Chrysler 41te Trans Control

A forum for discussing applications and implementations of the MegaShift transmission controller code for the GPIO from B&G. This can control up to 8-speeds and 6 shift solenoids (plus a 16x9 table for controlling a PWM line pressure valve). It has manual and fully automatic modes (16x9 load x speed table), with under and over rev-limit protection, and full data logging of all inputs and outputs (among many other abilities). A TransStim to test your completed board is also available.
Post Reply
tjabo
Posts: 68
Joined: Mon Jan 12, 2009 10:40 am

Re: Chrysler 41te Trans Control

Post by tjabo »

Alright Steve, assuming that I was right (above) about what you needed to know, here is how I see it so far. Hopefully Lance will correct me as needed.

First, you will need to provide the MShift a TPS signal. As we know from MSquirt installs on Neons, you can tap into the factory TPS line and have it shared between the PCM and the MShift (or MSquirt as we usually do).

Second, you need to give the MShift a load signal, since it won't be able to get the MAP information via CAN from the MSquirt. There is a circuit on the board that can be used optionally for either EGT1, or non-CAN MAP. This circuit is said to be a voltage sensing circuit for 0-5V, so maybe since the trans shouldn't really care about boost, you can simply tap into the factory MAP wire if you can figure out a voltage to kPa curve.

I'm guessing the FIC has it's own on-board MAP sensor that is boost capable, so maybe it would be more appropriate to take your MAP signal from the output wire from the FIC to the MAP input wire into the PCM?

As far as RPM, I'm guessing this is needed, even though you might be able to extrapolate to your shift points from VSS input. That doesn't sound very safe or precise to me, so I'm thinking the easiest way to go is to tap into a "tach signal" however people did it for MSquirt installs prior to the Neon/420a crankwheel decoder. You probably know exactly how that was done unless I missed my guess on that.

I **think** that's it!

Thad

EDIT: This makes me think I might be wrong about a couple of items in my post:
Lance wrote:There is one dedicated line pressure input, and more free ADC (voltage measuring) inputs that could be used simultaneously for more pressure sensors, depending on things like the shift lever signal (for example, if the shift lever is voltage based rather than digital, there are two more ADC voltage channels available in the current code). There is also an ADC channel freed up if using CAN (w/ MS-II) to get the load (MAP or TPS).
So for Steve's purposes, potentially using the AEM FIC to run his engine might limit his ability to datalog pressures from within the trans, since that ADC input would be tied up by the MAP or TPS.

Which brings me to my next potential "wrongness". This makes it look like an either/or thing as opposed to a "both" thing. . . . ?

Another EDIT:

After looking some more at Lance's prior posts in this thread, and looking around elsewhere, I think maybe the RPM input was only in my mind. :roll:
Turbo Neon, converting to 41te trans sometime?
Bernard Fife
Posts: 1696
Joined: Fri Apr 04, 2008 1:28 pm

Re: Chrysler 41te Trans Control

Post by Bernard Fife »

Thad,

Close! For operation of MShift without MS-II, Steve needs a single external load signal of some sort. This can be TPS or MAP and MAF or VAF or whatever, as long as it's 0-5 Volts or less (higher voltage singles can be brought down to 0-5V with a simple two resistor voltage divider) and goes strictly up or down with load (aka. "monotonic").

Engine rpm isn't needed for 'standalone' MShift operation (remember shifting is based on load and vehicle speed).

However, by running independently of MS-II, you lose a number of features of the code (things like under and over-rev shift protection, error checking, etc.,). This is an inevitable consequence of having much less information. However, the features lost are likely more important for activities like autocross/road racing than drag racing, I suppose.

Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
tjabo
Posts: 68
Joined: Mon Jan 12, 2009 10:40 am

Re: Chrysler 41te Trans Control

Post by tjabo »

Thanks for clearing that up Lance.

For Steve's initial install to test/develop the basic functioning, I don't think he'll be too affected by the loss of those features.

However, if doing it this way costs him a pressure logging channel, that might be a problem. . .

Thad
Turbo Neon, converting to 41te trans sometime?
Bernard Fife
Posts: 1696
Joined: Fri Apr 04, 2008 1:28 pm

Re: Chrysler 41te Trans Control

Post by Bernard Fife »

Thad,

We should still have the dedicated line pressure sensor ADC channel. And I will work out something for the unused temp sensor ADC channel to free it up. However, because the shift lever position logic is digital, the additional two ADC channels aren't going to be available, so there would be a maximum of 2 pressure sensor channels. I recall that Steve said that 2 was the acceptable minimum for line pressure sensors.

In theory we could free up the CAN pins for I/O, but they are digital, not ADC, so aren't much use for most logging purposes.

Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
racerstev
Posts: 19
Joined: Thu Jan 15, 2009 2:46 pm

Re: Chrysler 41te Trans Control

Post by racerstev »

I'll find a tach output somehow, there's a few options here. I have a MSD DIS2 I could wire in, there seems to be a unconfimed tach wire on the factory PCM as well. Or maybe tap into one of the
coil pack wires and set the Mega Shift into 2 clyinder mode (???) As a last resort there's a signal in the data bus I can get to with a remote engine start adapter.. At any rate I need a tach in the
van anyway.

For the load signal I think the TPS would be a bad idea with such a small turbo, part throttle boost is going to be there to screw things up. I have a spare GM 3 bar I can pop on there.

The test "car" is a 2001 Chysler minivan with a 2.4liter turbo engine in it, power should be around 275-300hp to start with..

Steve
tjabo
Posts: 68
Joined: Mon Jan 12, 2009 10:40 am

Re: Chrysler 41te Trans Control

Post by tjabo »

Steve, if I understand Lance's last post, we don't need RPM for your MShift application with the FIC. Since you're going to find one for your tach anyway, are you thinking that you'd like to see if we could implement the under/overrev downshift protection and other stuff that needs a tach signal and that still may be possible without an MS2 for the MShift to communicate with through CAN, or doesn't that really matter to you? I can't see how you can go all that wrong with the gear by load and vehicle speed table I guess. . . .

Lance, Is there a way to use Steve's extra 3 bar MAP sensor, or does the MShift really want some sort of a MAP signal that is less "raw" (for lack of a proper term) than that? I'd think the FIC unit must be piggybacking the MAP signal to the engine to avoid CELs, and assuming Steve doesn't want to use lower shift speeds at low boost and higher ones at high boost--in effect using the turbocharger as a pure engine size imitator, then the MShift shouldn't need to see positive pressure, should it?

This is interesting stuff, and I'm really getting the impression we can work it out. There will be a lot of happy Neon owners, and maybe the turbo Dodge guys who've started to use MegaSquirt now that it reads their crank wheel will be benefited by it too? I'm not sure on bellhousings and such.

Thad

For the To Do List (or not-to-forget list): Need solenoid PWM frequency. . . Lance, did you say some multi-meters might be able to measure that? I might be able to check it on my wife's minivan if I knew how/had the right equipment.
Turbo Neon, converting to 41te trans sometime?
Bernard Fife
Posts: 1696
Joined: Fri Apr 04, 2008 1:28 pm

Re: Chrysler 41te Trans Control

Post by Bernard Fife »

Thad,

You are 100% right on the rpm and MAP. As long as you want the same pressure about the max limit of the MAP sensor output (and it is a steadily increasing or decreasing voltage between 0 and 5 Volts), then it ought to work. Most MAP sensors work this way (it's usually the cheapest way to do it, and result from the piezoelectric crystal used at the heart of the sensor). It is also possible to grab the signal from an existing MAP sensor (unlike temp sensors, these can be shared between a processors - within reason).

The one danger with a speed by load shift table is that you have to get the calculations right to avoid over-revving the engine. With the CAN enabled, the engine will shift anyhow - because of the over-rev protection. But without that protection, the table has to be right. I plan to make a little javascript calculator that takes the gear ratios, tire sizes, etc. and spits out a table (when I have a bit of spare time).

Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
Bernard Fife
Posts: 1696
Joined: Fri Apr 04, 2008 1:28 pm

Re: Chrysler 41te Trans Control

Post by Bernard Fife »

Thad,

I forgot to add:

If your DMM has a Hz symbol (short for Hertz == cycles per second), then it can measure frequencies. This is a very handy feature in a DMM, and is often found on better ones, but not usually on the really cheap ones. I''m not sure where the price 'threshold' is though.

However, if you build your GPIO on the bench, then you might be able to use the VSS_Hertz gauge to measure this frequency - all you'd need is power (12V), ground (for the power and the signal), and then use Ampseal pin #2 as a probe to get the frequency (though it might get a little confused with the 8 millisecond 'refresher' events).

And I have thought of a GPIO based 'scope/voltmeter/function generator' - it should be a fairly straightforward project (especially since Phil's excellent MLV could be used in 'trail live file' mode as a display (the hardest part of the job, really). First things first, though!

Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
racerstev
Posts: 19
Joined: Thu Jan 15, 2009 2:46 pm

Re: Chrysler 41te Trans Control

Post by racerstev »

Is there a tach input on the modual or does it just come in though the CAN?

So there is no "full throttle" mode to shift off of rpm only? This trans has a pretty bad shift delay
on the 2-3 shift, we will need to lead the shift by .4 sec or so.
Matt Cramer
Posts: 55
Joined: Thu Apr 17, 2008 5:19 am

Re: Chrysler 41te Trans Control

Post by Matt Cramer »

Lance wrote:Thad,

I forgot to add:

If your DMM has a Hz symbol (short for Hertz == cycles per second), then it can measure frequencies. This is a very handy feature in a DMM, and is often found on better ones, but not usually on the really cheap ones. I''m not sure where the price 'threshold' is though.
I have a Triplett DMM that I picked up for around $35 at Fry's that can measure frequency. So this has started to trickle down to some rather affordable meters.
Post Reply