The CVT has been rougher and rougher on my HCH1 since the fluid was changed a few months ago at a nearby shop. (First time there for the HCH1, but we have used them for normal cars for years.) It has not quite reached the judder stage but it just doesn't feel as smooth as it used to. Not sure what they did, or maybe just a coincidence in some other part of the transmission starting to go, but I think a couple of CVT changes might be prudent. I would like to do these myself this time, since I'm planning on two changes, and don't want to have to take it into a (different) shop two times to do so. The Honda repair manual on page 14-196 says

and there is a picture of the flap hanging straight down what looks like about 12 inches. There are not 12 inches of clearance under the car, so that's obviously not going to work as described. There is an old thread on this forum describing the procedure

https://www.greenhybrid.com/forums/f...vt-fluid-4410/

but the pictures are long gone.

Is it actually possible to access the CVT drain plug without the car up on jacks or ramps?

If the car is up on jacks or ramps am I correct in assuming that the "fill" measurement at the dipstick will be off? I don't much trust that in my driveway when the car has all 4 wheels down either, as the driveway has a bit of a slant.

Is a special fill funnel needed? I'm most worried that the CVT fluid may dissolve part of a plastic funnel on the way through, carrying that into the transmission, which isn't something that transmission is going to be happy about.

 

That was interesting. I called the local Honda dealer to see if they had the CVT 1 fluid and crush washers in stock. I asked for two washers and the parts guy said I only need one because

Which makes me think that perhaps even the dealer's service shop does it that way. Seems like a pretty narrow hole to pour fluid through though.

 

Not sure about HCH1, but HCH2 hole was plenty big. It looks like the rubber seal on the HCH1 is about the same size as the HCH2. One of these works great:

 

I have a funnel with more or less that geometry, just need to check that the end fits into the dipstick hole. Also first I think a few drops from the CVT dipstick onto it to make sure the fluid doesn't eat it.

This morning on the way back from the Honda dealership with the CVT-1 fluid for my car the wife casually mentions that sometimes when she drives her 1998 Accord "the engine gets really loud and it doesn't accelerate well". It has been doing this "every once and a while for a long time". Shortly thereafter she revealed that she has no idea what the tachometer is for. Sigh.

So it looks like I will be changing a lot of transmission fluid in the next couple of weeks.

Also, does it make sense to try to do this without jacks, or is that just going to be too cramped a work space, even if it is theoretically possible?

Thanks.

 

I use ramps. Even if I were thin, it would be too cramped.

Unless your funnel is made of something weird that shouldn't be used with petroleum fluids, I can't imagine the fluid would eat it.

 

The dipstick hole on the HCH1 is much smaller than the tip of that funnel. Also the access path is quite narrow. With the dipstick out I was able to twist in a test piece of 3/8" clear vinyl tube. Then nothing in the house would fit on the other end. At the hardware store purchased:

1. 1/2" inner diameter "barb" to 3/4" in MIP (male) adapter
1. 3/4" to 3/4" threaded adapter (female to female)
1. 3/4" threaded (male) to filler 3/4" internal diameter slip right angle adapter
1. small funnel that can be held in one hand while also holding the adapters in the same hand. It does not fit tightly into the adapter.

The 3/8" vinyl tube has an outside diameter of 1/2", or very close to it. The three adapters were screwed together tightly making something that looks a bit like a tobacco pipe. Then one end of a 2 foot length of the tube was "catheterized" into the first adapter on the barb side. Hopefully it won't leak much.

Currently waiting to try it until later in the day, as the concrete in the driveway is way over 100F at the moment.

This morning with the car stone cold and tilted by the driveway so that the front wheels are about 2-3" higher than the rear ones the dipstick is reading above "full" in the hot position. The position is roughly as far above the top of the high "hot" line as that line is from the low "hot" line. The suggests to me that the CVT is probably way overfilled at present.

 

Hope it worked!

 

It did, but not fantastically so. The "bowl" of the "pipe" was a bit too small so that only a little bit could be poured down the tube at a time. It took about 4 minutes per quart. I think next time I will either pull the air intake box and use the designated fill port or modify the rig to have a much larger capacity. On the plus side, the vinyl tube didn't leak either where it went into the rig or the dipstick tube. No signs of damage to the vinyl tube, it was still perfectly clear and didn't look ragged or pitted at the end.

The unpleasant surprise is what came out before the new stuff went in. The fluid was jet black and there was a pea sized jelly like ball of what are probably tiny metal shavings on the end of the drain bolt. Worse, not counting the spills (1/4 cup ?) 4 qt 16 oz came out. The driveway only has a couple of inches slope between the front and rear wheels, and with the car up on a set of jacks at their lowest setting, at the jack points behind the front wheels, it must be pretty close to the angle everybody else is seeing. It didn't look milky at all, and the car has not been using either oil or coolant. My best guess is that the mechanic put in an extra quart 11 months and 4000 miles ago. For some reason I had thought the fluid change was more recent than that, but my records said otherwise.

Anyway. 3qt 4oz (minus maybe a tablespoon from drips) Honda CVT-1 fluid (green label, grey bottle) were put in initially (roughly the "3.1 qt" the manual says), but the bottom of the dipstick was dry after the car was put back on the ground. Drove it around until it had warmed up, although the fan just didn't want to come on, which resulted in an ambiguous reading but which also looked lowish. So another 4 oz were added (to 3.25qt) and I still couldn't get a clean reading. Each time there was fluid up one side of the stick and nothing down the other. Drove it a few more minutes and nope, still couldn't read it right. At least the fluid was still pretty close in color to what it was coming out of the bottle. Even if it is low, it can't be all that low.

The transmission didn't do anything untoward on the short drives, and I did not notice the "almost judder" from before the change. Still, I'm thinking at least one more change, just to be safe, given how much fluid came out, and how dark it was.

 

Here is a summary of the fluid changes for this car:

It is a little disturbing that the amount out doesn't match the amount in at the end of this series. In each case it was drained until it stopped dripping (well, less than a drop in 30s, I wasn't going to wait hours for the last slow drops to come out.)

I still cannot get a clean read on the darrn transmission dipstick. It is always smeared up both sides when measured hot or cold, there is no point which can be definitively identified as the fill level. What seems to be the natural position on the stick is rotated so that it is parallel to the ground (as if it was flat on the ground with a hinge at the tip, then rotated up). Perhaps if I twist it so that the blade is vertical it will drain better.

The manual says to put in 3.1 qt, others say to put in 3.4qt. The numbers above suggest that for this transmission the amount put in which will come back out again at the same volume is actually 3qt 16oz. (Ie, 3qt 8oz was 8 oz low so only 3qt came back out, 3qt 12oz was 4 oz low so only 3qt 8oz came out. Just two data points though, so not very convincing.) Whether or not that is the right fill level I don't know. The transmission was serviced at one point and the mechanic might have put things in oddly so that its capacity was slightly increased.

Regarding clearance, if the flap is left attached at the back and then dropped it is very much in the way of the catch pan. What I ended up doing was to leave in the center back fastener and remove all the others. Then it could be rotated to the side so that it was completely out of the way.

For my last two changes CVT fluid was added through the filler hole on top of the transmission. The crush washer was changed the last time and the part Honda now sells for that (different number) is maybe half the distance inner to outer diameter of the original washer, about the same thickness. My guess is that the greatly reduced contact area should significantly reduce the amount of force needed to break it free. Which is good, because the first time I had to use a 4ft cheater pipe to get enough torque, the breaker bar alone wasn't going to do it for me. (But I'm not very big or very brawny, a bigger stronger guy might have been able to do it with just that tool.)

It is driving a little better but I'm still not happy with the way the transmission is acting. It tends to be a bit jerky if the gas is applied suddenly. Not a judder, more of a sudden engagement. It used to be smoother, less of a clunk. Hmm, maybe I should check the engine mounts.

 

This weekend with the car cold, not having run since the previous day, pulled the dipstick, wiped it off, rotated it 90 degrees so that the blade was vertical and reinserted it. It did not want to go in as easily like that as it does when the blade is flat. When the stick came out there was for once a clear level line. Unfortunately it ran at a shallow diagonal from well above the highest dot to about where the upper "cold" level dot is. Worse, the wet part was on the top of the blade. Since gravity probably is not reversed within this transmission I'm going to put that last result down to a mind fart on my part, like I rotated the blade the wrong way when looking at it. In any case, the simplest explanation for my problems reading this dipstick is that it exits the fill tube at a shallow angle and not straight down.

The tube itself is bent in the middle (part number 9), which probably explains why it was hard to insert the blade vertically: