Transmission Tuning (advanced tech)

[engineermike]

I suppose it's the oncoming clutch ramp rate that would be the culprit

Regarding the flare? It might be but would it be the ramp time or the torque transfer rate? That’s why I started this thread because I don’t understand these relationships.

That said I accidentally might have figured something out. I noticed a long tie-up on a particular shift. The acceleration delay before rpm change seemed longer than normal. This is during the torque transfer phase of the shift. After digging, I found I had inadvertently used all the character 0 shift data but missed the stroke pressure. It was using alt 1/2 stroke pressure with char 0 shift data. My theory is it was engaging the incoming clutch too early, before releasing the offgoing.

All that to say that increasing stroke pressure might help with your flare issue. It makes sense intuitively. If something is different about your trans that delays grabbing of the next gear then more stroke pressure should help that happen sooner.

Sponsored

 

[shogun32]

idle question. if the computer knows wheel speed, knows what gear it's coming out of and what gear it's going into, why do we even get flare? the engine RPM should have already been forced down to the RPM suitable for the next gear, no? Just like 'rev matching' is done for stick?

 

[engineermike]

idle question. if the computer knows wheel speed, knows what gear it's coming out of and what gear it's going into, why do we even get flare? the engine RPM should have already been forced down to the RPM suitable for the next gear, no? Just like 'rev matching' is done for stick?

Rev (and torque) matching is accomplished by manipulating the throttle. Auto trans shifting engine rpm is controlled on shifts by the inertia phase of the shift, manipulating two solenoids to release one clutch and engage another, while leaving the throttle at the same state.

 

[Robottrainer]

idle question. if the computer knows wheel speed, knows what gear it's coming out of and what gear it's going into, why do we even get flare? the engine RPM should have already been forced down to the RPM suitable for the next gear, no? Just like 'rev matching' is done for stick?

Back in the good ol days, a flare, considering a mechanically sound trans, meant the timing of the oncoming clutch vs the off going clutch/band was off allowing the trans to be sort of in neutral. You could fix it by changing springs/number of springs, flow. On the flip side was a bind...when two set of clutchesband or bands were on for a split second. This would give you the tire bark you would here with a shift kits. A good shifts should be quick, positive and almost unnoticed in my opinion

 

[shogun32]

Auto trans shifting engine rpm is controlled on shifts by the inertia phase of the shift, manipulating two solenoids to release one clutch and engage another, while leaving the throttle at the same state.

right, so if there is a delay and effectively low/no load on the trans the engine will zoom. What's so hard about clamping the RPM (modulate the throttle?) so it doesn't go higher than its previous RPM let alone the RPM it will be at when new clutch lockup + a smidge? They way you described it they only implemented half the necessary logic and are hoping the clutch swap is timed "right" so the new gear will drag the RPM down and otherwise impart a 'surge' into the power train depending on how large a difference is between the 2 states.

You have all this control and sensors and they can't manage to hit a 100RPM window and adjust on the fly based on deviation from desired behavior?

IMO a properly programmed auto should be as subtle as a CVT - perfectly blending clutch release/takeup with throttle that you're left wondering (short of aural or gauge readout) did a shift happen? Hell, I can do that with 2 hands and my ear on my motorcycle, both up and down. (yeah yeah I realize a MC has fk all torque and rotational inertia and the wet clutch gives a huge window of variability)

 

[engineermike]

.

You have all this control and sensors and they can't manage to hit a 100RPM window and adjust on the fly based on deviation from desired behavior?

You’re pretty much describing the intent of the controls. They program what should work and a feedback control plus learning account for physical differences. But just like airflow, fuel trims, and cam timing, there are limits as to how much it can correct. There is most likely something out of tolerance in the OP’s car that is causing this issue. He’s basically just tuning around it because that’s easier and cheaper than a new trans.

The functionality doesn’t exist to control rpm during a shift using the throttle because the expectation is that the clutches do that. Fwiw torque modulation is controlled using spark timing retard.

IMO a properly programmed auto should be as subtle as a CVT - perfectly blending clutch release/takeup with throttle that you're left wondering

Between torque hole filling (torque phase) and torque modulation (inertia phase) It totally has the ability to do this, but it turns out that most people want to feel the shift. This is actually why I’m using mustang character 0 shifts in my f150.

 

[Weeee]

PSA - Ford and GM co-developed this transmission. Expand your research to include Camaro A10 threads; you'll find more helpful information.

Here is how one dude fixed the 4-5 flair in the tune.

 

[shogun32]

PSA - Ford and GM co-developed this transmission

Given the Ford product is such unbelievably sh+t compared to GM's I very much doubt Ford did any actual engineering. and simply bought the crate version with its half baked controls.

Ok maybe that's a tad harsh. But if I was counsel for GM I would send Ford a cease and desist letter to stop telling the world it was a joint program. And that their claims were damaging the GM brand.by being associated with Ford.

 

[engineermike]

What is it that makes the gm version better?

 

[shogun32]

What is it that makes the gm version better?

It actually works. It's night and day. Gm is a Rembrandt, Ford is a 5yr old finger painting. It is truly shocking how bad the mustang a10 is.

 

[engineermike]

It actually works. It's night and day. Gm is a Rembrandt, Ford is a 5yr old finger painting. It is truly shocking how bad the mustang a10 is.

I was hoping for specifics.

 

[Qcman17]

It actually works. It's night and day. Gm is a Rembrandt, Ford is a 5yr old finger painting. It is truly shocking how bad the mustang a10 is.

I wonder if the fact GM still uses a dedicated TCM has anything to do with it?

 

[engineermike]

I wonder if the fact GM still uses a dedicated TCM has anything to do with it?

The ford tricore has no shortage of processing power or speed.

I had heard some time back that the 10r in the Camaro was better so I pulled the tune file to compare. The shift schedule is drastically different. The Camaro shifts at much lower rpm, has no skip shift, and downshifts with less throttle input. These are all easy enough to copy over but do we really want to shift at lower rpm? This is why I’m asking for specifically what is better about the Camaro 10r80.

 

[markmurfie]

The ford tricore has no shortage of processing power or speed.

I had heard some time back that the 10r in the Camaro was better so I pulled the tune file to compare. The shift schedule is drastically different. The Camaro shifts at much lower rpm, has no skip shift, and downshifts with less throttle input. These are all easy enough to copy over but do we really want to shift at lower rpm? This is why I’m asking for specifically what is better about the Camaro 10r80.

Ford TCC apply and shift points for the A10

GM TCC apply and shift points for the A10

I think its obivous from comparing what GM tuned and what Ford didn't and why people think they are so different. Ford did not skip this on the 6R80 cars either.

I did modify some of the extraneous values to clean up the visual in comparing these. And GM is in MPH Ford is in OSS, so its not quite apples to apples, but similar.

 

[shogun32]

This is why I’m asking for specifically what is better about the Camaro 10r80.

I don't want to kill myself or drive the car into a lake after 30 minutes? The f150 was no gift to humanity either but a tune from Unleashed at least kept my teeth from grinding. Camaro Shifts were blended nicely and consistently up and down. When commanded they executed promptly. Even rapid fire down shifts didn't involve neck snapping.

The lower rpm thing is surely because of the torque curves between the 2 motors. And would vary widely based on throttle input.

The number and count of the clutch disks is probably different and the valve packs likewise not at all compatible.

Sponsored