Tuning a 2007 Mustang GT with Roush Stage 2 TVS Supercharger

[Littlegiant]

Hi guys!

New to this forum but been around the automotive world a while. Bought a new 2007 Mustang GT and over 2 years turned it into my version of a GT500. Aside from the body kits, hood, nose etc, I also worked on the powertrain & suspension. For the engine I bought a Roush stage 2 supercharger running about 12psi of boost from the new 2300 Eaton TVS rotors. When first installed it ran awesome and ran that way for 2 years. Since the car never sees winter much less rain, I store it, and bringing it out last year I ran into driveability issues. Since I could not get much help from Roush in Plymouth Mi, I purchased the HP Tuner VCM Suite and decided to have a go at what was happening. I even had Roush reprogram the PCM again, just to verify the calibration file was not corrupt, and it wasn't. I doubt the engine strategy is the issue, much less the tune, but I needed a powerful program to tell me what the PCM was seeing and what it was doing. In short, under partial load and low boost, the engine backfires through the exhaust as if it is running too rich. I replaced the obvious things from suggestions by Roush (MAF, ETC TPS, all 4 Narrowband O2s and spark plugs) but no change. I already have a boost gauge, wideband O2 system and fuel rail pressure senders. Due to Roush locking out the tune, I cannot read or edit the tune, which I was kind of ticked about since that was part of the expensive package.

However, I have used the VCM scanner to grab a bunch of run time data to show me what was happening. I am currently setting up a slightly better triggering system so I can tell within the plots when the engine backfires. In short, from preliminary data, it looks like the PCM is retarding the ignition timing to 0BTDC to even 5ATDC, and that is when the backfire occurs (sounds logical). Question is why, and what should I look at/for.

2nd question is does anyone have an unlocked supercharger tune for a 2005 - 2008 Mustang 3 valve 4.6L that I can view and play with? I purchased another PCM to program to my VIN so I can edit my own tune but would rather not start from scratch. Since HPTuners could not open my locked Roush tune and Roush would not send an unlocked one, not much choice.

I have a pretty good clue as to how engines and electronics work, based on working on cars & engines for 30+ years and being an automotive electronics engineer (I actually designed the hardware for my Ford PCM) for 25+ years, and I have done some side work tuning to practice plus hang out with the engine calibrators to learn. This issue throws me a little since I cannot figure out why the PCM is doing what it is. The Knock sensor amplitude does not go goofy right before or during the event either.

so, for the newbee to the board, HELP!!!!!!

Thanks guys for any input.


One other thing, attached is the tune from my Roush programmed PCM. With my VCM Editor I cannot open this. Can anyone else unlock it? I really appreciate you guys even reading my post.

[SultanHassanMasTuning]

a data log would explain alot

[Littlegiant]

a data log would explain alot

Here is the last log file I have from the dyno analysis I did. Basically when the spark timing hits 0 or ATDC is when the backfire occurs. Since the weather here has been really crappy I have not been able to get out on the road and get more data. I now have one of the inputs to the MPVI to go high when a backfire occurs so it will be easier to determine when the event occurs, but for now this is what I have.

If I need to be recording more or different please let me know. I take all input. Thanks

[toyoguru]

I was able to open the tune with 2.25 beta. I don't have anything to compare it with though. Also, you used the 2.24 scanner when you did your log file, which renders any knock values useless. Try scanning with 2.25 beta.

I did make a quick histogram and compared throttle angle vs rpm vs load and anytime the car was over 30% throttle, the load was 90% or higher, not sure if there is anything to that or not. Only reason I did that was I noticed that one of the instances in which the timing went negative (atdc), the load was 100%, at 1700rpm at 32%throttle. Something about that doesn't sound right, but I could be wrong.

[PST]

Haven't looked at the files, but one thing stands out to me. It's that the car apparently just started running like this out of the blue for no reason. Generally, this is not something to blame on the tune as you stated in your first post.

What is the wide band reading just before it backfires (again, I haven't looked at the logs)?
Do you have any fuel pressure data? What's it doing?
Are fuel trims outside of the +/- 5% range?
I would confirm that fuel supply is good first. After that suspect trashed cats if present or weak ignition coils.

[Littlegiant]

I don't have any computer data for the fuel pressure although I do have a gauge which is measuring rail pressure, and from what I am told by the Roush people, the rail pressure is right where they expect and tune it to be.

I do have fuel trims get as high as +/- 30% change which is way too high. I did pay to have Roush reflash the PCM and it made no difference. What is strange is that once I get past this backfire issue, which seems to be load/RPM related, and I get into true higher loads and higher boost, the car runs like did, like a bat out of hell, so it seems that whatever is happening is affecting low to mid RPM and low to mid load ranges, but not WOT max load values. Even at idle it will stumble some.

I did replace the spark plugs but not the smart coils since it did not seem that they were misfiring.

[PST]

Whats your plug gap? And for what it's worth, I have seen coils cause low rpm, mid range load misfires. Set your gap down to like .025-.028.

However the high fuel trims will likely be the best indicator to investigate. I will be able to look over your logs tomorrow afternoon. In the mean time, try to out good fresh gas in it and get another good long log. Log short and long term trims, maf, iat, ect, timing, spark source, rpm, mph, air load, o2 millivolts, knock and wideband.
Start it up cold, start logging, go drive it normally till warm, then make it backfire/act up and get some wot pulls in like 2nd-3rd gear.

Well see what it is from there. Oh yeah, and use the latest beta.

[GHuggins]

Whether it's varnish or trash built up in your fueling system, think you might want to do as PST said and double check your fueling. You can watch your datalog on the "dials graphings" and you should be able to see something pretty bad off. Atleast this is what I was seeing. Everytime you get into throttle, bank 1 goes rich as it should, but bank 2 barely gets above or around 200 or so mv's... Plus your air fuel gage is hitting anywhere from 16 to 21 which is where I assume your wideband maxes out at about the same time. Injector rails are easy enough to pull and check. I assume when you checked fuel supply, you watched fuel pressure when getting into boost as well as your normal pressure checks? I would make sure air and liquid will pass through your rails and rail crossover easily. With todays fuels, who knows what may have come apart and gotten lodged somewhere if that's what indeed has happened here. Do you have a fuel filter on this year model? Would replace that to if you do. Oh, and if it's running this badly, wouldn't be doing wot's until problem is sorted out.

Something to think about atleast.

[1202896042]

check if you have some rocker misplaced. the camshaft damaged a stuck valve, I hope it helps in something.

[Brice]

I've tuned 2 2006 Roush Mustang last week.

The first one was a roush kit stage 2 installed on a factory mustang (aftermarket)
The second one was a real roush stage 3.

They had exactly the same tune. (stockbeta.hpt to open with the 2.25 version).

The real roush one had LT Headers, pulley and no cats (the Customer ran 2 years without tuned and the issue and destroyed 2 cats)
The LTFT and STFT were out the place, switching from (-35% to +35%).
(the maf was out about 20%).
DTC : Lean and Fuel System error
I was not really confident to tune the car as it seems to have mecanical issue (alternance of black and white smoke on the exhaust pipe)

I've tuned the car with a wideband (about 20% lean) on open loop and the car ran perfect, fuel pressure was perfect too.
I've reenable the lambdas ... and from time to time (just after restart, the lambda went crazy again). The mecanics changed the lambda 2 times too.

I've finished to disabled to lambda and run open loop. (I've played with delay and transport without results)

[PST]

Ghuggins, good advice on the fuel filter/physical checks. That's part of what I was leading to with my suggestion to check the fuel system. The car sits for extended periods. Fuel filter and fresh fuel would be a priority every time my car came out of storage.
if a car ran great un-touched for years, then started acting up. The problem is 99.99% of the time going to be mechanical.

[GHuggins]

Ghuggins, good advice on the fuel filter/physical checks. That's part of what I was leading to with my suggestion to check the fuel system. The car sits for extended periods. Fuel filter and fresh fuel would be a priority every time my car came out of storage.
if a car ran great un-touched for years, then started acting up. The problem is 99.99% of the time going to be mechanical.

Yes with todays fuels and all the ethanol in gas, where the fuel itself is only good for about 3 weeks tops... Plus I've seen WAY too much damage over the last 10 years from where fuel was left in a tank or system and allowed to eat away at the factory rubber and plastic components. You can pull injectors and find what apears to be coffee grounds in their filters. Makes for an interresting multiple cylinder intermittent misfire hunt... Never fun.

Could see a stuck valve from sitting, but would have a constant misfire from the dislodged follower. Cats yes, but should have noticed something before storage the previous year. I would go after fueling issue just from 02 sensor data alone. It's kinda pointing to the problem. Now just have to figure out what that problem is...