[plain OM]

rasterman recently was able to log 5EAT operations (http://www.subaruoutback.org/forums/80-electrical-electronics/66857-log-analysis.html) and was kind enough share some with me. (Thanks!)

The data was interesting, from an academic point of view, but also in regard to the way Turbine speed sensor 1 and 2 are used.

In this first graph, we see that turbine sensor #1 is at zero except when in 4th gear. I checked several logs involving different driving conditions and all had this same characteristic.

I also found in the FSM 5EAT diagnostics section that the TCM does not register a turbine sensor #1 signal in 2nd gear, but does in 4th gear (the two gear positions at which troubleshooting tests are indicated). This is consistent with the graph.

Another observation is the relationship between the input clutch situation and the related gears. As was noted by power12 earlier, the clutch is engaged in 4th and 5th gears only. This is reflected in the following graph.

There's a lot of information available in the logs that tell us what the TCM is doing. This might lead to a more informed diagnosis. In this regard, I'm wondering what would be seen in a log of power12s 5EAT in the lower gears and then when the TCM attempts to shift up to 4 (and 5).

 

[rasterman]

If I have this right, the input clutch is the torque converter lockup... and if I remember it correctly, the torque converter does not lock in 1-3, and locks part of the time in 4 & 5. I've noticed the effect going up hills that get steeper- the RPMs will go up, but without a downshift occurring. I think this is just the TCU unlocking the input clutch to allow the torque converter to do its feats of multiplication.

I also seem to recall a towing thread on this board where Seabass commented that he wished he could lock the converter in 3rd gear for hill pulls without putting so much heat into the ATF.

That suggests to me that the input clutch is not required to work just to get into (or out of) 4th & 5th. It's just programmed to coincidentally lock up most of the time in those gears with some exceptions.

 

[plain OM]

rasterman:

If I have this right, the input clutch is the torque converter lockup...

The torque converter lock-up is different. In the logs, the related data for torque converter lock up control is under two columns headed "L/U Solenoid . . .".

The input clutch control is under "I/C Solenoid . . ."

The 5EAT transmission is made up of three planetary gear sets that are inter-connected. By varying which parts of which gear set are able to turn, or are locked to each other, the overall gear ratios can be set for gears 1 to 5. The input clutch is one of the controls, and is activated for gears 4 and 5.

Endwrench (a now defunct technician-related magazine) had some good articles on the 4EAT, explaining how the two planetary gear sets in that transmission, and their clutches/brakes, interact to provide the 4 different gear ratios. I don't have anything similar for the 5-speed, which has a third gear set, but the principle is probably the same.

Here's a graph of the clutch control signals (solenoid duty ratios) on my 4EAT as it shifts through gears 1 to 4:

In 1st gear, the Brake clutch, High clutch, and Load and Reverse Brake are high, while the Low clutch is low. To shift to 2nd, the Brake clutch is changed from high to low. From 2nd to 3rd, the Brake clutch goes from low to high, while the High clutch goes low. Finally, from 3rd to 4th, the Low clutch goes high, while the Brake clutch goes low.

Here's a similar graph of the 1 - 5 shifts from one of your logs:

Note how the I/C becomes active for gears 4 and 5. And, as charted earlier, the turbine sensor #1 signal also seems to be recognized by the TCM in 4th gear. I suspect that the TCM is constantly monitoring the turbine #1 signal, and as long as it is present, will allow the shift from 3 - 4. It then drops the signal (as seen earlier) for the shift from 4 -5, but resumes it again for downshifting from 4 back to 3. (This is also relevant in regard to your explanation for the torque converter lock up control.)

Here's another graph (from your logs) showing the relationship between the sensor # 1 and 2 signals, and the "determined" Turbine Speed data in the different gears. It again demonstrates how the speed sensor #1 data seems to be available, as an output from the TCM, only in 4th gear (perhaps implying that is when it is used for gear shift control).

 

[cardoc]

To make it simple, in an automatic transmission, in order to achieve a higher gear than 3, a clutch has to be applied to engage the 4th gear. The planetary set will only support 3 gear ratio changes. When shifting higher than 3, the first planetary is locked in the 3rd ratio and the second planetary reduces the ratio more as the gears are selected. The input clutch allows for engagement of 4th and subsequently 5th which are in a second planetary set. Like reverse, in order to change the direction of the output shaft, an independent clutch pack has to be engaged in order to reverse the planetary direction and if there is a pressure flow problem or worn clutches, you lose reverse.

When the clutch is worn, you don't get the engagement for 4th or you get slipping and partial engagement. No 4th means no 5th because if the input clutch can't apply forth, there is no way 5th will engage either. No 4th also means no TCC lockup.

With the way power12 is describing the shifts, the code and his electrical waveform, he has a burned input clutch which would most likely be caused by issue in the valve body.

Rebuild with valve body replacement or replace the complete unit with a known good transmission.

 

[plain OM]

Can't argue with that scenario!

What I did notice in the logs was the specific timing of "events". I've reproduced a small section of 3 -4 shifting in a log to better illustrate this"

Here we see that the TCM starts the change in the I/C solenoid signal at around 29.8 seconds, and it is fully "high" at 32 seconds. This is the only change necessary to shift from 3 - 4.

About a second after the TCM starts the solenoid change sequence, it also starts outputting the Turbine #1 data.

Two seconds after the I/C solenoid increase has started, the TCM indicates a shift from 3 to 4, (at the same time that the I/C has reached its steady-state "high" level. (I'm not sure at this point if the "gear" change indication is actually a verification of the fact, or just indicating what the TCM expects to be the case.)

This same sequence of events seems to appear for each shift. If the time difference was just a few milliseconds, it could well be due to timing of the Romraider logging, but one or two seconds seems too much to be explained this way. Instead, it appears this is an intended sequencing, which is corroborated by the chart I posted above for my 4AT.

If in power12's case the TCM is functioning to shift from 3-4, but the I/C clutch is slipping, we should at least see the I/C clutch solenoid data start to increase, and perhaps the Turbine #1 data about a second later. The actual "failure" would not be detected until the I/C solenoid pressure has reached a high enough point that the clutch should have fully engaged, which is where the "gear" indication would change.

So while it might well be that the I/C clutch is bad, it seems to me that the TCM data could help to confirm the possibility before going as far as dropping the transmission. But I do recognize your experience in this regard.

 

[cardoc]

That's good information.

If the solenoid has failed or is sporadic, you won't see a pressure rise. If the clutch is worn out, same thing. The TCM can command all it wants, but if the valve body has issues, or the clutch is a goner, its a no go. A weak solenoid would have disallowed proper pressure to the clutch allowing for slip over time until the disc have worn to the point of failure. The TCM isn't seeing a ratio error, its not seeing the sensor feedback. Its not engaging. So its either a valve body or a rebuild or both.

Given the nature of solenoids and how they rely on good electrical current flow to operate proper, if the car was run with a weak electrical system, the solenoid would not function proper which would reduce the pressure creating wear; i.e. not enough juice to open as the TCM would like.

This is why I always stress that the battery has to be in excellent condition and providing proper amperes all the time. At the first sign of electrical issues or any code, check the battery. The first thing to see wear in low current conditions is an automatic transmission.

 

[plain OM]

. . .
If the solenoid has failed or is sporadic, you won't see a pressure rise.

Although the 5-speed TCM data for the I/C solenoid reads out in psi, that's not actual pressure. There are no variable pressure sensors in the transmission. The psi indication is basically imputed from the (regulated) line pressure psi and the solenoid duty cycle. So, yes, if the solenoid is bad mechanically, the actual pressure on the clutch could be way off, but the TCM I/C data, as seen on the scanner, would look normal.

If the clutch is worn out, same thing. The TCM can command all it wants, but if the valve body has issues, or the clutch is a goner, its a no go. A weak solenoid would have disallowed proper pressure to the clutch allowing for slip over time until the disc have worn to the point of failure.

Agreed. Although, depending on the degree of slip, that would lead to incorrect gear ratios; that is, input revs versus output revs compared to the design ratio for that gear position.

The TCM isn't seeing a ratio error, its not seeing the sensor feedback. Its not engaging.

In post #24, power12 does report at least one code related to gear ratio error.

. . . Given the nature of solenoids and how they rely on good electrical current flow to operate proper, if the car was run with a weak electrical system, the solenoid would not function proper which would reduce the pressure creating wear; i.e. not enough juice to open as the TCM would like.

This is why I always stress that the battery has to be in excellent condition and providing proper amperes all the time. At the first sign of electrical issues or any code, check the battery. The first thing to see wear in low current conditions is an automatic transmission.

Good points there as well.

If the car could be driven so that the tranny changes from 1 to 2 to 3 and then "stalls" when 4 should be engaged, while at the same time logging the TCM data, we could compare it. If the TCM is commanding the necessary levels for all the shifts, and the turbine sensor #1 signal is also shown normally as the shift to 4 begins, then we can well conclude that the problem is not in the TCM control but in the mechanics, e.g. solenoid malfunction and/or clutch. But if there's no sign of the I/C solenoid imputed pressure starting it's upward journey, or there's no turbine sensor #1 signal, then that would have to be dealt with -- sort of like we can't expect the mechanisms to function properly if the controls to them are not correct.

On another track, I find the indication for Turbine sensor #1 interesting, in that it's present only in 4th gear.

As noted earlier,

"Turbine sensor 1 detects the rotation of the front sun gear, while turbine sensor 2 detects the rotation of the front carrier. Using the values detected by these sensors, the TCM calculates the turbine speed and uses it for shift control."

What is the front sun gear doing when the transmission is not in 4th gear? Is it stationary or free-wheeling? I'm wondering if, when the I/C is engaged (going into 4th) AND the D/C (direct clutch) is already engaged, the sun gear is turned. But, when the tranny shifts to 5th, and the D/C goes low, the sun gear is again freed (or stationary), thus no TCM output. Alternatively, the sensor could be detecting the sun gear's continuous rotation but the TCM isn't putting it through to the diagnostic connector. All academic, of course, and, as in other cases, trying to get a better understanding of what is supposed to be happening.

 

[power12]

thank you

thank you



I wont try to repair this transmission.


Anyone as already try to convert from au to a manual in this forum ?

It is posible to copy the ecu from the manual legacy to the auto legacy?

 

[cardoc]

Plain, 3rd gear is all the gears locked and rotating at the same time.

 

[plain OM]

Plain, 3rd gear is all the gears locked and rotating at the same time.

I appreciate your feedback.

If "at the same time" means the same rpm, wouldn't that be a 1:1 overall gear ratio in 3rd? If so, that would be consistent for the 4-speed AT:

1st 2.785
2nd 1.545
3rd 1.000
4th 0.694
Rev 2.272

but not the 2005 5AT:

1st 3.540
2nd 2.264
3rd 1.471
4th 1.000
5th 0.834
Rev 2.370

 

[power12]

Is the 5eat from a 2.5 same as the 3.0 r ?

 

[plain OM]

The 2.5 naturally-aspirated engine didn't come with the 5-speed automatic -- only the 4-speed or, now, the CVT. The 2.5 turbo and the 3.0 have the 5-speed, but they might have different final gear ratios (front differential) which would mean they can't be substituted without also changing the rear differential.

See attached from the Specifications section of the 2005 FSM. Note: I believe this applies only to North American models. Elsewhere the combinations might be different.

 

[plain OM]

The Mechanism & Function section of the 2005 FSM has a section on the 5AT with an explanation of how the clutches and brakes functions in each gear. I've been "studying" it and sort of getting the idea, but I certainly can't claim any expertise.

In any event, the "Front Brake" (F/B) seems to play an interesting role in regard to the front sun gear. The F/B is engaged in all gears except 4th. (See the chart in post #12 above.)

The F/B, when engaged, prevents the front sun gear from turning. As last noted in post #27 above, "Turbine sensor 1 detects the rotation of the front sun gear. . .".

Consequently, if the F/B is active in all but 4th gear, there would not be any Turbine #1 signal in the other gears.

I went back to rasterman's logs, and confirmed that the F/B is engaged in all but 4th. Moreover, it, along with the Turbine sensor 1 signal, seems to play a pivotal role in the upshift from 3rd to 4th. Here's a part of the log for one shift sequence:

Starting the analysis in the middle, in 3rd gear (the gear positions are marked along the green line near the bottom) at about 29861, (numbers are milliseconds into the log time) the F/B, which had been high, drops. At the same time, the I/C solenoid begins to ramp up, peaks, and then flattens out for a while.

Between 29861 and 31717 (all approx), the AT Turbine speed 1 shoots up from zero. When it reaches the same level as AT Turbine speed 2, the I/C solenoid valve pressure ramps up rapidly, and the gear position changes to "4".

I suspect that when the TCM is anticipating the upshift, it releases the F/B, allowing the front sun gear to rotate. At the same time, it starts the I/C solenoid on it's path to connect the front internal gear and mid internal gear, but it doesn't immediately lock them. After about a second, the AT Turbine 1 sensor signal, responding to the front sun gear now free to turn, ramps up. The TCM detects this, thereby confirming that the front sun gear is free to turn, and it then triggers the I/C clutch to lock front internal gear and mid internal gear. The combination provides 4th gear.

The same occurs when downshifting. Looking at the log in 4th gear (around 23942), F/B is low, I/C is high, and there is a AT Turbine 1 signal tracking with the #2 signal and the imputed Turbine Revolution Speed. (At this time the accelerator position is at zero, so the car is in a decelerating mode.)

At about 24678, F/B solenoid is increased slightly and then flattens out, while the I/C solenoid pressure drops precipitously. Within a second, the AT Turbine 1 signal drops (because the F/B is now applying a slowing pressure). When the #1 signal reaches a very low point, the F/B pressure rises rapidly, and the downshift from 4 to 3 is indicated in the "Gear Position" line.

The action of the F/B on the sun gear, and therefore on the AT Turbine 1 signal, might be the critical factor to the 3-4 and 4-3 shift. If the TCM is finding an error in the #1 turbine signal, it won't upshift.

Back around post #22, power12 reported that the oscilloscope showed a signal when in 4th gear but, strangely, only on deceleration, and after the fault code was registered, that there was a low frequency signal even in 3rd gear.

This might suggest that if there is a hydraulic problem it might be the F/B and not the I/C.

However, I'm not clear at this point if the oscilloscope test was made with the car moving on the street, or raised on a hoist. In the latter case, there would be little torque in the drive train (wheels turning freely) and no real deceleration (the car would "coast" while it slowed. There is mechanical one-way clutch at the front sun gear, and this could perhaps be a factor in the rotation of the sun gear when there's little or no load.

As before, it would be interesting to see logs of the transmission (all PIDs) and how the variables compare to rasterman's data. I still wonder if the logs might highlight something completely different. (Thinking of the "P0420 Diag" thread.)

Comments welcome . . .

 

[power12]

hello the oscilloscope test was made on the street. Not very safe... !!

 

[plain OM]

power12

Thanks for that clarification. Yes, it can be dangerous if there isn't two people in the car, one to drive and another to observe the scope. Good that you were able to do it safely.

I'd like another clarification, in this case, concerning the error code.

There's two transmission codes with the same numbers (digits):

P0715
P0751

In the first post, P0175 was mentioned, but because this is not a transmission error code I had asked if perhaps it might have been P0715 concerning the Turbine speed sensor (in other words, the digits had been inadvertently switched). As noted earlier, a failure in turbine speed sensor 1 can cause symptoms similar to what you reported initially.

However, there is also the P0751 code. This is:

SHIFT SOLENOID “A” PERFORMANCE OR STUCK OFF

Shift solenoid "A" controls the Front Brake (F/B), which, as was mentioned earlier, is applied to stop the front sun gear from turning in all but 4th gear.

If the solenoid is malfunctioning and the P0751 is set, the transmission is locked out from 4th and 5th gear. So we have symptoms that are similar to the P0715 but a different cause.

If the F/B is working properly, when it's applied the front sun gear should not rotate, and therefore there should not be any signal from the Turbine speed sensor 1 except while in 4th gear (or during the shifts to or from 4th gear).

The oscilloscope indication that there is a signal while in 3rd gear suggests that the F/B is not holding the input sun gear as it should, so it is able to turn under certain conditions. The front sun gear is also controlled by a mechanical one way clutch, and during deceleration, if the front brake isn't holding, the sun gear can turn in the reverse direction (counterclockwise). The oscilloscope signal of the Turbine speed sensor 1 would not indicate whether the rotation is clockwise or counterclockwise, but any signal at this time means the F/B isn't holding, when it should be.

If the correct code is P0751, and not P0715, then it takes the troubleshooting in a different direction. If the Shift solenoid "A" is malfunctioning, and the problem is not in the wiring to it, then the repair might involve changing only the valve body, and not the whole transmission.

Both the Front Brake [imputed]Oil Pressure and the Front Brake Pressure Switch signal are among the signals that are provided as outputs from the TCM and should be readable with Romraider Logger and FreeSSM. Comparing the status of the F/B and Switch signals might well reveal that when the transmission is calling for the Front Brake pressure to be "high", the Pressure Switch is indicating a "low", and that would lead to a P0751 error code.

Can you recheck the error code and let us know if it is P0751?

 

[power12]

only after 4th slip or 5th rev like N 0715
I have get not frequently p0734, only one time 0773

 

[plain OM]

Okay, so it is repeatedly P0715 pointing to the Turbine speed sensor 1. (The P0734 also points in that direction.)

Although the P0715 is described as "Input signal circuit of TCM is open or shorted." I guess it could also mean that the signal to the TCM isn't correct. That could be caused by a problem in the sensor circuit (wiring between the sensor and the TCM, or the sensor itself), or as cardoc suggested, a mechanical problem in the transmission.

 

[power12]

wiring seem good. My battery is dead lol
with freessm I can be able to recording the tcm valves commutation

 

[plain OM]

wiring seem good. My battery is dead lol
with freessm I can be able to recording the tcm valves commutation

FreeSSM does not record -- it only displays the information.

Romraider Logger can record the information (numbers) in a spreadsheet format. The data can then be examined separately, after the test drive. The charts I posted here earlier, which include the data for the shift controls (valves/solenoids) are made from a Romraider Logger recording.

(If you haven't already done so, read this thread -- it's a long thread but has information about both FreeSSM and Romraider. )

 

[power12]

kkl interface not working

Hello I have buy an kkl blue interface but its not working.....
I need clarify what's is the problem before replacing the gearbox....