aidandj

iTrader: (5)

It finally started raining (never thought i'd say that in oregon) so I can play with traction control tuning without destroying tires.

A couple things I noticed.

1. I have no clue what I'm doing. I can't get my head wrapped around slip% x time.
2. Pulling timing seemed to do nothing. I actually couldn't get it to pull more than 10*.
3. Spark cut actually worked really well. I did 25% spark cut and it stopped wheel spin perfectly. But this is pretty hard on components and stuff. It actually recovered really well.

@Ben any chance that you have some example reaction tables laying around that I could look at. Interested in the shape of them and how you tune it?

Also, how can I view spark cut in the log, it doesn't show up in spark advance or anything.?

psyber_0ptix

iTrader: (4)

Just reviving this for funsies. Does the following link provide any new insight?


https://www.diyautotune.com/news/tun...raction-world/



Also, on an NA6 shell, without ABS, could one simply reference the driven wheel speed off the gauge cluster like your tutorial (https://www.miataturbo.net/megasquir...6-miata-82518/) but then get a 10Hz GPS sender for chassis speed, or would this not refresh fast enough.

Some of us have zero facility to accommodate an abs sensor. Or just knuckle swap...

acedeuce802

iTrader: (1)

I don't know if it's actually been tested, but it's been discussed on msextra.com and the consensus is that it won't refresh fast enough.

shlammed

iTrader: (8)

put sensors on the hubs.

doesn't have to be an abs signal AFAIK. some people use sensors to the back of a wheel stud or something else that rotates and meets the space conditions for the sensor. I've seen multi tooth pickup wheels between the flanges of a driveshaft for driven wheel speed too if the cluster option doesn't work well. ideally you would use 4 wheel speed sensors and allow the ecu to use an average for driven wheel speed and an average for non-driven wheel speed.

psyber_0ptix

iTrader: (4)

Would one just use a reluctor ring on the driveshaft to get the average of the two rear wheels taking the final drive into account?

4 pulses per revolution is high enough resolution if taken from back if wheel studs?

Does megasquirt even accommodate more than two speed sensors other than for logging?

I'll search a bit more after work, I can't access some things at work nor can do I have the ms3x wiring handy.

acedeuce802

iTrader: (1)

Driveshaft would be average rear wheel speed times final drive, yes.

4 p/rev is not enough resolution for low speed stuff. It equates to about 1Hz/mph, so at 40 mph it will be about 40 Hz which is likely enough resolution. At 5 mph, not so much.

Yes, you can select whether the input for traction control is a single wheel speed, averaged, or highest wheel speed between the two.

shlammed

iTrader: (8)

I cant comment on megasquirt.

I have been doing all of my system design and configuration for LINK Thunder ECU.

psyber_0ptix

iTrader: (4)

Thanks! So as long as one has enough digital inputs, it will see 4 wheel sensors? I need to revisit my harness to see how much free i/o I have.

aidandj

iTrader: (5)

You need 4 VR inputs. Or 4 conditioned inputs into digital inputs. Megasquirt can't handle 4 individual wheel speed sensors independently right now, but it can do things like average, minimum, maximum, etc.

Its all in the manual/tunerstudio.

acedeuce802

iTrader: (1)

I had thought about using a ground speed sensor, but never really put in the effort to find one that's affordable. You could put it in the center of the vehicle, so you only need 2 inputs, the GSS and driveshaft speed sensor.

Something like this: http://ph.parker.com/us/en/tgss-true...ors/0740059ecd

Not sure how much that one costs or details on it (it only goes up to 44 mph so that specific one isn't useful), but it'd require a lot less effort than swapping hubs/WSS or adding a custom ring/sensor.

psyber_0ptix

iTrader: (4)

But using the NA6 gauge cluster to output into [Cam in/Datalog in/ Tableswitch in/Launch in] would already average the rear; then one could use PT4 and or another input for either or both the front wheels for Tuner Studio to average? Sorry for all the questions. I'm reading the manuals now but my laptop stopped working, so I have no access to tuner studio at the moment. Is the RSW not populated on a pin in the plug behind the cluster (from your other tutorial)?

aidandj

iTrader: (5)

My NA6 output and ABS speeds read differently. Not sure why, maybe slop in the speedo cable

psyber_0ptix

iTrader: (4)

Understood; thanks. I got all excited. Now I'm sad.

aidandj

iTrader: (5)

You have an osg in the back right? Should keep side to side slip pretty solid. Do one and sensor on each end.

psyber_0ptix

iTrader: (4)

I'm flying up to MA this weekend for work, selling a clutch and talking to a dude and possibly scoring some NB knuckles up there. TSA is going to love checking my bags. if only referencing one side, will traction control be accidentally triggered while turning if it's not averaging across sides on the respective fore/aft end?

I don't know if I have very enough inputs free but I'll have to double check. I haven't touched the car for so long due to travel; i'm forgetting all the things.

EDIT: I just found the page on additional I/O internally; hence the reference to JS10 above

acedeuce802

iTrader: (1)

A Miata has a minimum curb-to-curb of 30.8' and a track width of 4.6', which means the center of the car has a turn radius of 28.5'. The driveshaft will be the average speed, which means it equals the speed at the center of the vehicle. So the percent difference is between 28.5 and 30.8 is 7.5%. Traction control strategies often start at 7-8%. Traction control is usually turned off under a certain speed, a speed that's much higher than what you could do curb-to-curb at, so when traction control is active you'll always be at lower than 7.5% difference. That means traction control will never come on just from turning, but it does mean that you are falsely closer to your activation slip ratio (or further away). The tightest corner at my local track is 60' (which outside-inside-outside you'd be at a largest radius, but we'll do 60' for worst case), you're down to a 3.5% difference. That means if you tune to hold 10% slip, MS may actually be targeting 13.5% while turning one direction while 6.5% the other way.

OCD Engineer aside, in my opinion, a single front wheel is acceptable if your goal for traction control is to prevent "OH ****" moments or for drag racing. Utilizing traction control to maintain slip ratio through a corner will absolutely need an average L+R signal, otherwise the amount of grip/torque will be unpredictable corner to corner.

Leafy

iTrader: (1)

I have a nice little box wired up with 4 vr conditioners on it for my car. It needs to become part of the car and I need to build a case for it. Interested in figuring this out. And like Andrew I'd like the rolling cylinder fuel cut, I know spark retard isnt enough and I worry about sound restriction issues with spark cut.

psyber_0ptix

iTrader: (4)

Dude, thanks for laying it out. I've lost all my engineering working for the government. Makes me wish I had ABS as well.

shlammed

iTrader: (8)

doing average front or average rear you will still need two sensors to be able to average. otherwise its just a driven wheel speed.

configuring it as an average doesn't help if you only have one sensor. if you need an ECU with 4 wheel speed sensors, lmk. I have sold my project car and no longer need the LINK G4+ thunder which is leaps and bounds better than most ecu used in standard Miata.

Matt Cramer

Current MS3 firmware does allow for four separate VSS inputs. However, a driveshaft speed sensor will always return an average of both rear wheel speeds since the diff averages the speeds mechanically. There is one mode, "side 2 side," that compares slip on each side and would require sensors on all four wheels. For a car grippy enough to lift an inside front tire, though, using the higher of the two front wheel speeds plus the differential speed is usually your best bet.