545RFE, harsh 2 to 3 upshift fix?

ebbnflow · **Posted:** Mon Apr 21, 2014 8:16 pm

Thanks for the reply. When it starts shifting bad, do your CVI values change?

I was suspect of the transgo kit since my problem started about 10k after I installed the transgo kit, but I guess I can rule that out now.

I ran AE while cruising around last night. My TPS looked fine, but I did notice my line pressure drop off when it went from 2-3. Mine hold pretty good around 120psi, but during 2-3 it would briefly drop down to 80 for a split second and then recover. I didn't notice this pressure drop during other gear shifts.

What would cause this brief pressure drop? Valve body? Or something major? I'm not a slush box experts.

lars0247 · **Joined:** Sat Oct 12, 2013 12:09 am **Posts:** 98

ebbnflow wrote:

Thanks for the reply. When it starts shifting bad, do your CVI values change?

I was suspect of the transgo kit since my problem started about 10k after I installed the transgo kit, but I guess I can rule that out now.

I ran AE while cruising around last night. My TPS looked fine, but I did notice my line pressure drop off when it went from 2-3. Mine hold pretty good around 120psi, but during 2-3 it would briefly drop down to 80 for a split second and then recover. I didn't notice this pressure drop during other gear shifts.

What would cause this brief pressure drop? Valve body? Or something major? I'm not a slush box experts.

I do not know my bad CVIs, but will post them when the problem starts occurring again. Did you graph the desired line pressure versus the actual pressure? If the desired line pressure was steady, but the actual dropped, then I'd guess the pump is insufficient or there is some type of restriction on the input side of the pump. I also noticed a brief drop-off in pressure on some of the bad shifts, and as I recall, the desired line pressure during this time remained constant, in other words, the pump could not keep up with where the transmission was trying to keep the pressure level at. If this were to occur, it might intermittently cause the CVI value to drift higher as the TCM may(?) try to compensate for the lower fluid pressure by keeping the solenoids that actuate the 2C and OD clutches open longer during this shift. On the other hand, I drove it tonight extensively around town and on the highway under all sorts of normal driving conditions, and I never once saw any anomalous deviation between the desired and actual line pressures nor did I have one bad shift.

The other thing weird I saw at all was the tranmission temperature briefly, maybe for 1/2 second, on the graph went to zero, and then came back up to its normal value. This never repeated itself though.

Also, I graphed the pedal position 1 vs. pedal position 2, and found they track very close (if you scale the one that is lower by a factor of 2). I do not think it the TPS (accelerator pedal position sensor) is bad.

ebbnflow · **Posted:** Mon Apr 21, 2014 11:52 pm

My desired line pressure is always at 120psi. It never changes under any driving conditions that I have noticed.

I think that during the 2-3 shift a fluid port opens somewhere and has a big enough leak to bleed off line pressure that the pump cannot compensate for. It happens so fast that the pump duty cycle does not even spike. My duty cycle hangs around 15% and fluctuates +/- 5%. Just thinking out load here.

My Jeep always seems to behave better when AE is plugged in. This makes it harder to diagnose. Is this a coincidence?

I also notice that my pedal position 2 reads lower than pedal position 1, but it is only off by a few percent. If I scaled it by 2, it would not match.

lars0247 · **Joined:** Sat Oct 12, 2013 12:09 am **Posts:** 98

My desired line pressure is always at 120 psig when starting out from a stop, or idling in drive, or accelerating, but once a steady speed is reached, the desired and actual will move to 50 psig. It was like this to before I relearned it, and it was shifting badly. Perhaps the control schemes are different between 2005 and 2006, because on mine, the pedal position voltages are related by a factor of 2 but they track together closely.

GreenDieselEngineering · **Joined:** Sun Jun 07, 2009 5:10 pm **Posts:** 1111

The accel pedal has primary and secondary sensor potentiometers. The secondary is half the value of the primary. This is a safety redundancy on electronic throttles. If one flakes out the diagnostics will flag a code and but you in limp mode.

SargeIndustries · **Posted:** Tue Apr 22, 2014 2:20 pm

It seems like mine will shift hard under light throttle and below 1500RPM. If you are accelerating hard, it smooths out when shifting. Makes me wonder if it is a pressure related issue.

ebbnflow · **Posted:** Tue Apr 22, 2014 4:26 pm

GreenDieselEngineering wrote:

The accel pedal has primary and secondary sensor potentiometers. The secondary is half the value of the primary. This is a safety redundancy on electronic throttles. If one flakes out the diagnostics will flag a code and but you in limp mode.

Thanks for the reply Keith. I'll have a look at mine again, but I swear both are within a few percent of each other.

ebbnflow · **Posted:** Tue Apr 22, 2014 4:28 pm

SargeIndustries wrote:

It seems like mine will shift hard under light throttle and below 1500RPM. If you are accelerating hard, it smooths out when shifting. Makes me wonder if it is a pressure related issue.

In my experience, it definitely shifts harder under light throttle. It is better during hard acceleration, but still very present. Unless it is in a good mood, then it shifts just fine.

flman · **Posted:** Tue Apr 22, 2014 5:32 pm

ebbnflow wrote:

GreenDieselEngineering wrote:

The accel pedal has primary and secondary sensor potentiometers. The secondary is half the value of the primary. This is a safety redundancy on electronic throttles. If one flakes out the diagnostics will flag a code and but you in limp mode.

Thanks for the reply Keith. I'll have a look at mine again, but I swear both are within a few percent of each other.

Potentiometers can get flaky, what is one more part? :5SHOTS:

lars0247 · **Joined:** Sat Oct 12, 2013 12:09 am **Posts:** 98

Mine definitely shifts with a noticeable bind under light throttle on the 2-3 upshift at low RPM (when it is shifting badly). The pressure testing recommended in the FSM is much more complicated than just reading the line pressure sensor, as there is an adapter that replaces the pan and I believe allows you to take pressure readings on each of the circuits of the valve body, and there are at least two pressure ports up near the right side by the bell housing.

This much I know, when it is shifting badly, I can see a momentary drop in the actual versus desired line pressure on the 2-3 shift some of the time. I suspect it might be there all of the time during bad shifts, but does not last long enough for the AE program to catch it on the graph.

One other interesting piece of information is that I have had two different TCMs and three different programs in my Jeep. The original TCM with the Jeep or GDE program both drift and shift badly after maybe 150 to 300 miles of city driving (cold weather seems to make it occur sooner), and both shift from 2 to 3 at a fairly low RPM. I did have a TCM from a 2003 Ram 5.7L in this Jeep for a few months, and with the original Dodge program it never shifted badly, but the 2 to 3 shift point was much higher (maybe 2500 RPM) and the pump would be able to supply more oil at a higher pressure under a higher RPM. This same Ram TCM currently has the GDE program in it, and now with the lower shift point from 2 to 3, the shifts are just like the original Jeep TCM once they drift bad.

ebbnflow · **Posted:** Tue Apr 22, 2014 11:26 pm

I checked my pedal positions again. Pedal 1 is double Pedal 2 when I look at the voltages. When I look at percentages, they are almost identical. I was looking at the percentage value when I posted above. I must concur that the TPS has nothing to do with the shifting issue.

Back to line pressure and AE. I guess I haven't stretched my CRD's leg long enough to see the desired pressure drop to 50psi. I'll have to log some more data.

Lars. In AE is there a way to monitor what gear you are in using the Transmission Control Module. I have something called Gear Position Indicator, but it just reports a 0. I've been using engine speed to correlate the shifts with line pressure, but when logging line pressure, desired LP, throttle position and engine speed, AE seems miss a lot of the action and it makes it hard to tell when the shift occurs on a graph. On my next run I'd like to just read line pressure and gear position if I can find the right parameter. Hopefully this will capture a little more of the action.

ebbnflow · **Posted:** Wed Apr 23, 2014 1:02 am

Here's a graph of my latest run. I only logged Engine Speed and Line Pressure. This jeep was shifting hard for most of the drive and the line pressure did not seem to budge. Toward the end of my drive I could see the line pressure go all over the place and the shifting was at its worst.

I put line markers on the graph to show where I think the 2-3 shift occurred. I am assuming the desired line pressure was 120psi during this as I did not log it so I could capture more data points. I don't understand how it can shift a little hard and not effect line pressure unless AE is not logging fast enough. Then toward the end of my run, something (my front pump?) started to really act up and allowed AE to see it misbehaving.

ebbnflow · **Posted:** Wed Apr 23, 2014 12:24 pm

Reading the other thread started by Tedder, it looks like Truckbouy had his TCM, Valve body and front pump replaced and this did not solve the problem. Seems like all that is left, are the tranny internals. If that's the case, I think I'll drive it until it breaks and then put a different tranny in.

lars0247 · **Joined:** Sat Oct 12, 2013 12:09 am **Posts:** 98

ebb,
It definitely appears there is a pressure drop during the shift. Whether this has any bearing on the bad shifting is hard to say. I had hoped that someone out there who had this problem knew the solutions, but it appears that Jeep does not know either from Truckbuoy2's post, and Jeep did not bother to do the full blown pressure test described in the FSM that would test the individual circuits, instead they just hooked up the DRB which of course just reads line pressure sensor, and they just threw parts at it, instead of determining what was really wrong.

If I knew for certain that a reman tranny would fix the problem, I would order one ASAP. I guess for now, I will just continue to quicklearn the tranny when the problem crops up and investigate as time permits.

truckbouy2 · **Posted:** Thu Apr 24, 2014 9:59 am

" Jeep did not bother to do the full blown pressure test described in the FSM that would test the individual circuits, instead they just hooked up the DRB which of course just reads line pressure sensor, and they just threw parts at it, instead of determining what was really wrong."

I know the tech that that worked on Laurie's Jeep, he's about a 35 yr veteran and the ONLY guy they have that knows his stuff (?). I can't say for sure if he did the circuit tests per the FSM but I would venture a guess he probably did. Her Jeep was in their shop about 6 times for this. And you're right on the parts tossing. Valve Body, Front Pump, TC, TC Mounting Plate, TCM, Radiator (How about that one), Solenoid Pack, ANOTHER TCM, Learning, Re-learning, Re-Re-Re Learning. I drove the Jeep for a week never letting hard shift (accelerated through it), Results, none. Same issues.
I live in 4WD country and every one I run across with 05-06 gas Jeep or Dakota (I imagine this transmission was used throughout the line) I ask how their's shifts. Talked to a lot of folks and haven't stumbled across an unusually hard shifter yet.
I'm inclined to believe it's a software communication between the TCM and the ECM exclusive to this CRD model. And I would say, if a solution is found it will be right here and only here.

railguy · **Posted:** Sun Apr 27, 2014 7:28 pm

Since the day that I took delivery of my (then) new 2005 CRD, I have had issues with oftentimes very rough shifts when downshifting from 3rd to 2nd gear on downgrades. Sometimes the transmission will wait for as much as 10 seconds before downshifting. I know that it is not a speed or RPM issue, as it will do it even at relatively low speeds. It is also not consistent--some downshifts will be smooth. I was told several years ago by a pretty knowledgeable Jeep tech that rough 2-3 or 3-2 downshifts were just a characteristic of the somewhat sloppy mating of the transmission with the CRD's electronics and not much could be done about it. The rough downshifts have not broken or damaged anything, they are just uncomfortable. Oh, and my CRD did this both before and after doing the GDE tune.

truckbouy2 · **Posted:** Mon Apr 28, 2014 9:29 am

"I was told several years ago by a pretty knowledgeable Jeep tech that rough 2-3 or 3-2 downshifts were just a characteristic of the somewhat sloppy mating of the transmission with the CRD's electronics and not much could be done about it."

I guess chasing shadows is over. Thanks for the info RG. My suspicions could have been accurate and not just classic paranoia :dizzy:

This will have to be left to the computer whiz kids here at LOST. And I bet they could find the glitch.

lars0247 · **Joined:** Sat Oct 12, 2013 12:09 am **Posts:** 98

Here is what a poster called TransEngineer on the Dodge Ram and Cummins forums told me relative to the problem with the 545RFE;

Sorry for the delay in response. I was checking with some other people for information.

Unfortunately, I don't have much to recommend for you, for the harsh 2-3 upshift. The 2-3 shift has historically been prone to random harsh shifts at light throttle. There are no TCM updates that will correct that problem. The main remedy available is to run another quicklearn when the harsh shifts come back.

Of course, it's a pain (and an expense, I imagine) to keep going back to the dealer for more quicklearns. There are two other options. One, you can drivelearn the trans by making repeated light-throttle upshifts (from a stop, until you get into 3rd gear). Start with the trans warm, and try to keep the throttle constant through the shifts. This should re-learn the OD clutch volume. It may take 10-15 upshifts to smooth the shifts out, and this still may not be as effective as a quicklearn.

The second option would be to get your own scan tool that can perform the quicklearn. I'm not sure whether this is even feasible, but it may be a possibility. Dealer scan tools are really expen$$$ive, and I don't think most aftermarket scan tools will do a quicklearn. You would want a scan tool that can be programmed to send commands to the TCM, and display response codes from the TCM. I'm not sure such a tool is even being sold (other than to dealers and shops at big buck$). But if you find one that can be so programmed, I can tell you how to set it up to run a quicklearn. Or maybe you could find a used dealer scan tool on eBay or something.

I'm really sorry I don't have a better answer for you!

I guess I will quit chasing ghosts and relearn with AE when the shifting gets too annoying. I would guess this might have been part of the impetus for them switching over to the New Generation Controller that integrated the ECM and TCM into one control unit. If you Google "545RFE rough 2 3 shift", it appears this problem also occurs on Dodge Ram trucks and Jeeps with the 545RFE.

Yeti · **Posted:** Wed Apr 30, 2014 10:10 am

Hi guys , I have the same problem , but i think I have found the problem ........or best ......the solution , In a engine ECU there are a map that determine the torque in a low and high load on the gear , in the next few day I'll try to tune the map lowering the torque on low load at low rpm , and I let you know if this work

ebbnflow · **Posted:** Wed Apr 30, 2014 11:17 am

Yeti. Thanks for looking into this problem. I hope you find a solution.

545RFE, harsh 2 to 3 upshift fix?

Who is online