a few feature requests

'Microsquirt' Transmission Control development and support.

Moderators: jsmcortina, muythaibxr

Post Reply
ashford
Super MS/Extra'er
Posts: 1603
Joined: Sun Apr 27, 2008 4:29 pm

a few feature requests

Post by ashford »

i've been putting off making this post for awhile now. i have a few things i would like to see get added or changed

first the easier things
1. be able to change leaving torque locked during shifts. leaving it locked on shifts on high stall converters is a necessity to avoid rpm flares between 3-4 shifts while converter is unlocked. for example light acceleration running at 1800 rpm in 3rd unlock for 4th rpms jump to stall speed for that throttle position(abut 2500) then drop to 1400 for lockup. the car actually looses acceleration during this period.
when there is a stockish converter the 3-4 shift is harsh.

2.have a generic output or cooling fan control


the last part is a bit more in depth about shifting timing.
i did a big upgrade for my friends 4l80e. a few hard parts,roller conversion, friction upgrades and a d3 transbrake. line pressure is now fixed, no accumulators and the trans should be reliable to 1000hp. the on thing i didn't expect is the increase in time in the 1-2 shift. a 1-2 shift is taking .7-.8 seconds not slipping just a delay, i called the manufacturer and they said that is normal. 2-3 is about the same and 3-4 is quicker than the stock valvebody with shiftkit. i looked into this after the first few runs at the track looked goofy
too short time.png
then i overlayed the 1-2 shift from before and after the upgrade. the darker colors is before and the greyed is after
beforandaftertransbrakeoverlay.png
i increased the shift time some and had him take a run he said it was acting funny during the shift, the timing retard came back right before the shift and boost target get upped right at the actual shift occurs, it was working against what we are trying to do, so i just disabled it till hopefully the code can get a few changes.

this is my proposal on this. splitting the shift time into 2 sections. ill just call it A and B for the time being.

1-2.png

in region A the beginning is when the shift is commanded and the end of it is where the shift actually starts. for all intents and purposes A is still in the previous gear transfering torque without slipping or ratio change. this seems to me to be the time in which it takes for the field in the solenoid to colapse plus the time it takes to vent the signal line and move the valves nessesary to make a shift.

region B is where the shift actually occurs.

for region A there is no point in doing torque reduction(timing retard), i would like this to happen in region B.
there is a possibility that once a 1-2 shift has been commanded it can return to 1st if it is reapplied while still in region A, if that is useful or not.

with this trans and what i'd like to see done there would need to be different shift times for different gears

i would also like the "GEAR" to be either split up into "commanded gear" and "in gear" where "in gear" is what is sent to the engine controller so boost by gear dosent come before it is actually in gear or have "GEAR" that is broadcasted after shift time is completed insted of when it is comannded.
ashford
Super MS/Extra'er
Posts: 1603
Joined: Sun Apr 27, 2008 4:29 pm

Re: a few feature requests

Post by ashford »

i forgot a simple one

can there be a 2 or 3 point idle step/pwm adder it would be idle adder vs vehicle speed. at some point around 4-8 mph engine torque draw becomes zero at idle and is no longer needed. if i use it so that n vs d results on the ms closed loop idle is doing nothing i get a nice cruise control around 25mph
LT401Vette
Super MS/Extra'er
Posts: 12697
Joined: Sat Jul 16, 2005 8:07 am
Location: Moorseville, NC
Contact:

Re: a few feature requests

Post by LT401Vette »

i didn't expect is the increase in time in the 1-2 shift. a 1-2 shift is taking .7-.8 seconds not slipping just a delay, i called the manufacturer and they said that is normal. 2-3 is about the same and 3-4 is quicker than the stock valvebody with shiftkit. i looked into this after the first few runs at the track looked goofy
I see the very same thing in my car.
I have to target shifting by 6500 RPM to shift by 7200.

That is a work around, but it has other fallout issues.
1) Ignition retard is rendered useless, as it retards when the shift is triggered so you are back to full timing by the time the shift actually starts to happen.
2) changes to rate of acceleration such as a nitrous pass or adding/removing boost make your delta RPM change and shift RPM change.

If there was a configurable shift time delay for each gear, it would help, but to help with #2, it really needs to predict the delay based on delta RPM.
Phil Tobin
EFI Analytics, helping to simplify EFI
Next Generation tuning software.
Supporting all MegaSquirt versions and firmwares.
http://www.TunerStudio.com
http://www.efiAnalytics.com/MegaLogViewer/
Support the firmware running your engine:
http://www.msextra.com/doc/donations.html
whittlebeast
Super MS/Extra'er
Posts: 2221
Joined: Tue May 04, 2004 8:20 pm
Location: St Louis
Contact:

Re: a few feature requests

Post by whittlebeast »

I think it comes down to expected shift time (from the time the ECU commands a shift until the time that the clutches in the transition and what is that a function of?

Torque from the motor (generally a function the PW)
HP from the motor (generally a function of the duty cycle)
RPM per sec accel rate of the motor (generally a function of mass of the car, gear ratio of the gear that you are coming from and the two things above)
And if the throttle position is changing (this one may be the really tough part to model)

And if there is a better way to model this in some sort of table.

Andy
Raymond_B
Super MS/Extra'er
Posts: 1394
Joined: Thu Mar 06, 2014 2:17 pm
Location: Texas
Contact:

Re: a few feature requests

Post by Raymond_B »

Have you noticed *more* of a delay with the MS over other trans controllers? Or you are simply trying to eliminate any delay? The reason I ask is when I used to do chip tuning you never actually used the specific RPM you wanted the shift to happen because of the mechanical delay in the trans (like what whittlebeast posted). So I would always command it a few hundred RPM lower, sometimes even more if it was a stock trans.
1995 Ford Lightning. Dart based 427 Windsor, Novi 2000, full sequential, E-85, etc. MS3X/v3.57
http://www.buildpics.org/
ashford
Super MS/Extra'er
Posts: 1603
Joined: Sun Apr 27, 2008 4:29 pm

Re: a few feature requests

Post by ashford »

Raymond_B wrote:Have you noticed *more* of a delay with the MS over other trans controllers? Or you are simply trying to eliminate any delay? The reason I ask is when I used to do chip tuning you never actually used the specific RPM you wanted the shift to happen because of the mechanical delay in the trans (like what whittlebeast posted). So I would always command it a few hundred RPM lower, sometimes even more if it was a stock trans.
im not really worried about the delay itself, i can tune around it. my gripe is that torque reduction comes at the wrong time or for a long time, it only needs to be in the last 1/3rd or so of the comannded to completed time. the other one is that if doing boost by gear. boost comes up at the comanded time so it is there in the last part of the gear and during the shift.

i have tuned oem pcms, gm adapts line pressure to meet programed shift times to an extent and learns shift time like a ltft and adjust torqe reduction to match the transition of gears. ford has a goofy torque based system.
Raymond_B
Super MS/Extra'er
Posts: 1394
Joined: Thu Mar 06, 2014 2:17 pm
Location: Texas
Contact:

Re: a few feature requests

Post by Raymond_B »

Gotcha, and yes Ford's torque stuff is weird!
1995 Ford Lightning. Dart based 427 Windsor, Novi 2000, full sequential, E-85, etc. MS3X/v3.57
http://www.buildpics.org/
Post Reply