SMG Torque Reduction on v3zz

Posts : 15 Join date : 2014-02-26

I need help getting SMG Torque Reduction working. This unfortunately is the make-or-break feature for me. If I can get this to work successfully, I will be continuting down the path of E85, more boost, and ProEFI. If not, however, I will be forced to sell the ProEFI and re-think what I want to do.

That all said:

www drobnak.com /proefi/ADIDCA_3zzS54_Matt_3-Current_State.xcal is the current calibration.
www drobnak.com /proefi/2014-02-26-08-55-01-473-to-train-station.csv is a log with clutch slip and other fun items.

I'm asking anyone who knows how to make this work for some help. Garth sent me over his calibration but it was for _3ib, and the logic seems to have changed since then...

On another note:

www drobnak.com /proefi/2014-02-26-08-52-54-867-cold-flex-bad.csv is a log of what happens when I turn on flex fuel (I have no sensor) to blend the two tables...I assume there is special "cold" logic in the ECU, and it doesn't play nice? If I turn on flex fuel blending after it's been warmed up a little, the car cranks fine.

-Matt

Had to mess with links cause my account is too new.

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: I need help getting SMG Torque Reduction working. This unfortunately is the make-or-break feature for me. If I can get this to work successfully, I will be continuting down the path of E85, more boost, and ProEFI. If not, however, I will be forced to sell the ProEFI and re-think what I want to do.

That all said:

www drobnak.com /proefi/ADIDCA_3zzS54_Matt_3-Current_State.xcal is the current calibration.
www drobnak.com /proefi/2014-02-26-08-55-01-473-to-train-station.csv is a log with clutch slip and other fun items.

I'm asking anyone who knows how to make this work for some help. Garth sent me over his calibration but it was for _3ib, and the logic seems to have changed since then...

On another note:

www drobnak.com /proefi/2014-02-26-08-52-54-867-cold-flex-bad.csv is a log of what happens when I turn on flex fuel (I have no sensor) to blend the two tables...I assume there is special "cold" logic in the ECU, and it doesn't play nice? If I turn on flex fuel blending after it's been warmed up a little, the car cranks fine.

-Matt

Had to mess with links cause my account is too new.

It's pretty simple to set up, just find the slip point at which you want to trigger the torque reduction, and enter the amount of reduction you want, along with the time you want it for.

Posts : 15 Join date : 2014-02-26

Ok, so the question is:

Do the curves in the Torque Reduction tab matter the most now? Because it used to be a standalone table for the SMG stuff. Now that's gone.

Torque Ramp in rate - was set to 1, I tried 75 as per older examples...

Hold time - all set to 1s currently, is that a good starting point?

I have the min set to -10, max set to 20 I think, and what's happening now is it seems to trigger on RPM falling (I turned on fuel cut once to immediately know when it was happening, and that's when it kicked in...)

It's changed a few times since v2, hence why it's confusing...

Admin Posts : 258 Join date : 2008-02-22

Posts : 15 Join date : 2014-02-26

That's helpful.

So now it's all using the auto trans torque reduction tables, and the logic to use (shift points or clutch slip) is determined by Transmission type and SMGTorqueReudction_Enable being Enabled? If so, yes, that actually makes it simpler.

What are the units on the RampInRate map? ft/lb/sec?

What's the difference between the hold time and the TorqueReudctionSwitch_Time_OnTime_cal ?

-Matt

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: That's helpful.

So now it's all using the auto trans torque reduction tables, and the logic to use (shift points or clutch slip) is determined by Transmission type and SMGTorqueReudction_Enable being Enabled? If so, yes, that actually makes it simpler.

What are the units on the RampInRate map? ft/lb/sec?

What's the difference between the hold time and the TorqueReudctionSwitch_Time_OnTime_cal ?

-Matt

Ramp in Rate is in time (mic/sec)

Hold time is just that, switch time is from the switch (trigger).

Posts : 15 Join date : 2014-02-26

So OnTime = Duration of spark retard / fuel cut

Not sure what you mean by mic/sec. Microseconds? so 2000 is 2ms then?

So the problem I'm having is that during shifts, SlipPercent is 0. So the trigger never occurs.

What exactly is SlipPercent computed from?

Posts : 15 Join date : 2014-02-26

www drobnak.com/proefi/2014-02-27.csv

That was my ride home. Tried messing with trigger a little, tried changing the delay times...no go.

Is there an alternative to telling the ProEFI a shift is occurring?

-Matt

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: So OnTime = Duration of spark retard / fuel cut

Not sure what you mean by mic/sec. Microseconds? so 2000 is 2ms then?

So the problem I'm having is that during shifts, SlipPercent is 0. So the trigger never occurs.

What exactly is SlipPercent computed from?

If your slip is zero, then you don't have your trans gear ratios set up properly, and or you aren't getting wheel speed input properly.

Posts : 15 Join date : 2014-02-26

Wheel speed is definitely there.

Transmission ratios are from here:

www bokchoys.com/differential/GearRatios.htm

330i 6spd E46 Sedan manual
4.35
2.5
1.66
1.23
1.00
0.85
3.93
2.93
6500

which is what's in the ProEFI.

What are my options at this point?

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: Wheel speed is definitely there.

Transmission ratios are from here:

www bokchoys.com/differential/GearRatios.htm

330i 6spd E46 Sedan manual
4.35
2.5
1.66
1.23
1.00
0.85
3.93
2.93
6500

which is what's in the ProEFI.

What are my options at this point?

Log your wheel speeds (driven, non driven, and vehicle speed), it can't just be working, it has to be working correctly. log your gear ratios (mt calculated gear) and make sure it's correct. If all of this is correct, it will work. You can't have 0 slip on a gear change, it's physically impossible, so something isn't set up correctly.

Posts : 15 Join date : 2014-02-26

The log from yesterday shows non-zero slip, but you can see the flaring and the waviness as the clutch of engages..

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: The log from yesterday shows non-zero slip, but you can see the flaring and the waviness as the clutch of engages..

O.k. so take the point off of zero at the gear change, and that is where I would start your cut....adjust accordingly til it feels right.

Posts : 15 Join date : 2014-02-26

The problem, as you can see, is that there IS slip detected at the first shift, but not during the later ones...you can see (which I've highlighted) slip falls to zero.

Why is the min value a negative number, if slip percent never goes below 0?

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: The problem, as you can see, is that there IS slip detected at the first shift, but not during the later ones...you can see (which I've highlighted) slip falls to zero.

Why is the min value a negative number, if slip percent never goes below 0?

You can clearly see the slip in in that log...you are looking at the point when the clutch is already fully engaged, where the rpm goes up is wehere the slip is. You should probably get this to Garth so he can set it up.

Posts : 15 Join date : 2014-02-26

Compare behavior:

SMG Torque Reduction on v3zz Dme_spark

This is with the DME controlling spark. While it's not 100% perfect, you can see the difference here.

Maybe this is something I missed earlier - is the torque reduction intended to reduce torque as the next gear is ENGAGED? (The red circles) If so, that's a secondary problem for me.

I'm trying to stop the RPM flare, as you can see the DME pulls torque as soon as the shift is commanded...

Can this type of behavior be done with the clutch slip detection mechanism?

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: Compare behavior:

This is with the DME controlling spark. While it's not 100% perfect, you can see the difference here.

Maybe this is something I missed earlier - is the torque reduction intended to reduce torque as the next gear is ENGAGED? (The red circles) If so, that's a secondary problem for me.

I'm trying to stop the RPM flare, as you can see the DME pulls torque as soon as the shift is commanded...

Can this type of behavior be done with the clutch slip detection mechanism?

Your setup isn't right, get it to Garth, he can help you. While the clutch is engaged (normal driving) and in gear, the clutch slip should be REALLY close to zero. You can CLEARLY see in you log that it is measuring slip, and you can clearly see when the rpm flairs that the slip is following it (that's when the trigger should be happening). Torque is reduced while the gear is changing, and we are CLEARLY capturing that time. Your gear ratio's aren't exactly right, you need to remember, those are calculated values, so you will need to adjust the scalers or the gears to get zero's in every gear for it to work properly.

Posts : 15 Join date : 2014-02-26

I can do that, but I'm not expecting miracles. I'm going by published information, and what the configuration of my car is supposed to be.

How much time is involved to implement the CAN torque reduction requests (if it's on the shared can bus)? I see that the functionality is in there on 3zz.

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: I can do that, but I'm not expecting miracles. I'm going by published information, and what the configuration of my car is supposed to be.

How much time is involved to implement the CAN torque reduction requests (if it's on the shared can bus)? I see that the functionality is in there on 3zz.

You have everything you need to do this CORRECTLY.

Obviously the shift happens before what you are trying to say, or the RPM wouldn't be dropping down prior to that. The shift happens as soon the clutch slip shows it...you are looking after the clutch is engaged. You need to get your ratios set up properly so the slip values are accurate....end of story.

Posts : 15 Join date : 2014-02-26

Something I missed earlier - you mentioned scalers - is that wheel speed scaling items? It is possible that as in the case of coolant temp, the CAN datastream id is the same, but the math is different, no?

I will compare my wheel scaling items to Ace's.

Edit: All I see are 1's in the scaling items. However, he had 3.15 gears, and I have 2.93 gears. It looks like all the gears show 4% slip...should I change each gear, or just the overall drive ratio?

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: Something I missed earlier - you mentioned scalers - is that wheel speed scaling items? It is possible that as in the case of coolant temp, the CAN datastream id is the same, but the math is different, no?

I will compare my wheel scaling items to Ace's.

Edit: All I see are 1's in the scaling items. However, he had 3.15 gears, and I have 2.93 gears. It looks like all the gears show 4% slip...should I change each gear, or just the overall drive ratio?

You want it right, so adjust what is wrong. I.E. if you are driving down the street at a verified 60mph (GPS, or similar), and your front wheels say they are going 55 and the rears say they are going 62, adjust the scalar until they are correct.

Posts : 15 Join date : 2014-02-26

Comparing the a DME log using BMWLogger and ProEFI's log..
1st gear 1883 - Dr 10 / ND 11 / DME 11
2nd gear 2490 - Dr 14 / ND 14 / DME 14
3rd gear 2207 - Dr 21 / ND 20 / DME 21

I compared it in multiple places, up to about 35 MPH, and it matches within 1 MPH.

This is with 2.93 as the rear end ratio, 1 for the CAN scaler on ND and Driven wheel...

So if driven wheel speed and non-driven wheel speed seem correct, what's the next part to check?

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: Comparing the a DME log using BMWLogger and ProEFI's log..
1st gear 1883 - Dr 10 / ND 11 / DME 11
2nd gear 2490 - Dr 14 / ND 14 / DME 14
3rd gear 2207 - Dr 21 / ND 20 / DME 21

I compared it in multiple places, up to about 35 MPH, and it matches within 1 MPH.

This is with 2.93 as the rear end ratio, 1 for the CAN scaler on ND and Driven wheel...

So if driven wheel speed and non-driven wheel speed seem correct, what's the next part to check?

Like I said, all of the data must be accurate, it's off somewhere or it would be calculating correctly, you need to find out where it is off.

Posts : 15 Join date : 2014-02-26

And I'm asking for the next place to check.

Admin Posts : 258 Join date : 2008-02-22

mdrobnak wrote:: And I'm asking for the next place to check.

Mathew, I can't walk you through the entire process. I gave you a list, gear ratios, and wheel speeds, so whatever you haven't checked and confirmed with data is the next step. If you are having a hard time understanding that, get with Garth, and have him do it for you.