Full time lock up mode

[Bernard Fife]

All,

The fixed 2.00Ex code is here: http://www.msgpio.com/manuals/mshift/fi ... Ex.abs.s19

The problem was that the speed (102 mph) is in 1/10th, and multiply this by 64 gives 1020*64 = 65280, and much above this we hit the wrap-around limit of 65535 for an unsigned int. The new code fixes this (the calculated values were scaled), and appears to work as planned in brief testing.

The default VSS smoothing factor was also lowered to 8, and the range restricted in the INI to be from 2 to 12.

Lance.

[gurov]

TCC pwm is not being reported anymore with Ey code. i see little spikes, and it goes to 100% in non-pwm mode.

with vss smoothing of 4-12, i could not get any speed readings in 1st gear, so it would never shift out of 1st gear below redline.

[Bernard Fife]

gurov,

I will have a look.

Lance.

[gurov]

attaching some datalogs of this happening.

[Bernard Fife]

gurov,

Yes, it's pretty obvious! I had 'cleaned up' some parts of the TCC code (without changing the logic at all, I thought), but the compiler doesn't like the way I 'clean' apparently.

Attached is 2.00Ez code:

Monitor_200Ez.abs.s19

It appears to work properly in my quick bench tests. It uses the same INI, but I will be releasing a new one shortly that opens up the VSS mask to 200 again (as before, this should help your low speed issue). The line is like this:

Code: Select all

vss_mask              = scalar,  U08,   548,                "%",         1.00,    0.00000,     0.0   99.00,        0    ; VSS interrupt mask (%)

but it will become:

Code: Select all

vss_mask              = scalar,  U08,   548,                "%",         1.00,    0.00000,     0.0   200,        0    ; VSS interrupt mask (%)

Also, make sure your minimum speed is set fairly low.

I will do some testing here, and let you know if I find any issues.

Lance.

[gurov]

gurov,

Yes, it's pretty obvious! I had 'cleaned up' some parts of the TCC code (without changing the logic at all, I thought), but the compiler doesn't like the way I 'clean' apparently.

Attached is 2.00Ez code:

Monitor_200Ez.abs.s19

It appears to work properly in my quick bench tests. It uses the same INI, but I will be releasing a new one shortly that opens up the VSS mask to 200 again (as before, this should help your low speed issue). The line is like this:

Code: Select all

vss_mask              = scalar,  U08,   548,                "%",         1.00,    0.00000,     0.0   99.00,        0    ; VSS interrupt mask (%)

but it will become:

Code: Select all

vss_mask              = scalar,  U08,   548,                "%",         1.00,    0.00000,     0.0   200,        0    ; VSS interrupt mask (%)

Also, make sure your minimum speed is set fairly low.

I will do some testing here, and let you know if I find any issues.

Lance.

my vss mask was still 200 in these logs as i change that by hand still. any idea on why speed did not get picked up in 1st gear with smoothing 4 and higher ?

[Bernard Fife]

any idea on why speed did not get picked up in 1st gear with smoothing 4 and higher ?

gurov,

I believe this is because you have your minimum speed set to 10 mph. With averaging it take a while for the speed to catch up, and if the gap is large it can stick at 0 for a while. I would try a minimum speed of 3.0 mph (the default), or maybe even less.

However, I don't think the restricted vss mask was helping in your case. The vss_mask was limited in the 2.00Ey code to 99, even if it was 200 in the msq, etc. The new code re-activates the 200 setting (internally to the code, and the INI is a separate issue). Values above 200 cannot be used because they lead to negative values in the calculations, which will then wrap around. However, 200 is almost no filtering at all, anyhow.

So your low speed mph should work as before once you load the 2.00Ez code.

Lance.

[gurov]

any idea on why speed did not get picked up in 1st gear with smoothing 4 and higher ?

gurov,

I believe this is because you have your minimum speed set to 10 mph. With averaging it take a while for the speed to catch up, and if the gap is large it can stick at 0 for a while. I would try a minimum speed of 3.0 mph (the default), or maybe even less.

However, I don't think the restricted vss mask was helping in your case. The vss_mask was limited in the 2.00Ey code to 99, even if it was 200 in the msq, etc. The new code re-activates the 200 setting (internally to the code, and the INI is a separate issue). Values above 200 cannot be used because they lead to negative values in the calculations, which will then wrap around. However, 200 is almost no filtering at all, anyhow.

So your low speed mph should work as before once you load the 2.00Ez code.

Lance.

wait... 200 is almost no filtering at all ? i'll have to read through the past threads, sounds like.

minimum speed i can ever see is 10 mph. below that it sits at 0.

i got up to 25-ish mph with vss filtering set at 4 and the speed still read 0 until i shifted.

i'll try the new code tonight and see if it makes it better.

thank you.

[Bernard Fife]

gurov,

Yes, higher numbers mean less filtering. You specify the percent the signal is allowed to vary from the current average. This is covered here: http://www.msgpio.com/manuals/mshift/V2tune.html#gv

The code has a user setting for 'Min. Vehicle Speed' for the minimum speed that will register, below that the code considers you to be stopped. In your case, I believe you have this set to 10 mph. The default is 3.0, and I would set it to that or lower. You can read about it here: http://www.msgpio.com/manuals/mshift/V2tune.html#gv

Lance.

[gurov]

just tested Ez.

TCC does not unlock on UPSHIFTS. it unlocks on downshifts.