process. This is why people shouldn't view this as a "negative", but merely the result of normal testing procedures designed to push the limits of boost, timing and fuel octane on 2 US test cars. This information has now been used to dial in the US production kits with the correct timing and boost settings that will be well within the failure threshold regardless of the fuel octane being used. If these were actual production kits that blew up dozens of engines, I would agree with the harsh skepticism that some have erroneously levied on G-Power's supercharger kit, but that's not what's going on here.
So why did these engines fail? The answer is simple: too much spark advance, not enough octane. As has been discussed ad-nauseum in various threads, the MSS60/65 engine management system on the S65/S85 engines employs a different type of knock-control system, one based on measuring ion-flow across the sparkplug’s electrode, as opposed to one based on piezoelectric accelerometers that measure the actual vibration or “shock” of detonation (like the type of knock sensors found on previous generation M engines). In essence, the spark plug serves a dual purpose in this engine. It not only provides the spark necessary to drive combustion, but also helps the ECU determine whether knock is occurring based on the level of ionization produced inside the combustion chamber. It does this by applying voltage to the plug when it's not being fired during the power stroke.
This type of knock detection system was never designed to work in conjunction with forced induction and can fail to work properly if subjected to high combustion temperatures for an extended period of time or under heavy loads (such as the track or dyno). Due to this, the ECU’s ability to sense knock can be compromised on low octane fuel since combustion chamber temperatures rise significantly as more air is injected into the engine. This in turn can adversely affect the level of ionic current seen by the ECU across the sparkplug's electrode and cause it to stop working properly. If the DME doesn’t detect knock properly, it will fail to retard timing enough to prevent piston failure. Since low octane fuel such as 91 AKI has a higher propensity to pre-ignite under higher temps/pressures, detonation can occur if the engine is pushed continuously under heavy loads (which Jon’s and Josh’s cars were).
As a side note, I understand there’s a debate surrounding the use of methanol and FI on the S65 and whether that caused Josh’s engine failure. I can’t say for certain if that contributed to the piston damage, the jury’s still out as to the effects of methanol on ion-sensing knock detection systems, but G-Power has told me their contacts within BMW have verified this as a no-no, especially with forced induction. Until there is hard scientific evidence to support this (not just hearsay), the debate will surely rage on as to the effect of meth injection on boosted S65s. Personally, I wouldn't run it with 91 AKI, not on this engine. There are simply too many question marks surrounding meth and ion-sensing knock detection. I'd rather run 100 AKI each time (which is what I've been doing).
For those interested, here’s some more info on this type of knock detection system:
http://www.fs.isy.liu.se/~larer/Projects/main.html
http://************/m574m3
http://www.dynotechusa.com/docs/spar...ionization.pdf
Using the data from the 2 test cars, G-Power has now developed a new software update which limits the amount of spark advance to a safe level (regardless of the fuel octane being used) and richens up the fuel mixture in order to reduce combustion temps even further (as an added failsafe). You might ask why didn’t G-Power do this from the beginning? The answer is they wanted to develop a software program that could be used with various types of octane, from 91 AKI to 100+ AKI. This way a customer could take advantage of being able to run both pump and race gas. With the new software update, race gas will not yield a major performance gain as timing is limited to a safe level designed for low-octane fuel only. Customers can still request the other program, but running anything less than 94 AKI will not be recommended as it will interfere with proper knock detection if subjected to heavy loads for an extended period of time.
I know people have their loyalties to various brands and will have their opinions (right or wrong) regardless of what I say, so I'm not going to sit here day after day and try to "pump" G-Power’s products like a cheerleading shill. However, I will defend them when I know from personal experience that when the right type of fuel is used, their kits do not pose a danger in any way and are as bulletproof as can be. This is backed up by the cars running in Europe, the UK and the UAE (as well as my own personal car). I’ve been on these boards (and others) long enough to know that flame wars serve no purpose. I’m sure some will try to pick apart what I’ve written here, but I have no intention of getting into a pissing contest about which kit is better (that’s why you don’t see me post very often). I’m simply stating what occurred since I was actually involved and dealt with the shop and customers that were part of this project last year.
In the end, I’d rather prove the kit works by putting miles on it myself (both track and street) and not waste time engaging in needless banter online. There is a AAA Dragway event open to the public on April 3rd in Fontana. I intend to run the M3 there and provide real world ¼ mile timeslips. Other cars should take advantage of this as well. Even if you’re not confident in your ability to run a solid ET, posting a solid trap-speed is just as valuable. While VBox data is useful as a secondary benchmark, 0-400M is what people ultimately want to know and is IMO the primary metric to use when gauging a car’s straight-line performance (whether ET, trap-speed or both).