Logging slip

[tobmag]

Hi Lance.

I was reading thrue the values possible to log.
I've seen that "slip" was in there and took for granted that it was possible clutch/brake slip in the tranny but its slip over the converter.
Is it possible to get a log of brake/clutch slip as well? ofcourse you need a ISS and VSS to do it.

Also when slip is detected is it possible to;
>Trigger an output for a warning light,
>Increase main pressure?
>Retard timing to cut torque?
>Possible to link this into MS3 to reduce boost?


BR//Tobmag

[Bernard Fife]

Tobmag,

The outpc structure (to the PC/TS) used to include an estimate of the slip, but it was of limited use and removed a while ago because:
- a lot of trans don't have an input shaft speed (ISS) sensor,
- each byte of data in the outpc structure has to be transmitted from the trans controller to the engine controller (via CAN) and then to the PC (serial). So we want any values that could be calculated by TunerStudio (using the much faster PC processor rather than the trans or engine controller) to be done in the INI rather than in the embedded code (and internal slip falls into this category).
- the speed sensors (ISS and VSS) are often troublesome to get right in non-OEM applications because of electrical interference issues./ As a result, a fair bit of filtering is required to get a usable signal,. This filtering degrades the accuracy of the information the signal is carrying. For the VSS, if the speed is off by 1 or 2 mph, that's no big deal, it can be adjusted in the shift table, etc. But if we are comparing VSS to ISS, then the algorithm is much less reliable.
- <edit>slip during shifts is necessary (because the trans is briefly in two gears simultaneously), so without more information we would have to ignore slip during the shifts. But often it is excessive slip while shifting that burns up friction elements, not necessarily slip while in gear. So if we hope to prevent friction element damage, at a minimum we would need a very good estimate of the slip magnitude and what a reasonable threshold would look like (and this may vary by each gear change, without which all of this may accomplish nothing in the end.</edit>
- In general (and coming from a road race point of view) I'm not sure we should be slowing the car down if there's a problem. We should either stop and fix it if we are driving on the road, or if it's a race - retire or drive it until it blows up - there's no sense racing slowly. People often overlook the most sensitive and sophisticated 'smart sensor' in the car - the driver - they will usually know something is wrong and take the most appropriate action long before the sensors ring an alarm. And there's a philosophical angle to this too (again from a race point of view): don't put anything on the car that doesn't make it faster (or isn't required by the rules, of course!).

One unintended consequence of 'fail-safe' measure is that a an innocuous sensor failure (or a failure of its wiring or connector), that would simply be noted in a datalog and fixed later, now becomes a 'show stopper', and it can sometimes be difficult to know why the vehicle has slowed or stopped. Even GM didn't put low oil pressure engine disabling sensors in place - instead they use oil pressure sensors to make sure the fuel pump relay is on when the pressure is high as a secondary measure (i.e. it makes sure something is on under a couple of conditions, rather than shutting it off if there's a problem).

However, I believe you can make an indicator in TunerStudio using the gear ratios, current gear, and ISS/VSS teeth. B&G code has a way for the trans code to add/subtract fuel or timing by writing to certain bytes in memory in the engine's controller. I do not know if the extra code has implemented these bits, though

<edit>Apparently it does, I just checked here: http://www.msgpio.com/manuals/mshift/V22tune.html#oa - so I could potentially put in an option to allow the slip to affect fuel/spark - you would have to let me know what behavior you would prefer, such as how much slip would be necessary action is taken, whether it monitor slip only while in gear or also if shifting, what action to take, etc.</edit>.

In any case, it would be better if the engine controller calculated the slip (which it can do because it can get the entire outpc and calculate slip itself, just like TS) and used a more sophisticated algorithm to limit boost/engine rpm/timing/whatever. So that would be a job you could ask the extra developers to look at.

Lance.

[tobmag]

Lance.

I understand the issues with the ISS/VSS signals but that might be better in the MicroTCU?

My idea was basically to have a "look-up" window directly after the gear change since as you said then slip is most possible to happend.
Today we allready lower PC at gear shifting and we could then trigger the window after x millisek after pressure has returned according to table.
Even so it will be very tricky to get this to work for all transmissions since they will have different cluts clearnace giving different applying times..
Also different oils, friction material etc will make a part as well... so for how long can we accept the slip or more correct how long does it take to change gear??
To get it right it has to be somthing self adapting....(ZF does this today with their adaptive controllers.)

I'm quite sure the extra guys can do something regarding this but giving this a second thought we might be adding more possible errors that arent really there.

BR//Tobmag

[Bernard Fife]

Tobmag,

What I can do for a start is add a user option to report/log the internal slip (between ISS and VSS) instead of torque converter slip (between the engine/tach and ISS). At very least, this will let us get a handle on how much slip is acceptable in different circumstances. I will put this in the next release (4.110, due shortly - check in the 'Latest Code Updates' thread here: viewtopic.php?f=4&t=240).

Lance.