GPRP Shift System

[Matteo Tagliaferri]

Sorry. This is the run on a straight.

https://drive.google.com/file/d/1aCCdF0 ... sp=sharing

This is the poin in which the Source from the sensor becomes the estimate, that is not correct because for the estimate the gear is the second.

https://drive.google.com/file/d/1rw4leg ... sp=sharing

As you can see here, the wheel speed and the gps speed are more or less the same.
So i really can't figure out where the problem is, given that the gear estimated ratio differs from the gear ratio of +0,025 in second, third and fourth. So even if this difference is small, why the estimate still presents a wrong gear?

The followings are the values set up for the Gear Position Sensor.
Could the problem be the fact that at 95% (the percentage in which the source from sensor becomes the estimate) is set the Reverse gear and at 100% the fourth?

https://drive.google.com/file/d/1sK2tGd ... sp=sharing

[David Ferguson]

OK, I have looked at your logged data file posted in the first post in this thread.

Your wheel speeds are off by about 3.5$%, which is not insignificant. You are obviously geared very short and don't have long straights to get the calibration, so you are still accelerating at the end of the straight. You can see the GPS offset, so I used the max GPS speed and the Max non-driven wheel speed to calculate the difference.

I suggest you change the Wheel Speed Front Circumference from 50.709in to 48.91in to account for this difference. Since the Wheel Speed Rear Circumference is the same, I would also change that to 48.91 in. (multiply by .9647 if you are using different units).

Next, looking at your M1 Config, I am surprised to find that your Reverse Gear is located between 3rd and 4th Gear, according to the Gear Position Sensor calibration. This doesn't seem correct.

Do you really have a reverse gear? The Gear Minimum is set to Neutral not Reverse as I would expect if you had a Reverse gear.

Tell us about your Gear Position Sensor -- looking at a graph of the Gear Position Sensor, you see when it can't quite get into the next gear (either going up or down). The voltages seem clean. I see at 18 minutes into your log file, the Gear Position channel change to 95% several times as you attempt to downshift.

I notice that Gear Position Tolerence is 2% for everything except 3rd and 4th gear, where it's 5% and 6%. I suggest you keep it at 2% for all gears, so that it only selects the gear when it's actually in the gear with dogs engaged.

Now your Gear Estimate Tolerance is 0.5 ratio, but there is only a .25 ratio difference between 3rd and 4th gears. I believe the Gear Estimate never detects 4th gear because of this large tolerance. I would suggest you set the Gear Estimate Tolerance to 0.120, so the correct gear is estimated.

I suggest you set the Gear Estimate delay to a shorter value, like 100 - 125 ms. Only increase this if you see bad estimates that get corrected if you had waited longer.

It's good that you are logging all of these channels at 100hz (and have pro-analysis -- thanks!).

Try these changes and post up a log of your results.

[Matteo Tagliaferri]

Thank you for the time you spent on analysing my log file.

I don't have reverse gear, for this reason i asked you if it is correct to have its percentage between that of the third and that of the fourth (set up done by another guy). Minimum gear is actually the neutral.

For the gear position tolerance i tried in the workshop to decrease the tolerance and below the 5% and 6% for third and fourth they can't be detected when manually engaged. This is the reason why the third and fourth have 5% and 6%.

Could you tell me what infos you need about the gear position sensor please and i will post them.

I have already calculated the value of Gear Estimate Tolerance and will do a run the day after tomorrow with this value set up correcly and do other runs with the other values you adviced me and i will post the results. Thank you again!

P.s. Pro-analysis it's almost a must when you have to keep under control several channels! It's the first time i use MoTeC and i'm really satisfied with the type of analysis that let you have!

[David Ferguson]

I would use 0% for Reverse just like 5th & 6th. I was only asking about the gear sensor thinking it wasn't clocked right on the gear box or something. No need to provide further info on that.

Please read the help for Gear Estimate Tolerence -- I think that is the main issue.

[Matteo Tagliaferri]

Alright, thank you.

I'll fix these values, get back on track and test them.

Hope to solve this problem!

I'll update you in the next days with new log file.

[Matteo Tagliaferri]

Hi David,

Went back on track and test out the changes: the problem between the third and fourth gears (the second detection) seems to be solved. Moreover the source for Gear is always the sensor, that is correct. This is the log file:

https://drive.google.com/drive/folders/ ... sp=sharing

Now that the things are correctly set up i would like to know in which ways i can optimize the shift process in terms of timing. For this i was asking in the first post, if was correct that the effective engagement of next gear (for example for upshifts) happens in the recover or in the rearm phase, since having read MoTeC GPRP guide i understood that the effective engagement has to mark the end of the shift phase (in the worst case).

May this can be related to other setup parameters or just to the timing of the actuator phases and shift?

[David Ferguson]

Looking at your data, I see the Gear Estimate is much better. The Gear Estimate Ratio still has some noise in it, because the wheelspeed data has noise. I think perhaps your wheelspeed triggers are not quite evenly spaced. My suggestion, change the "Wheel Speed <corner> Sensor Sample Teeth" to something like 4 or 6. I was able to filter the data in I2 (using filter, samples 5), and you will get a much better idea of what is happening with the speeds when you shift.
I think you could change "Gear Estimate Delay" to 25 ms now, and "Gear Estimate" would match "Gear" nearly all the time.

Since the shift system is closed loop, there is no disadvantage to making your Gear Shift Timing in the "shift" phase to be much larger (particularly the Upshift Power On). I suggest you set these to 150 - 200 ms from the 40-70 ms you currently have. The shift phase will automatically end when the shift completes. Looks like it takes about 110 ms for your 1->2 shift, 120ms for 2->3, and 3->4 shifts.

Your engine speed is not dropping very fast despite having a 100% ignition cut -- is your Hybrid motor providing torque while the shift is occurring?

I would play with "Gear Shift Ignition Timing Retard" (increase this to soften upshift harshness), along with increasing the "Post Shift" and / or "Recover" timing. Look at your wheelspeed data and engine speed data to tune these values -- driver feedback is important as well.

I would lower the timing for Gear Shift Actuator Timing for Preload Power On (and OFF) 5-10 ms may be ok for both.

Looks like you are on your way.. Good Luck!