47RE Torque Converter Lockup Adjustment

[kzimmer]

Hey All,

 

I would like to find a simple, inexpensive and automated way to adjust when torque converter lockup occurs. I am not interested in off the shelf transmission controllers. I have an electronics and industrial automation background. I am interested in hearing ideas that pertain to a DIY install and long term reliability, as well as ease of use for someone else driving the truck.

 

I have 3.54 gears and 35" tires. I'm fully aware what that means for my final drive ratio, and I'm not entertaining a regear at this time. Please don't suggest obvious large mechanical changes.

 

The truck has a lockup switch, which I generally don't use. While cruising at around 60km/h, in 3rd gear, my TPS is hovering right around the area where lockup is allowed (12% ish). There is only maybe a 2% hysteresis here, and the converter will unlock (thanks Chrysler). This is especially obvious if I set the cruise. Lock, unlock, lock, unlock. Couple this with the fact that my RPM's are too low in this situation, and a little bit of surging can occur. Really, if this trans was properly controlled, it'd be in 2nd lockup at this point. But let's not dive into that. 2nd gear might as well not even exist in this heap.

 

A similar but less severe situation occurs at 85-95 km/h in Overdrive. I'm not as concerned about that, I usually lock out O/D until at least over 100 km/h. I mostly want to clean up 3rd gear mid speed cruising.

 

So ideas. I could adjust the Quadzilla's 0-5 psi fuel percentages. If I decrease these percentages, it will force more TPS % to go the same speed. This only fixes half the problem; it'll still lockup too early, just won't keep unlocking because of low TPS. And there's the obvious side effect of increased spoolup time because of less fuel.

 

I could tap into the TPS analog signal, and use a potentiometer to make an adjustable voltage divider to trigger a low voltage relay, and use the contacts of the relay to open up the lockup circuit until the TPS % is higher. This would be the ideal situation where possible. It would force lockup to be delayed until higher throttle, which would mean higher speed and higher RPM. Except, this may still allow the lock-unlock-lock-unlock situation, just moved into a higher TPS range. It might even make it worse. I could maybe use another relay to allow unlock at a different TPS but that's getting a bit complex.

 

If there was a DC wheel speed signal somewhere, I could do something similar with a speed trigger instead. But I'm guessing all tappable speed signals are AC (or rather pulsed DC).

 

I'm sure there are more reasonable inexpensive solutions that I am not aware of. I'm going to keep thinking about this, and I plan on implementing something when the snow is off the ground.

 

Thanks in advance for your ideas.

 

[W-T]

6 minutes ago, kzimmer said:

Hey All,

 

I have an electronics and industrial automation background.

 

There is only maybe a 2% hysteresis here, and the converter will unlock (thanks Chrysler). This is especially obvious if I set the cruise. Lock, unlock, lock, unlock. Couple this with the fact that my RPM's are too low in this situation, and a little bit of surging can occur. Really, if this trans was properly controlled, it'd be in 2nd lockup at this point. But let's not dive into that. 2nd gear might as well not even exist in this heap.

 

A similar but less severe situation occurs at 85-95 km/h in Overdrive. I'm not as concerned about that, I usually lock out O/D until at least over 100 km/h. I mostly want to clean up 3rd gear mid speed cruising.

 

 

Thanks in advance for your ideas.

 

 

I'm sorry...a little clarification would help further in understanding what you're experiencing...when you are just driving along in 3rd gear are you experiencing this "lock,unlock and lock" syndrome? Is this experience occurring often or on a regular basis in your opinion?

[kzimmer]

2 minutes ago, W-T said:

when you are just driving along in 3rd gear are you experiencing this "lock,unlock and lock" syndrome?

 

Yes. The PCM (or ECM, which ever it is) allows lockup to occur around 12 or 13% TPS, and then unlocks around 11%. Something like that. So if I'm cruising along around 60 km/h in 3rd gear, at 13% it'll lock, throttle will have to back off a little bit to maintain speed now that the trans is locked, lose a couple % TPS, then unlock. More TPS is now required to maintain speed. Repeat.

 

6 minutes ago, W-T said:

Is this experience occurring often or on a regular basis in your opinion?

 

Yes. I mean I can bump the shifter into second but I'm not a huge fan of doing that, or trying to remind my wife to do that.

 

[W-T]

Very well...I believe you'll find excellent answers for this subject matter on this site. @Mopar1973Manhas posted a comprehensive test that directly answers this question. 

 

Take a look at 47RE Torque Converter Lock and Unlock post. I believe you'll find this very valuable information in the archive Articles. Take a look and provide feedback on this thread...I would be curious regarding your impression and observation. 

[Mopar1973Man]

Might talk to @Dynamic ask him about it as well. He might shed some more light on the idea.

[Me78569]

His issue isn't related to the typical ac noise, rather the tuning the PCM.  

 

auto guys with BIG injectors have this type of issue at cruise under very light load. 

[Mopar1973Man]

1 minute ago, Me78569 said:

His issue isn't related to the typical ac noise, rather the tuning the PCM.  

 

 

Why I mention Dynamic to draw him in and maybe he's got some insight into the method of tuning the lock up. 

[kzimmer]

5 minutes ago, W-T said:

Very well...I believe you'll find excellent answers for this subject matter on this site. @Mopar1973Manhas posted a comprehensive test that directly answers this question. 

 

Take a look at 47RE Torque Converter Lock and Unlock post. I believe you'll find this very valuable information in the archive Articles. Take a look and provide feedback on this thread...I would be curious regarding your impression and observation. 

 

Yessir, I'm fully aware of the information available on this site, and I am aware of most lockup issues. This is not an issue of AC noise or bad APPS. I do appreciate the advice very much; however I'm more looking towards a solution to this particular situation, and ideas to adjust when lockup occurs, rather than what else can cause various lockup issues. 

 

 

7 minutes ago, Me78569 said:

i've thought about this for some time now.   I really wanted to build some sort of lockup controller, but I never got around to it. 

 

 

You could in theory tap tps input, rpm, speed, into an arduino.  They output ground to the lockup wire when you want to based on some code / maps.   

 

 

Absolutely. This would likely be the most effective method. Just debatably not not simplest... But this might not be possible to over-simplify. It's a conundrum.

 

10 minutes ago, Me78569 said:

Not knowing gear makes it somewhat hard though.  all it takes is one hard 2-3 shift locked and your input is toast.

 

Totally. I don't want to risk that. Mind you, if one were to use vehicle speed, you can be fairly sure you are fast enough to be shifted from 2nd to 3rd.

 

 

6 minutes ago, Mopar1973Man said:

Might talk to @Dynamic ask him about it as well. He might shed some more light on the idea.

 

Good idea, thank you. Maybe there's some secrets worth sharing, haha.

 

4 minutes ago, Me78569 said:

auto guys with BIG injectors have this type of issue at cruise under very light load. 

 

True story. I've been known to try to have my cake and eat it too... Haha.

 

[Me78569]

you would have to compare rpm to speed and qualify gearing....as you might notice the quadzilla does this HOWEVER it is not %100 accurate and I have noticed some points where I am in 2nd when it thought I was in 3rd.  the resulting locked shift would have been no fun.

[kzimmer]

1 minute ago, Me78569 said:

you would have to compare rpm to speed and qualify gearing....as you might notice the quadzilla does this HOWEVER it is not %100 accurate and I have noticed some points where I am in 2nd when it thought I was in 3rd.  the resulting locked shift would have been no fun.

 

Makes sense. I guess that's another point in favour of creating a circuit that disables lockup in certain situations, rather than attempting to fully control it.

[Me78569]

might talk to @Chris O.  I think his CCD tool will tell you current gear according to the pcm.  It also reads road speed, and oss speed / tps etc etc etc.  he could build on heck of a lockup controller.

[dave110]

I realize you're not new here, and I know you said you have a lockup switch, but have you ever noticed the Cadillac version of it that @IBMobile created? Sounds like that might do what you wish to accomplish. There's an article on it. Just a thought.

[kzimmer]

29 minutes ago, dave110 said:

I realize you're not new here, and I know you said you have a lockup switch, but have you ever noticed the Cadillac version of it that @IBMobile created? Sounds like that might do what you wish to accomplish. There's an article on it. Just a thought.

 

I did see that, and I thought it was really badass. Great idea. Only problems are that I don't want to have to teach someone else driving the truck to manually activate lockup, and I don't want the risk of locked shifts.

 

 

Below is a screenshot of the factory service manual. Not shown in this screenshot is the APPS, which is wired to the ECM. It looks like, for some reason, the APPS (TPS) signal is repeated by the ECM and retransmitted, hard wired, to the PCM. I have no idea why they chose to do it this way instead of sending it over the data bus. I also don't know how it's scaled. Assuming 0-5 VDC, just like the ECM. Can anyone verify? I'll measure if need be.

 

So, maybe I can break the signal at this point and use a voltage divider to reduce the TPS voltage that the PCM sees. This would definitely delay lockup. However, does anyone know what else the PCM might use this signal for?

 

Edited March 15, 2018 by kzimmer

[notlimah]

[jlbayes]

Throw voltmeter on the circuit and move the apps through its travel. Yes the apps is a 0-5v sensor. The tps wire to the pcm should respond in a similar fashion. I have stopped trying to figure out wtf Diamler designed on these trucks long ago. Just leads to massive headaches and excessive drinking.

[Dynamic]

Yeah, you're dealing with the factory programming, which is not that great. Your situation is exacerbated by your gearing.

 

It is my understanding that Chrysler received the engine assemblies from Cummins complete with electronics (APPS, ECM, etc.), and ready to drop in as a pre-configured unit. The ECM simply outputs a 0-5V TPS signal (based off of the APPS signal the ECM sees) and crank sensor signal (to reference engine RPM) to the PCM for use in controlling the transmission.

 

There are a myriad of different - let's call them "anomalies"...to be polite - that I would find "work-arounds" to if I was a smarter guy. This is definitely one of them. I have, on one truck that I built a transmission for, played around with some fixed resistors on the TPS signal line to delay the shift into OD, which did work, but that truck was a 12 valve with a 47RH, so I was just manipulating OD and TCC lockup. I would surmise that the same could be accomplished on a 47RE with some ingenuity.

[kzimmer]

Thanks for the responses guys.

 

@jlbayes That's exactly when I planned on doing a couple weekends from now when I get the time, I'll throw a meter on and see what the range is. I just need to find the right logic ground reference in that harness first. The main PCM ground might have to do.

 

3 hours ago, Dynamic said:

I have, on one truck that I built a transmission for, played around with some fixed resistors on the TPS signal line to delay the shift into OD, which did work, but that truck was a 12 valve with a 47RH, so I was just manipulating OD and TCC lockup. I would surmise that the same could be accomplished on a 47RE with some ingenuity.

 

On the 47RH did it affect lockup and OD or just OD?

 

I think this will do what I want, and if it affects both, that's just icing on the cake.

 

I'll report back with my findings when I learn something.

Edited March 16, 2018 by kzimmer

[kzimmer]

@Me78569 Could the Quadzilla be programmed to filter TPS? I'm not sure if it's possible to intercept and overwrite TPS.

 

That way 0% in could be 0% out, 100% in could be 100% out, and the curve could be modified, possibly with a single adjustable parameter which could identify the midpoint of an S curve.

 

Thoughts?

 

Edit: I just realized that would also affect fueling, and as a result, do nothing. Lol.

Edited March 16, 2018 by kzimmer

[Me78569]

all sensor data that comes from the j1939 bus under the hood is 1 way. 

 

[kzimmer]

Well then... Here comes the side cutters, resistors, and butt splices...