What is the shftBut field in datalogs?

[Bernard Fife]

Addicted,

What was the last code version you were using (there have been 5.099a through 5.099w). The last change to any of the speedo code functional snippets was quite a while ago; and there were no changes at all to the VSS/speedo section of the 5.099 code between the u and w versions (at least no intentional changes). The latest code does show the right speed for a given input signal on my bench using the default settings.

I can't say anything about errors in speed based on a log, as I can't see what the actual speed is supposed to be.

You might post an MSQ, and I could check to see if I get an unusual behavior in the speedo on my bench with your settings.

Lance.

[Addicted]

Ref speed:

2007rpm in third should be 32mph. The datalog shows 36.2mph

1887rpm in fourth should be 44.3mph. The datalog shows 69.2mph

The previous code I was using was 5099s

One of my previous posts has the datalog and msq from Sunday where everything was spot on...

Martin

[Bernard Fife]

Addicted,

I had a look on the bench. With the *default* settings and a VSS input of 1583 Hz, I get 60 mph (as always).

With *your* settings loaded and the power cycled, and the same 1583 Hz VSS input, I get a speed of 47.6 mph, which is the expected value for that input. (I calculate your VSS rate at 119951 pulse/mile; 1582*3600/119951 = 47.5 mph, which is within the margins of error of the calcs and the rounding of the code.) I did this with both the 5.099u and 5.099W codes with the same results.

However, please note that if I just load your MSQ *without* cycling the power, I get 73.7 mph at 1583 Hz on the VSS input, and perhaps that's what you are seeing. If that's so, you have to remember to reset the power after changing the VSS setup parameters (even if these VSS changes are part of a larger MSQ rather than edited in TS) in order for the calculations to work out.

If that's not so (i.e. you did cycle the power after loading the settings) then I am not seeing your issue, and we will have to dig a little deeper.

Lance.

[Addicted]

Hi Lance,

Due to extreme timezone differences between us I won't be able to have a go until tonight (GMT)

If its simply because I didn't cycle the controller I will be extremely dissapointed with myself..

Martin

[Addicted]

Oh the shame..

works perfectly! datalog attached to prove (GPS speed included to compare)

Just needs tuning now.

What setting should I alter to smooth avload? or limit its rate of change? There's a point on the log where, when in fourth gear, the avload instantaniously jumps from 60kpa to 70.5kpa which causes a downshift. Is that the load smoothing factor?

Also there's an area on the log where TCC cycles on and off in third gear where I cannot see any criteria for TCC to be engaged at all, any thoughts Lance?

Martin

[Bernard Fife]

Martin,

Yes, see the 'Load Smoothing Factor' description here (just updated): http://www.msgpio.com/manuals/mshift/V5tune.html#gh Note that increases in load are instant, decreases are smoothed. This makes the trans sufficiently responsive to 'flooring the throttle'.

I didn't find an area in the most recent log (2015-03-10_18.06.34.msl) where the TCC cycles on and off. Can you point it out to me?

Lance.

[Addicted]

The reason you didn't see it is because I didn't attach it!!

[Bernard Fife]

Ah, that's good; I thought my eyes were failing me!

[Bernard Fife]

Addicted,

I don't see any input in your datalog that has caused the TCC to cycle, so I suspect I have messed up something in the code when implementing the full-time TCC in 4th. I will have a look and post fixed code here when I have a solution (it may have to do with how the hysteresis is handled, as the fluctuations happen shortly after the hysteresis is released in 3rd gear - but then there is no hysteresis at all for a few seconds and the TCC oscillates).

Lance.

[Bernard Fife]

Addicted,

I found a small problem in the code (relating to the indeterminate referencing of the current gear during a shift) and have fixed it in the attached code. I wasn't able to replicate your particular issue's behavior on my bench (though I have no doubt it exists in your datalog, of course), but the problem I did find could conceivably be the cause. If you are willing to give it a try, that would be great. I will keep working on the code here in case there is something else at the root of this issue.

Monitor_5099x.abs.s19

(82.91 KiB) Downloaded 508 times

GPIO_MShift_5099.ini

(277.7 KiB) Downloaded 514 times

Lance.