Can't get rid of harsh 1-2 shift in 4l60e

[rhino]

Make

The engine/transmission is 1994 Chevy LT1/4L60e swapped into a 1985 Jaguar XJ6

Model

Jaguar XJ6

Year

1985

Miles

Less that 500 miles on total rebuild

Engine

1994 Chevy LT1 with 4L60e transmission

Hello, newby to the Forum with a transmission issue.

Short back ground: I have wrenched on cars for over 60 years and just finished my last one. For the last 3 1/2 years I have been working on a total rebuild of a 1985 Jaguar XJ6, plus engine swap. I have done all the work myself except the engine machine work, the trans rebuild, and welding up the exhaust. It now has a 1994 LT1/4L60e power train and runs very well.

The trans was a total rebuild by a Tech at the local GM dealer who does all their trans work and also side jobs at home. It works very well in all areas except one, it has a very hard 1-2 shift, even with very light acceleration. He commented on first test ride that it seemed hard and went to the 1-2 accumulator and removed one of the springs (small one) thinking that would soften the shift. It really didn't. After that, he has been very reluctant to come up with solutions, but that's my problem.

I don't know how technical I should get, but I will lay out everything I know
- The first thing we (my Son is a tech at the dealership too) did was hook up a scan tool and measure PCS current.
- At idle in neutral: 1.06a
- Minimum load acceleration: .88 to .90a
- 3/4 throttle acceleration: .43a
- The shift is hard for all tests.

- My tuner states that the PCM is not commanding a higher line pressure that would cause a hard shift. He could see in the videos I sent him that the shift was too hard for the PCS current displayed.
- The TPS has been adjusted to about .58 on idle stop and 4.488 at WOT.
- The IAC is about 50 counts at idle.
- The MAP voltage is 5.039 w/engine off. 1.409 w/engine at idle. (grey/lt. green wires respectively)

- The builder installed a "Corvette" servo.
- He drilled the "#8 separator plate hole" to .070 (I don't know what it is to begin with, but I know it's common. I've done it in the past and noticed a difference)
- All parts are high-grade ie-Sonnax, genuine GM, Transgo, etc.

I told him I wanted a firm shift prior to starting the rebuild, but this is a "drag race" shift, and it shifts like that all the time. I don't want that just driving around town, from stop light to stop light.
After all this, here's what I don't understand, sometimes it will shift normally with a smooth controlled shift. It will make this normal shift all the time when the trans oil is cold (first startup of the day) and then start the harsh shifting when the oil temp comes up. But even after the trans is at operating temp, it will sometimes make a normal shift. If I make a hard acceleration, the shift is hard, as it should be.

I'm looking for help with identifying the cause of this issue or steering me in the right direction. My tuner thought replacing the Corvette servo with a Truck unit would cure the problem, but I've read the Corvette unit also helps the 2-3 shift and I don't want to loose that.

If you are still reading this post, thank you. Any help or suggestions would be appreciated.

Dave

 

[billr]

I did not read it thoroughly, word-by-word, but I do have a few preliminary questions:

Did you transplant the original Chevy PCM along with the engine/trans? If so, any codes being set? Can you read/log live-data? Have you been monitoring line-pressure with a gauge? Do you have the ATSG manual for the 4L60E?

 

[rhino]

PCM, harness, engine, trans all from same vehicle.
No codes shown on CEL or scan tool
I attached two short video's of live scan tool data, maybe that's not what you mean. Explain further.
Have not monitored line-pressure with a gauge.
I do not know what the "ATSG" manual is. I'm not a transmission builder.

Thanks for replying, I appreciate it.

Dave

 

[billr]

I had not looked at the videos before, only skimmed through your original post. Yeah, that is the type of live-data I am asking about. I think I am already seeing that there is no need to check line-pressure just yet. The "PCS current" is pretty much at max (~1 amp), so the LP should be near max, and that is verified by the hard shift.

Since the PCM/engine/trans combo is stock, I would like confirm that it can set codes. Disconnect something, like the cable to the TPS or knock-sensor and see if the proper code is set. Disconnect a cable from the trans, see what that sets.

I am attaching the manual, it should be of some help in our discussion.


OOPS! I just peeked into the manual and realized I made a big mistake in the above: The PCS at 1A should yield minimum LP, not max. You better get out a gauge and verify LP is max, not the min commanded. Also, you could probably use an ammeter (preferably a clamp-on) to see if there is actually 1A going through the PCS coil. Have you checked the PCS coil resistance? (value in the attached manual). How about the TPS and VSS, how do they look in live-data?

I gotta say, I am surprised there are no codes being set. Again, I urge you to confirm that codes can be set.

 

[grcauto]

Did you guys flash the TCM and PCM?

 

[billr]

Are there separate ECM and TCM units, or only one (PCM) here?

Grc, note the edit to the my previous post, I made a big error at first.

 

[rhino]

There is just the PCM, '94 was the first year to use programmable EPROM. '93 and earlier used replaceable ROM-PROM chips. My PCM has been tuned to operate best with the modifications I've made to the engine/car. By the way, the shift points and shift feel are perfect in the 2-3 shift, the converter lock and the 3-OD shift, just as the tuner and I talked about. The 1-2 shift point is right-on, just too harsh for daily driving.

Yes, you caught yourself in time, LP is inverse to PCS current. That seemed unnatural to me too when the tuner explained it. Testing the LP with a gauge is the "absolute proof" of actual pressure, I can do that. I think I have info on what the pressure should be.

As far as disconnecting something to see if a code can be set, I can do that too. Current status of CEL is on with key on, off when engine starts. Scan tool shows everything from water temp to O2 open/closed loop operating properly. My Son did say that sometimes a fault won't set a code because of the frequency or magnitude of the fault. Beyond my pay-grade.

TPS in scan tool showed exactly what I had set with DVOM, .58 on idle stop.

I'm going over everything I've done and learned in the last 3 1/2 years and this thought popped up, might have absolutely nothing to do with my issue. This is my first "modern" computer controlled build and as I read everything I could get my hands on, I came across how the knock module retarded the timing when it sensed pre-ignition. But, a different knock module was used in the one-year 1996 Corvette with the LT4 engine. Sadly, this module is NLA through regular channels. It was supposed to ignore the noise of roller rockers and louder exhaust manifolds, and maybe some more stuff. Well, I built my engine with roller rockers, genuine '61 Corvette rams horn exhaust manifolds and a slightly better cam that the LT4. I found a NOS LT4 knock module and bought it....no, I won't tell you the price. Do you think there is any programming in a knock module that could influence the trans shift? Just a WAG.

I'll hook up a pressure gauge when I can get some time on my Sons lift. I'll unhook something tomorrow to see if I can set a code.

Thanks again for the help.

Dave

 

[grcauto]

I was thinking these were OBDII.

 

[billr]

"My Son did say that sometimes a fault won't set a code because of the frequency or magnitude of the fault."

That makes no sense to me, and I have never heard of it before, of course. We really need to verify if there are no trans codes, just no codes reported. If the PCM fails to report codes, then we have to question all the rest of its functions.

However it does appear it is commanding low LP (high PCS current) when the shifts are hard, so I'm not focused on the PCM trans control so much. But, without codes being reported, then we may be missing some useful clues.

Sweep the TPS through its whole range, make sure readings make sense and that the voltage or % reading changes smoothly.

What about the VSS, have you been looking at that?

There are a variety of ways to fool/disable the KNK sensor to eliminate it as a problem; but then you would not be able to rely on it to prevent spark-knock if you want to run the engine harder than it can tolerate with the fuel used. Honestly, I don't see how this hard 1-2 shift could be caused by the spark timing.

I am familiar with that hard 1-2, that is the classic symptom when my 4L60E sets code P1870 ("Internal Slippage").

Again, check resistance on that PCS coil, if it is OK, then check current.

 

[rhino]

No, '94 and early '95 were OBD1. OBD2 was in full force starting '96.

I swept the TPS slowly through it's full range already, test OK. I agree the knock module is out of the picture.

Does the pan have to come off to check PCS coil resistance/current? If we see the PCS values rise and fall correctly with engine load wouldn't that indicate the coil is working? Just asking.

Dave

 

[bp042665]

ok what you could do is take the 700 R4 1/2 Accumilator and replace it with the 4 L60 E put the servo in the bottom of the housing and use the blue spring that will help have seen this before with the corvette servo being used the early 4L60E use a 700 R4 Assembly the double spring system didnt come in to play until late 96

 

[rhino]

That sounds like what he has already done, I mentioned that in my original post. He didn't say whether he swapped to a 700 R4 housing, I doubt it, but maybe he did. Can you tell me the difference is between the two housings?

 

[billr]

Before you wander in the weeds/pan again, why not check the PCS resistance and current? The resistance is very easy to do, see pages 22 and 109 of that manual I posted. Resistance should be 3.5 to 8 ohms, across pins C and D of the trans connector. Read with the harness disconnected. Readings current may be a bit more difficult, depending on if your have a clamp-on meter, but is more meaningful. Live-data displays a certain PCS current, but I (we?) have no clue if that is commanded current or an actual measurement of the current at the PCM driver. Even if it is the measured current at the driver, we still have no assurance it isn't being dumped in a short in the wiring, not making it to the actual solenoid.

Likewise, checking line-pressure is very easy and meaningful. (See page 9) If LP is "stuck" high, there is no mystery why shifting is harsh.

Have you tried "forcing" the PCM to produce a code yet? VSS?

 

[rhino]

You are right about not wandering into the weeds, I would like to stay focused on finding something that points to the problem. Then figure out the solution.

I don't have any problem checking the resistance/current of the PCS, I just don't know the procedure and that's why I asked about having to remove the pan. The Jag has a very narrow tunnel and getting to the trans connector, past the exhaust, is not easy, but my Son was able to get his hand in there when we tested the harness for continuity. You mention a clamp-on meter, my DVOM will read current with the correct attachment, would that work? Also, would I clamp around the whole harness or do I have to open the harness sheath and isolate the appropriate individual wire?

When you say "LP....stuck high", do you mean a physical part being stuck and not allowing pressure to fall off?

I just finished testing the ability of the PCM to recognize a fault and light up the CEL. It will and it does. Disconnected the TPS, started engine, CEL came on in less that 5sec. Reconnected and started engine again, no light. Disconnected MAF and started engine, CEL came on in less than 5sec. I have not grounded "B" to see what code was set, but I feel confident that a fault can be recognized and turn on the CEL. From the very first start about a month ago, the CEL has never lit up after the engine starts, I guess that's why we don't see any codes with a scan tool. I will read the codes that my test has set and let you know what they are.

Can you tell me what the role of the VSS is in this issue? Here's how the tach/speedo chores are handled in this car.
The Jag is an analog vehicle, the only exception was the electronic FI, which is no longer being used. The GM power train is all electronic and is now the source for the tach/speedo, the two worlds don't mesh without help. The all-GM power train works as normal; output shaft to VSS to PCMin to PCMout and on to a GM speedometer. Only there isn't one, so I have installed a Dakota Digital SGI-100BT interface that will take a variety of inputs and convert them to a variety of outputs. I can match the GM PCM output for the speedo (4000pps) to the needed input for the Jag speedo (8000pps) and then do some fine tuning for tires, gears, etc. This Interface will do the same thing with the tach signal from the PCM by converting that to a signal the Jag tach can use and can be fine tuned.

The stock GM systems operate exactly as they did in the Caprice donor car except at the tail-end of the speedo/tach output wires. The PCM never knows the difference. The VSS does work very well and I have it adjusted for the correct speed on the Jag speedo, same for the tach. But if there is something else I am missing that you want me to test, I will if I can.

Thank you again for your help,

Dave

 

[billr]

Please, peek at pages 6 and 11 of the manual regarding the VSS! The VSS puts out two signals, one for the engine control and a separate one (at much different frequency) for trans control. I will never claim that manual (or any other) is faultless, but let's not ignore it.

Yes, check for what code numbers are set when you disconnect things, make sure they correlate; and be sure to try disconnecting the trans to set a trans code. Codes may be stored, even if the CEL is not on, so it would be great to get a real code scan, not just the "blinky codes".

To use a clamp-on ammeter (either AC or DC) you need to have it around only the one wire you are trying to measure current in.

By "stuck" LP, I mean only that it doesn't change as expected, stays high/low or whatever, but doesn't move much at various driving conditions. If "stuck" it could be mechanical or electrical, first step is to see if it is stuck. How about in live-data, do you seeing the PCS drive current varying a lot there?