4L60E solenoids always on

[turbolambo]

i have a strange problem with my GPIO board. The 2 shift solenoids are always on even if the screen in tuner studio says it should be in second or 3rd gear. I have checked the tip 120 transistors and they are fine but i am getting a 5 volt sig from the processor. I have reloaded the code a couple of times and checked the msq several times. Please check the attached msq and give me your thoughts. Thanks Greg






please

[Bernard Fife]

tl,

Are you testing on a bench or in a car? I have loaded your msq on my bench setup. You have 'enforce rpm checking' enabled. This is preventing shifts (on the bench at least, I am not sure how you testing). On my bench, your settings force the solenoids to be 'always on' because that what you has specified for first gear, and the trans never shifts out of first (because of the rpm checking).

When I disabled rpm checking, the code shifts appropriately with this msq, and the solenoids switch on and off as set in the 'Shift Output Patterns'. That rules out a code issue.

If you could post a datalog as well as your msq, that would be helpful (it would tell us if the code is commanding the solenoids to change, and if not, why not). I would also re-download the code from the site, and reload it onto the GPIO board in case it was corrupted anywhere in transit.

In the car, rpm checking enable is highly recommended (and is tested and works as designed), but on the bench it should be turned off because the relationship between the speed and rpm is often not appropriate for the shifts to occur. If you are testing, I would turn it off until you get this sorted.

Other possible issues are a short in the PCB, or wiring. If this was mine, I would investigate this by removing the solA and solB jumpers at the 25x2 header. That should remove the 5V signal to the TIP120. If the solenoids stay on, you have a short 'downstream' of the header. The jumpers are:

- processor PE4/ circuit VB3/ Amp pin 23
- processor PM2/ circuit VB4/ Amp pin 35

I would also make sure that the solenoids are connected to the correct ampseal pins, and that the harness side of these pins are not grounded when it is not connected to the GPIO (which would indicate a wiring short).

But before doing any of that, if you have not thoroughly cleaned the PCB using 99% isopropyl alcohol (or equivalent), you should do that first.

I would also check to make sure that the processor pins for PE4 and PM2 were not shorted to adjacent pins with excess solder during manufacture. PE4 is pin #14 on the processor, PM2 is pin #43 (pin #1 is on the side of the CPU nearest the edge of the PCB, and is the one nearest the boot jumper header. The pins are numbered counterclockwise around the chip from #1).



Lance.

[turbolambo]

I was checking the car on a hoist with the engine running.
I pulled the jumpers and the solenoids go off so it is from the jumpers to the processor.
In tuner studio it is showing the ecu shifting gears , 1 to 2 to 3 but both solenoids stay on.
I just redownloaded the firmware before the datalog.
Here is a log i just did.
Thanks for your help. Greg

[Bernard Fife]

Greg,

The sol states always show 3. 3 means sol A on, sol B on, and any others off. This is here: http://www.msgpio.com/manuals/mshift/V2tune.html#d So there is no shift expected to happen, according to the controller.

The tGear value (which is the target gear) show potential shifts to 2nd, and even 3rd briefly. But cGear (the current gear) is always one, so the shifts are being blocked. Shifts can be blocked for a few reasons - over-rev, under-rev, hysteresis not met, etc.

In the highest speed in your log (around 418 seconds) the speed is 47.2 mph, and the target gear is 2. The rpm at this point is 4055, but dropping rapidly (even though you are still going 45 mph ten seconds later, the rpm, has dropped to 984 rpm).

With the lower rpm limit of 1500 and the 1st/2nd gear ratios of 3.06/1.63 you would need an rpm of at least 4055*1.63/3.06 = 2145 rpm is the lower setting. Yours is set to 1500, so it theory it's fine as long as the pre-shift rpm is greater than 2815, but the rpm in your log has dropped below this value within 0.6 second.

On a hoist everything is changing so rapidly that the code might reject the shift because the rpm has dropped by the time the load has caught up (the request for a shift happens over a few seconds, and the rpm is only checked once the shift is commanded). Things like load are averaged so that the trans won't try to shift 10 times a second if the inputs vary slightly or there is noise in the signals.

You can reduce the averaging, but I think this will cause a fair number of issues once you get off the hoist and onto the road.

In a real car, the rpm would stay relatively high at a constant 45 mph in first gear, and the shift would happen as expected with the current setup - it does on my bench.

If this was mine, and it was possible (car safely assembled, licensed, etc.) I would road test it, I think it will shift just fine. If you want to continue testing on the hoist, disable the rpm checking (http://www.msgpio.com/manuals/mshift/V2tune.html#gr) and send another datalog for us to have a look at.

Lance.

[turbolambo]

I am sorry i was driving the car on the street and the rpm limits were off.
Here is another log i just did and the msq.
Trans stays in first gear and both solenoids stay on. TS shows it shifting into second.
Also the trans will freewheel at speed if i let off the gas , Thats why it showed 40mph and idle speed. This is normal for the 4L60E.
Thanks for the time Lance

[Bernard Fife]

Greg,

The log never shows the controller shifting out of first. The current gear is always 1 in your log. Don't confuse this with target gear, which is the gear you would be in if the gear table was being used (but it isn't always used, because rev limiting or hysteresis may prevent shifts, or it might be in manual mode). You are always in auto mode in the log, however, so something else is blocking the shifts.

I have loaded your latest MSQ on my bench, where it shifts fine (i.e. the current gear increments/decrements, and the solenoid outputs follow).

You have 2 for your number of VSS teeth - I believe this should be 40 for the 4L60E. Having that wrong can cause all sorts of problems.

You are also getting some error 25 in your code, this is 16+8+1 = VSS errors/resets, and quite a few error 3, which are dropped CAN comms. However, you have changed the VSS max error setting to 5, and I would reset the VSS error max back to 100 (or higher), as the resets can impede the function of the controller (after a reset, lots of things have to start counting again, so it's better to have errors accumulating than resetting all the time).

What I would do is re-load the code and use the stock settings for now (or, slightly less preferable - use the default msq below). The defaults are set up for a 4L60E. Don't change anything, just give it a cautious try.

Here is the default settings for 2.107:

default_2107.msq

Have you tried using the shift buttons at all? You should be able to shift regardless of the load or speed and see the trans go through the gears (if rpm checking is disabled it should always shift).

If the above tips don't help, I will send you some custom code that has more diagnostics enabled so we can get to the bottom of this. It may be that some of the MS-extra CAN comms are interfering with the MShift code (and we can fix this).

Lance.

[Bernard Fife]

Greg,
In looking closer at your logs, something else is going on too. When you are at 0% throttle in the engine controller, the trans controller is seeing this as wide open throttle, and the WOT flag acts much like hysteresis (on the assumption that if you just hit WOT, you probably want the lowest gear a possible for a while).

It looks like the throttle value passed from the msextra is wrapping under (so small negative numbers become large positive numbers), and the trans controller sees it as a large throttle opening. I had previously corrected this for users using engine TPS as the load, but I didn't correct the WOT calcs when you specify MAP for load. I will do this shortly and release new code as 2.108 (likely tomorrow).

In the meantime, you can recalibrate the engine controller so that the TPS never goes negative.

Lance.

[Bernard Fife]

I have fix for the TPS issue (at least it works with negative TPS values using B&G code on my bench). You can get the fixed 2.108 code and INI here in a few minutes: http://www.msgpio.com/manuals/mshift/V2code.html

If you want me to look further into this, let me know how the msextra code you are using on the engine controller reports the TPS. There will be a line in the INI like this:

Code: Select all

   tps              = scalar, S16,   24, "%",     0.100, 0.0

and I need to know what that line says. In the case above (B&G code) it is telling us that the engine is reporting the TPS in x10 values (so 78.9 is reported as 789) and in percent (%) rather than ADC counts, or some other unit. It can be negative (it is signed = S) and it is 16 bit long (=2 bytes = -32768 to +32767). It is located at offset 24 in the outpc table.

Also, note that you probably can't use the 'Get only required vars' CAN setting with the extra code (http://www.msgpio.com/manuals/mshift/V2tune.html#oa), you likely have to use 'Get all vars' unless they have updated the extra code recently.

Lance.

[Bernard Fife]

I will be releasing a new 2.109 code/ini shortly, this will give users the ability to disable the WOT hysteresis. It will be disabled by default (a setting of 0 disables the WOT flags), the previous value was 10 seconds. The INI allows values from 0 to 20.

The code will be released here: viewtopic.php?f=4&t=20&p=1976#p1976

Lance.

[turbolambo]

Ok i checked everything and took it for a drive and low and behold it shifted. did a lot of adjusting shift points and firmness and called it a night. Next day drove it out of the shop and worked fine. made some changes later in the day and buttoned up the ecu,wiring ,ect and went for a drive and just first gear. So went through anything i could of changed and all looks fine. But i noticed the screen showed the upshift button is pressed and it is in manuel mode. I have no buttons or wiring for that matter wired up. I did build the circuts but have not used them.
If i turn off the car the next time i start it it might be in man mode or not.When this happens it is in park and auto mode , i shift into drive and man mode pops up and no shift , Or it works perfect. Upshift button always shows pressed even when it is showing auto mode and working. So how can i disable this or check the upshift circut and repair.
Thanks again Lance.