AT545 MT643 Electric Modulator Issues - Theory

[Simplicity]

I have delved deeply into why the Allison electric shift modulator used on the AT545 and MT643 are not dependable.

My friend (Rob deserves a lot of credit) and I have come to three main reasons and theories for two of those reasons. Depending on the year of your bus, you may use the T444E Diagnostic Manual version 125 or 190. Both refer to this electric modulator. For other engine types, I have no information.

Main Issue:
ECM does not communicate to modulator relay. Why this doesn't occur, I don't know.
Relay burns out.
Modulator burns out.

Relay Pins:
85 - Coil ground
86 - Coil positive from the ignition
30 - Switch positive from the ignition
87a - Switch closed position allowing 12v to the modulator
87 - Switch open, activated by coil being energized

Modulator Activation/Deactivation:
The modulator is a simple on/off, 12v, magnetic plunger that pushes the modulator pin in the transmission. The pin is spring loaded. When the modulator is not receiving 12v, the plunger is de-magnatized and the spring loaded pin pushes the plunger back into the modulator shaft.

The ECM is set to activate the relay at approximately 80% engine load, deactivate at approximately 65% engine load. Although, a 40+ year Allison guru swears those should be 60% and 40% respectfully.

Theory:
Because the ECM is energizing the coil whenever the engine load is below 80%, which is likely the majority of the time, our theory is this:

Low Engine Load - Per the diagnostic manual, the ECM sends 12v signal to relay pin 85. This energizes the coil between relay pins 86 / 85, which pulls switch (pins 30 / 87a) from closed to open (30 / 87).

High Engine Load: When the engine load hits approximately 80%, the ECM stops sending the 12v, the coil deenergizes, switch 30 / 87a closes and 12v are sent to the modulator.

Due to the high amount of time the relay coil is activated, the coil burns out much sooner than it should.

When the coil burns out, the switch automatically closes and now 12v are being sent to the modulator 100% of the time. Now the modulator burns out sooner than it should.

Once the modulator burns out, it's not too long, unless someone notices the bus isn't shifting properly and stops driving it, until the transmission burns out pre-maturely.

Testing:
Modulator in transmission: Apply 12v (polarity matters) and you should hear a gentle clunk of the modulator plunger. If you can't, remove the modulator for bench testing. WARNING: Copious amounts of transmission fluid will spill out, so have a bucket ready.

Modulator Relay:
Remove relay, test continuity between pin 30 and 87a, should tone.
Apply 12v + to pin 86 and - to pin 85, coil should energize. With coil energized, continuity between 30 and 87a should NOT tone, continuity between 30 and 87 should tone.

Modulator Socket:
Follow the testing for the modulator socket in the diagnostic manual. I believe with KOEO with positive lead of multimeter to pin 30, negative to solid ground (this is important) you should have 12v, same with pin 85. 30 / 87a should tone.

ECM Connector Pin (NOT the ECM socket):
Depending on the year of bus, you could have a different wire number or color into your relay socket. Mine was light brown. For my 2000 T444E, the ECM pin was 56. I used a tone tracer to trace from the light brown relay socket wire to the ECM Connector pin 56.

Solutions/Work Arounds: These are listed in order of least to most expensive.

A) Hopefully it's simply the relay coil burned out and you need a new relay, ~$20.

B) If everything checks out, then it's likely the ECM isn't sending the signal to the relay socket. Workaround is to use a micro switch from the accelerator directly to the modulator. W.W. Williams made such a switch, but you can make one yourself (see below).

C) Buy a new modulator, ~$250

D) There are three (3) types of modulators that are used on these transmissions; Linkage, Air/Vacuum (not sure which) and Electric. I "believe" you can swap the electric with one of the others. I'm not sure what this may cost.

Workaround:
You can either buy the W.W. Willams micro-switch or do the following:

Basically all you are doing is running a wire from the positive of the modulator to a micro-switch at the accelerator and a ground on the bus. You'll have to play with the placement of the micro-switch to estimate where the pedal position for simulating 80% and 65% engine load.

The other suggestion is you wire in an idiot light onto your dash that lights up every time the modulator receives 12v. This way, you know the modulator is getting energized.

You still need to know what your transmission should shift like and when and how it should shift, so you don't have to rely on the light, but it's nice to have.

How well the work around works:
I am awaiting an actual W.W. Williams switch, but in the mean time, we wired from the battery positive to the modulator positive, then the battery negative to a toggle switch and a 15a blade fuse to the modulator ground terminal.

I drove and my friend sat in the back (RE) and when I told him, he flipped the toggle switch. OMG! My transmission worked! So pumped!


So, that's my story and I'm pretty much sticking to it!

[cadillackid]

these relays are continuos duty,they shouldnt fail due to time used.. however like anything they sure can. im guessing you are right though that the solenoid in the modulator itself is not continuous duty so if the relay does fail the modulator is expected to operate continuously in which it will get hot pretty quick.



knowing that the transmission will fail if operated in unmodulated state under heavy loads, navistar built the system to be "fail to safe", there is no damage to the transmission if the modulator is active all the time.. it just shifts a little rough and of course later than normal.. most drivers probably pay little asttention.. driving a bit rough is expected for a commercial vehicle.. esp a school bus or box truck.. if the same happened to a consumer truck it would be in the shop with a customer complaint that they spilled their frappucino when the trans dropped from 3-2..


Navistar operates the electric clutch fan in the same manner.. if the relay fails the nthe fan runs continuously.. of course those clutches are designed to run all the time vs intermittently.



I still need to get that switch pulled off and sent to you.. work has gotten rather busy again so i havent gone over and pulled it off yet. if we were closer I could update the firmware in your ECM, apparently there was a navistar TSI about this issue according to someone in an allison group im in online.. navistar addressed it in the 2003+ ECMs or as an update to the earlier ones. (Not sure why it took them so long, considering these modulators were never sold new in a truck after 2003)...



part of me now wonders if lots of the later AT545 failure are related more to the modulator than to the tranasmission itself.. im far from the only one who finds that the earlier AT545s were much more durable than the later ones.. most of us just chalk it up to "cheaping out" on internal parts.. but perhaps these modulators are partly to blame..



one of my AT545 failures i can attribute to a broken modulator, the other one was on a mechanical engine with a REAL modulation.. that trans I believe the front pump bearing seized and broke the pump. my 3rd AT545 (actually an original AT540).. is vacuum modulated and from what i can tell is still original to the bus that is 43 years old.. (theres no evidence a socket has ever loosened the bell housing bolts or the secondary support)... that bus odometer reads 98,000 and very well may be original..

[Simplicity]

As always, thanks for the clarification and education.

I think you're indicating that an ECM firmware update may address the pin 56 to relay signal issue?

If so, can I update the firmware?

Per the switch, I will be out of town from 4/3 - 4/7, so no rush at the moment.

TTYL.

[Simplicity]

Original Text:
Low Engine Load - Per the diagnostic manual, the ECM sends 12v signal to relay pin 85. This energizes the coil between relay pins 86 / 85, which pulls switch (pins 30 / 87a) from closed to open (30 / 87).

Updated Text:
Low Engine Load - Per the diagnostic manual, the ECM sends 0v signal to relay pin 85. This energizes the coil between relay pins 86 / 85, which pulls switch (pins 30 / 87a) from closed to open (30 / 87).

[Soaringcook]

Can anyone tell me where is this rely located in my bus? What does it look like?

[Hooker]

I'm pretty sure this great running 2000 Chevy BB bus I have right now if having modulator issues. On a quick take off, it's already in 4th gear before I even cross the intersection shift way to soon and then seems to be like it's slipping until the speed finally catches up with 4th gear. Seeing that modulation points are 65 and 80% throttle pedal, I just started doing light throttle takes, like maybe 30-40% and it seems to do so much better. For sure there needs to be a work around for this. I have a feeling, these bad modulators are killing more of these transmissions than heat that everyone complains about. Mine has an electronic modulator but I think I'm goin to take a look at the throttle pedal and see about converting it to a manual modulator with the cable and work around it that way.

[cadillackid]

many people dont even realize the modulator is dead.. and on IH the ECM was completely flaky about when to send modulator on and off signals.. thus why WW williams created the micro switch on the throttle pedal that over-rode the computer... my bus had it however the modulator itself died and I never knew it.. I did replace it and it was too late the transmission died anyway.. I ended up giving that spare modulator away to someone here on this board.


the modulator is a simple solenoid.. if you want to test yours simply put 12 volts DC across the 2 pins (I believe its polarity sensitive and marked).. and it should make a definite "snap".. in the case of mine, it did not.. when I took it apart it was rather simple inside but the electromagnet on mine had coil issues and would not move the solenoid so I got a new one..

testing the full system.. the transmission should definitely shift at higher RPM when floored vs very light throttle.. these are on / off so really not modulators.. technically electric Kickdowns..


below is an image showing the + and - on the unit so if you wantto test yours just hit it with 12 volts... probably a good idea to ohm it out and make sure its not shorted first or the 12 volts might be a good spark show,,

[Hooker]

Originally Posted by cadillackid

many people dont even realize the modulator is dead.. and on IH the ECM was completely flaky about when to send modulator on and off signals.. thus why WW williams created the micro switch on the throttle pedal that over-rode the computer... my bus had it however the modulator itself died and I never knew it.. I did replace it and it was too late the transmission died anyway.. I ended up giving that spare modulator away to someone here on this board.


the modulator is a simple solenoid.. if you want to test yours simply put 12 volts DC across the 2 pins (I believe its polarity sensitive and marked).. and it should make a definite "snap".. in the case of mine, it did not.. when I took it apart it was rather simple inside but the electromagnet on mine had coil issues and would not move the solenoid so I got a new one..

testing the full system.. the transmission should definitely shift at higher RPM when floored vs very light throttle.. these are on / off so really not modulators.. technically electric Kickdowns..


below is an image showing the + and - on the unit so if you wantto test yours just hit it with 12 volts... probably a good idea to ohm it out and make sure its not shorted first or the 12 volts might be a good spark show,,


Attachment 74026

It just seems to be a bad problem all the way around with 3 possible common failures and any 1 of the 3 will kill your transmission. #1 signals from ECM don't seem to be reliable, #2 Relays seem to burn out, and #3, the modulator themselves go bad. I like the idea of adding the pilot light so you know when it SHOULD be working, but even then, all that proves is, is that your ECM send the signal and your relay is still working. Doesn't mean the modulator is going to do it's job. I haven't checked my pedal yet, but I have a feeling it's just going to be the typical drive by wire pedal assembly. I think I really prefer to go with the cable modulator. If that don't work, then the WW Williams micro switch idea might be what I do. Currently with only 43,000 miles, this thing runs and operates like brand new so long as I don't do a heavy throttle take off. As said, to save the transmission, just doing light throttle take offs for the time being til I get this fixed and not letting anyone else drive it.

Pretty sure I'm selling this one anyway but I want it right when I sell it so it's not someone else's broken dream. But if I do keep it, I'm sure I'll be converting it to whatever good transmission donor bus I can score super cheap with a bad engine or other problem. 2000, 2500, MD3060 or even a manual if it's got the right pedals for my model to hang the clutch pedal.