Car accelerates on its own?

Sometimes.. and Im not sure how to do this on purpose, when in first gear, I will just want to get the car moving, as in edging up a bit from a full stop. So I just give it a lil bit of gas, and then go for the braek. as I'm about to push on the break, the car keeps goin, even accelerating, as if I was still pushing on the gas! Can someone explain this to me? Also, Im a computer programmer, not a mechanic, so if it could be dumbed-down-as-possible that would be great. Thanks.

It is due the the drive by wire throttle - it can be very annoying at slow speeds in traffic.

My RPMs will also increase to ~1500-1700 when negotiating a turn while the clutch is in. Dunno if it's also related.

I get that same "surge" when I try to pull up my drive way.

DBW sucks. I wish we had a throttle cable.

Ah yes.. forgot to mention, it does seem to happen more when going uphill just like SuperCleanLGT said when going up his driveway. This is downright dangerous isnt it?? Is it a manual trans thing or just another loveable Subaru quirk? :-)

Could someone explain why drive-by-wire causes this?

 

My other thought was it could be a bad MAF - that can cause weird surges - but it usually doesn't happen at low speeds.

If it wasn't for my ninja like reflexs I would have parked in the kitchen!!!

have the same issue, assumed it was the DBW, but didnt have the problem with my TSXs DBW.

1st gear surge....everyone has it ....deal with it....

Yup, gotta disengage the clutch in 1st gear around turns because the e-throttle "sticks." Just another potentially deadly issue Subaru never cared to fix in our surging, bucking, stuttering, stumbling cash pits.

DBW sucks. I wish we had a throttle cable.

 

+1

But without DBW we wouldn't have gadgets like SI-drive to artificially neuter the driving experience!

You have an Audi 5000. Try another forum...or 60 Minutes.

It is NOT a bad MAF, it is Subaru's imitation of Microsoft

 

Sorry, I didn't mean to imply that Subaru ever did anything that wasn't perfect

The only time I have an unexpected increase in RPM is when the clutch energizes for the compressor to run the AC or defrost.

Let us not forget the turbo wanting to take off too. I stab the clutch alot.

I really doubt that there is some inherent property of DBW that makes it suck.

 

However I do suspect that the tables that determine how DBW works are making it suck.

 

Real soon now I'm going to get off my ass and test this theory.

 

Here's the table that I think is causing the problem. This is from a modified ROM, but IIRC the unmodified table is pretty much the same. I don't have the stock ROM handy right now though so you'll just have to take my word for it. Or better yet, get yourself EcuFlash, RomRaider, and a Tactrix cable, and see for yourself.

 

http://i193.photobucket.com/albums/z151/Legacy_NSFW/RequestedTorque.png

 

Notes:

 

The 'requested torque' value has arbitrary units. It ranges from 0 to 400, but it would be just as meaningful to divide those values by four and show them as 0-100%.

 

(I think in my stock ROM the values range from 0-320, but again I don't have it handy.)

 

There's a second table that maps 'requested torque' to 'throttle plate angle,' it's also got RPM as a vertical axis.

 

I suspect that the cruise control signal is added to (or subtracted from) the 'requested torque' value, so that cruise control gets affected by the 2nd table but not the 1st table.

 

Note that at 25% pedal angle this table calls for about 50% of the available torque.

 

I think this is why the accelerator pedal is so touchy at low RPM.

 

On one of the tuning forums, someone posted a European or Australian ROM (I forget which) that had the opposite curvature in this table. Like, 50% torque didn't happen until 75% pedal angle, or something like that.

 

And finally, I should add that the people who actually look at the ROM code are finding new stuff semi-regularly, and it's possible that the current interpretation of of these tables will be tossed out and replaced with something more accurate in the future.

 

Anyway, I'm going to make the throttle-pedal-angle / requested-torque table increase linearly, and we'll see what happens.

Addendum: the values in the table above go up to 400, but that's because they were modified by a tuner who didn't understand the table very well. No disrespect intended though - remember what I said about learning new stuff all the time.

 

The torque-to-throttle-plate-angle table only operates on inputs up to 320.

 

At one time, the throttle-pedal-to-torque table was shown only going up to 80, and the torque-to-throttle-plate-angle table hadn't been discovered. So some people make the 1st table go up to 100, thinking they knew something that Subaru didn't (OK, perhaps just a little tiny bit of disrespect intended here). So anyway, the table above actually made the twitchyness problem a little worse. But the shape of that table is pretty close to the shape of the original table, it's just that the raw values have been scaled up a bit.

sorry this is off topic but it bugs the hell out of me...it's "brakes" not "breaks" :-/

First time it ever surged on me, I was in stop and go traffic and I damn near rearended the car in front of me before I dumped the clutch and hit the brakes.

 

I simply learned to adjust my driving habits (following distance and throttle inputs that seem to create the problem) at stoplights or in stop and go traffic to account for the 1st gear surge when it occurs.

Does this happen on the Automatic transmissions as well? I wouldnt think so right?

Hey shiggins- You gonna mod that bish?

why you guys gotta keep bringing up flows in my subaru?

I have the same problem at slow speeds. It sucks, but i have gotten used to it. However, one time i almost got into an accident.