Mercedes-Benz 722.6 (NAG1) Automatic Transmission

[Bernard Fife]

Vesku,

Someone has sent me a file on the NAG1. The basic shift logic apparently looks like this:

NAG1 (722.6) shift logic

NAG1_shift_logic.jpg (140.19 KiB) Viewed 20801 times

The current code does not have the ability to 'pulse once' or 'toggle' the outputs. However, we do intend to support this trans if at all possible (it looks straightforward at first glance), and I certainly hope we can add that capability to a future code version fairly soon. The CAN shifter likely is a problem, but it may be possible to work around that (1996-1999 were not CAN bus enabled, apparently).

Lance.

[Bernard Fife]

After thinking about this, the 5.00X+ codes could be used with the 'clutch outputs'. Clutch outputs can be set as active (with user PWM settings) or inactive for the duration of a shift, then they turn off. When outputs are used as clutch outputs, they have two timing elements: a fixed line pressure delay, plus a load dependent shift completion delay. So those delays could be used to adjust the duration of the shift signal as required by the NAG trans.

There is a show-stopper in the code at the moment, though. All current codes determine the current gear by polling the output states. Since all the outputs would all be off after the shift period, the controller would not know what gear it was in after a shift was done.

However, I can easily implement a 'NAG mode', in which the controller would assume the current gear is the same as the last commanded gear. With that and the clutch outputs, I think we are within reach of NAG transmission control (other minor adjustments may be needed, such as for the allowable duration of the delays, etc.).

I will work on this for the next beta release code.

Lance.

[johnboy426]

After thinking about this, the 5.00X+ codes could be used with the 'clutch outputs'. Clutch outputs can be set as active (with user PWM settings) or inactive for the duration of a shift, then they turn off. When outputs are used as clutch outputs, they have two timing elements: a fixed line pressure delay, plus a load dependent shift completion delay. So those delays could be used to adjust the duration of the shift signal as required by the NAG trans.

There is a show-stopper in the code at the moment, though. All current codes determine the current gear by polling the output states. Since all the outputs would all be off after the shift period, the controller would not know what gear it was in after a shift was done.

However, I can easily implement a 'NAG mode', in which the controller would assume the current gear is the same as the last commanded gear. With that and the clutch outputs, I think we are within reach of NAG transmission control (other minor adjustments may be needed, such as for the allowable duration of the delays, etc.).

I will work on this for the next beta release code.

Lance.

This is great news! The factory controller "knows" what current gear is by monitoring the input/output speed sensors and an output (some are a direct output sensor while others are vehicle speed via CAN BUS i.e. wheel speed). Any speed error outside a certain percentage is seen as a ratio error and the transmission enters limp in mode turning off power to all solenoids which defaults the transmission to the current gear. Is there a way to have the code poll the input speed sensors and a wheel speed sensor(or some other type of speed) instead of the output states to make the 'NAG mode' possible?

[johnboy426]

After reading the discussion again, I realized there is more than one gearset for the NAG1. The first post lists:
1st Gear 3.93
2nd Gear 2.41
3rd Gear 1.49
4th Gear 1.00
5th Gear 0.83
Reverse Gear 1 -3.10
Reverse Gear 2 -1.90

Chrysler service information lists:
1st Gear 3.59
2nd Gear 2.19
3rd Gear 1.41
4th Gear 1.00
5th Gear .83
Reverse Gear 3.16

[Kma4444]

Hello all, I am new here and am wanting to use a NAG1 with a stand alone controller. Wonder if there has been any success in making one of these work.

Thanks a lot

[Bernard Fife]

Kma4444,

We are working on the NAG mode for MShift right now. There is a fair bit of work to do, but the basic shifting code is done. I hope we will have something for people to try before too long (hopefully within several weeks if all goes well).

Lance.

[Kma4444]

Fingers crossed that you have good luck with getting it up and running. I am to the stage of installing the NAG in my vehicle and would love to use this controller.

Thanks
Kevin

[anthonyh]

Fingers crossed too, I'd love to use the 722.6 behind my v12 M120.

[Mclain]

Any news about this little charm?

Or is it possible to get a beta for testing?

[Bernard Fife]

Mclain,

The best place to find out how the code is progressing is in this thread: viewtopic.php?f=4&t=240&p=5051#p5051 At the moment we are waiting for feedback from bench testing with a NAG trans. So things are progressing quietly.

Lance.