- Subject: Erratic 2-3 shift, code P0780
- Unit: BAYA 5-speed automatic
- Vehicle Application: 2003 Honda LX
- Essential Reading: Rebuilder, Diagnostician
- Author: Jesse Zacarias & Roy Delfran
For a while now I have been interested in finding out what is causing third-clutch failures in Honda five-speed transmissions. An opportunity finally presented itself when a customer brought in his 2003 Honda LX with a 3.0L engine and a BAYA five-speed automatic transmission. We had overhauled this transmission three years ago when the vehicle had 60,000 miles. Now with 134,000 miles the vehicle had an erratic 2-3 shift and code P0780 “Problem in Shift Control System.” When we disassembled the unit the only problems we found were burnt third clutches (Figure 1) and badly burnt fluid.
The customer was not too happy with a vehicle that had experienced two transmission failures in 134K miles – the original one at 60K and now the rebuilt that lasted 70K. So I bought the vehicle because I thought it would be perfect for finding the cause of the third-clutch failure, since we knew the vehicle’s history.
I wanted to put the transmission as close as possible to the condition in which it came in but in a workable state to obtain the readings we wanted, so we repaired the transmission only. That is, we replaced the third clutches and steels with Honda original equipment, installed an aftermarket rebuilt converter and thoroughly cleaned the valve body. We did this to find the original cause of the failure.
When we rebuilt the transmission three years ago, we installed an overhaul kit with Honda original clutches; aftermarket rebuilt converter; new linear solenoids A, B and C; and new shift solenoids A, B, C and D. We did not replace the bushings in the mainshaft or the countershaft, since at that time we did not have the tools to replace them. We also did not re-flash the PCM to the latest software or replace the pressure switches.
I connected our test equipment to the vehicle (Figure 2) to obtain as much data as possible. I obtained main line-pressure readings along with the pressure reading for each application clutch, amperage reading at linear solenoids A and B, and voltage readings from each shift solenoid and the third and fourth pressure switches. I even monitored the duty cycles of linear solenoids A and B during upshifts and downshifts.
I looked for any pressure lost in the third-clutch circuit while the transmission was hot, and as Figure 3 shows, third-gear oil pressure was within the 10-psi difference allowed between application pressure and main line pressure. In this case it was equal to main line pressure, proving no pressure lost in circuit 30, which leads to the third-clutch pack. This is interesting, because as we mentioned at the beginning we did not replace the bushings in the countershaft or mainshaft, and by this time they had 134,000 miles of wear.
After obtaining all this data, I sat down for hours analyzing it. I noticed something peculiar when looking at the recordings of the pressure readings taken during upshift and coast downshift. On the third clutch, pressure was present during a forced 4-2 downshift but not as much during the 5-2 downshift (figures 4 and 5).
When I looked at the recordings made with the third pressure switch connected to the scope, I noticed that the pressure present was enough to close the third pressure switch (Figure 6).
This third-clutch pressure was present from when fourth pressure was connected to CPC A pressure (Channel 3), through 2nd gear (Channel 1) and until third gear was fully applied (Channel 2), for a total of about 2 seconds and reaching at times 50 psi. This third-clutch pressure reaches its highest when second-clutch pressure is fully applied at 125-130 psi. That could not be good for the third-gear clutches, but was it a cross leak? It was not present during all the upshifts or coast downshifts and not even during the 5-2 forced downshift. So that almost ruled out a cross leak, but how was this oil pressure getting into the third-clutch circuit on a 4-2 downshift?
To eliminate the possibility of a software update that might solve this problem, I had the PCM re-flashed to the latest software per Honda’s bulletin. When I ran the same pressure test, no difference was apparent other than crisper shifts, so that ruled out a correction with a re-flash.
After studying the hydraulic charts I discovered the reason for the presence of this oil pressure. Remember that the PCM modulates linear solenoids A and B to control the gradual engagement and disengagement of the clutches that are coming on and going off. It does this by slowly increasing or decreasing the amperage. For more details see “A View into the Honda Five-Speed’s Shifting” in Transmission Digest, August 2011.
Let’s look at the PCM solenoid strategy during a 4-2 downshift, using the sweep reading of the scope in Figure 7 as an example. As you can see it takes three steps to accomplish a 4-2 downshift.
We start with all solenoids off in fourth gear.
1) Solenoid C (blue) is turned on for about 0.3 second. This action moves shift valve C, connecting fourth-clutch-circuit oil to CPC A oil pressure, already at high amperage (high pressure), and makes the solenoid shifting strategy the same as for a 3-4 shift: SSA off, SSB off, SSC on.
2) Next, as solenoid C (blue) is being turned off again, solenoid B (green) is being turned on at the same time for about 1 second. This gives us the solenoid strategy for the 2-3 shift: SSA off, SSB on, SSC off. This solenoid strategy moves shift valve B and connects fourth-clutch pressure that is now CPC A oil pressure to exhaust port H4X at shift valve B.
3) Finally, solenoid A (yellow) goes back on, giving us the solenoid strategy in second gear: SSA on, SSB on, SSC off.
SSA off SSB off SSC on
3-4 shift strategy
Now let’s look at what is going on in the hydraulic circuit during a 4-2 downshift(Figure 8). When SSC was turned on for 0.3 second, CPC B oil pressure that was beginning to modulate to lower amperage (Figure 9) increased in pressure. Because of the way the shift valves are placed by the solenoids, this CPC B pressure starts to enter third-clutch circuit through shift valve C circuit 5C, through shift valve B, then through shift valve A and finally through circuit 30, third clutch.
When SSC was turned off again and SSB was being turned on (0.7s on the time scale in Figure 7), CPC B oil continued uninterrupted into circuit 30 third clutch, this time through circuit 5D at shift valve C, through shift valve B, through shift valve A and finally to circuit 30 at shift valve A (Figure 10).
Finally, solenoid A comes on, giving you second-gear solenoid strategy (SSA on, SSB on, SSC off). Third-gear oil exhausts at H3X through shift valve A when shift solenoid A is turned on (Figure 11).
What makes matters worse is that, as figures 6 and 7 show, the 2-3 shift can start immediately if the gas pedal is backed off slightly, by turning solenoid C on. This does not give the CPC B oil present in circuit 30 enough time to exhaust. Because of the 2-3 shift strategy CPC B oil starts to enter again into circuit 30, or third clutch.
On the basis of this finding, it is my opinion that this CPC B oil pressure present on a 4-2 forced downshift is dragging the clutches and contributing to an early third-clutch failure. If we keep in mind that the third clutches are also applied momentarily during the shift from Neutral to Drive (Figure 12) and it requires only 36 psi to engage, we can imagine what 50 psi can do with the second clutch on with up to 130 psi. It also does not take much throttle increase to get a 4-2 downshift on these vehicles. On my many road tests on this car, 4-2 downshift was a normal shift when I entered the freeway.
It is also my opinion that by increasing the size of the third piston in 2004, Honda made the problem worse, since now you have about 1 square inch more application area. I believe the solution would require a software change for the PCM or else the third clutches would have to be made to take this kind of abuse without sacrificing the 2-3 shift comfort.
In the past four months I have been trying to find a solution. In looking into the PCM strategy, it would have to be programmed not to lower the amperage on linear solenoid B during a forced 4-2 downshift (Figure 9). In other shift situations linear solenoids A and B are modulated independently of each other. There may be other reasons why Honda has decided to program it this way, but trying to talk with the company did not produce any results.
My next approach was through the clutches. I tried talking with some clutch manufacturers, but the only one that took the time to talk was Raybestos. The company’s GPX plates seem to be the best choice at the moment, because of their unique material and special grooved pattern that makes them run cooler when third clutches are being dragged. Raybestos told me that with this new information they’ll look for options that may result in even more durability. I still have more testing to do, so my research continues.