kb58
HalfDork
4/29/13 8:35 a.m.
Background: Honda K24 engine + K20 transaxle with straight-cut gears and dog-engagement. New twin-disc clutch was installed per instructions along with the supplied throwout bearing, and the clutch-stop was set at 1/4" past disengagement, also per instructions.
Problem: As the drivetrain warms up, the gears gradually start to become more difficult to engage, with reverse being the worst, followed by first which isn't as bad, and so on. It's as though there is insufficient clutch disengagement. but only when hot. However, adjusting the clutch for more throw doesn't fully solve the issue once it starts happening.
The puzzle: As the car cools off, shifting once again becomes perfect.
Things I've tried or noticed:
1. Clutch feel is always the same whether it disengages properly or not (pedal never feels different.)
2. "Pumping the pedal" makes no difference.
3. When it refuses to go into reverse, if I shift into first and then back to reverse, it'll go right in.
4. Holding the clutch in for an extended time as part of #3 above makes no difference - it doesn't "spin down."
5. I've bled the clutch several times - there are no air bubbles.
6. I have confirmed that there are NO leaks in the clutch system.
I'm stumped.
Could this be a pilot bushing issue? Binding with temperature maybe?
It was also suggested that perhaps when cold, the thicker oil keeps the gears from spinning, and that it's just how dog boxes work. Could that be it? Seems unlikely...
I know enough about how the clutch, throwout bearing, and pilot bearing work that I'm dangerous, hence the questions. I'll take it apart if I know what the problem is, else I won't know what to look for.
Thanks... it's humbling having to throw myself at the feed of the Collective Intellegence, but this issue has kicked my butt...
As your lube gets hot and thinner, the synchros are unable to get enough friction to properly work? That's assuming there are synchros.
Double clutch/rev-match to eliminate this as a possibility.
Vigo
UltraDork
4/29/13 9:15 a.m.
If you're still able to engage first at a stop you probably wouldnt even notice this problem when shifting from a roll unless it's just increased shift effort. I dont think it's a synchro issue anyway. Reverse is unsynchronized and if you cant get into it without gear clashing it's a sign of a clutch issue.
I think it's a clutch dragging issue. It may NOT be a disengagement issue, and everything from the pedal to the pressure plate may be working fine. I think you're on the right track to suspect the pilot bearing/bushing. Moreso if it's just a bushing but anything's possible. Since it's a twin disc i guess it's also twice as likely as normal that one of the discs is wobbling and dragging. How quickly does the problem go away once you stop and wait?
kb58
HalfDork
4/29/13 9:26 a.m.
Thanks for the comments.
The drivetrain has to cool off and since it's a big lump of metal it's at least 3-4 hours. Because it's dog-engagement there are no syncros, and though reverse wasn't swapped, I'm told that it doesn't use syncros in either case.
e_pie
HalfDork
4/29/13 10:09 a.m.
IIRC you have to shift dog boxes like a shiny happy person or else the gears will grind, no granny shifting. I'm also pretty confident you don't need the clutch when shifting a dog box.
kb58
HalfDork
4/29/13 11:39 a.m.
True, and shifting is fine when it's cold or warming up. Once hot though, the only way to get it into reverse is either to have it there when it's started, or briefly put it into first.
Wait. So, it grinds gears whether you're using the clutch or not? As you know, dog box typically means lift throttle, punch the shifter, then throttle back on. If it's grinding when you're not using the clutch, then the clutch isn't your problem. It's the trans, or the fluid in it, or your technique.
Have you tried calling whoever makes the dog box?
Have you tried using thicker oil that won't thin out as much as it warms up? Or maybe a fluid cooler? I always thought dog-geared transmissions required a fluid cooler...
Vigo
UltraDork
4/29/13 1:55 p.m.
Once hot though, the only way to get it into reverse is either to have it there when it's started, or briefly put it into first.
And when you do this, you can feel that things are still spinning when you put it into 1st?
If both gears are unsynchronized (major oversight on my part to not notice that before my last post..), then a clutch dragging issue would give you the same kind of problems going into ANY gear and shifting to 1st before reverse probably wouldnt change anything. So, now im not sure what's going on.
kb58
HalfDork
4/29/13 2:18 p.m.
Thanks for the comments.
The more I think it through the simpler it gets - the transmission input shaft is spinning when I don't want it to. There are only two paths for engine torque to spin the input shaft: the clutch and the pilot bushing, and one or both aren't fully "letting go." The questions are, which one's causing it and why only when it gets hot? Either seems to require disassembling the engine/transmission to have a look. My concern is that I get in there and everything looks fine... then what?
As has been mentioned, a bubble in the slave cylinder could do it, growing in size as the temperature rises, reducing clutch travel, but since it's been bled a couple times that seems unlikely. I can use a dial-indicator to compare the throwout bearing lever travel cold vs hot, though it's hard to keep the pedal depressed when measuring... have to cut a custom-length piece of wood.
Or... squirt the clutch cylinder with freeze spray, that would show real fast whether it's the cause. My guess is that the lever is moving sufficiently in both cases, but guesses aren't allowed at this stage. Pulling the engine is a big deal and it would be nice to know what's going on before diving in.
kb58
HalfDork
4/29/13 2:24 p.m.
I should have added in the first post that I don't know for SURE whether a pilot bushing was installed. If it's missing, I don't see how that would cause an issue when hot. Missing lubrication, maybe... and partially binding at temperature?
Dog-ring transmissions bark :) Shift it with some gusto and try using a thicker viscosity oil once you verify that the the clutch isn't lightly dragging... as the oil gets thin with temp it would drag the shaft and make synchronization difficult in all gears. With the wheels off the ground in gear, clutch in... do the wheels stop spinning?
What trans fluid is being used?
kb58
HalfDork
4/29/13 9:18 p.m.
pinchvalve wrote:
You need a cooler.
If that's true then I have serious problems because the driving was very mild around residential areas... probably 50mph max.
kb58 wrote:
True, and shifting is fine when it's cold or warming up. Once hot though, the only way to get it into reverse is either to have it there when it's started, or briefly put it into first.
That's normal behavior fr an unsynchronized reverse gear.
kb58
HalfDork
4/29/13 9:21 p.m.
Paul_VR6 wrote:
What trans fluid is being used?
Want to say stock Honda tranny oil. Everyone has totally different ideas of what's best, even when it's a contradiction. The gear maker says one thing and the maker of the limited-slip diff says something much different. I guess I'm skeptical that oil brand is root cause since it's not being run hard, at all.
kb58
HalfDork
4/29/13 9:23 p.m.
Kenny_McCormic wrote:
kb58 wrote:
True, and shifting is fine when it's cold or warming up. Once hot though, the only way to get it into reverse is either to have it there when it's started, or briefly put it into first.
That's normal behavior fr an unsynchronized reverse gear.
I'd believe that if it acted the same way hot AND cold, but it shouldn't change if it's a design "feature. "
kb58
HalfDork
4/29/13 9:31 p.m.
New data point: confirmed that the clutch pedal is moving the throwout lever the exact same distance hot or cold... whatever is going on is inside :(
Oh, and I jacked up the rear tires to see what they do in this condition and reverse and first barely turns the wheels with the clutch in, but they do turn. Once it's cooled off it doesn't do that.
Cold the fluid is thicker and slows the rotating mass down faster, hot the fluid is thinner and lets things rotate more freely, if you stand on the clutch for a minute it would probably go in.
Stock Honda oil is thin to begin with too, I would be running a 75w90 gear oil in that trans.
kb58 wrote:
Oh, and I jacked up the rear tires to see what they do in this condition and reverse and first *barely* turns the wheels with the clutch in, but they do turn. Once it's cooled off it doesn't do that.
Sounds like a clutch disengagement issue that the thicker oil is masking when cold, or the disks in the clutch are expanding enough to prevent disengagement when warm.
The trick for setting up a twin disk on a C5 Corvette is to put the car on level ground, fully disengage the clutch via the pedal, and floor it. If the car moves forward at all, the clutch isnt fully disenagaged.
I would agree with checking disengagement. I run my stop just before te release hits the cover.
Stock Honda oil is light in a stock trans. Look into the Torco mtf or rtf or Reline shockproof (light) for a dogbox.
kb58
HalfDork
4/30/13 8:22 a.m.
The problem is one of contradicting requirements. The gear manufacturer recommends shockproof and friction-modified type oils... Meanwhile, the LSD manufacturer recommends NOT using shockproof or friction-modified type oils... and there I sit in the middle.
Gears are PPG straight-cut with dog-engagement. The LSD is a "WaveTrac", and because I had to pick something to move on with things, used Honda gear oil.
If you guys have gear oil recommendations for the above combination I'm all ears.
Oh, and I confirmed that hot or cold, the clutch pedal is moving the throughout lever the exact same distance, so whatever's going on is internal :(
Another thing, and this may be a big deal. When I'm cruising in 4th at constant speed, there's very little load on the straight-cut gears, and there's a fair bit of gearbox noise. I know what straight-cut gear whine is and this is different; it's like something's rattling or vibrating. It may be completely normal but because I've never run such gears, I don't know normal is. What do straight-cut gears sound like when speed is near-constant, with no speeding up or down? Is there a clattering associated with the gears going in and out of contact with each other?
Vigo
UltraDork
4/30/13 9:23 a.m.
Ive had chattering/vibrating noises from lugging a car down really low rpm in gear. I think if theres enough time between power pulses from the crank you will get some noise from any gearbox as things load and unload.
But, that's LOW rpm. I doubt that is your exact issue.
The rtf is a 75w90 thats a little on the thicker sife and could help if too low viscosity at temp is a problem. Call Torco but I am pretty sure it doesnt have any the friction modifiers that would be needed for clutch lsds.