Honda Insight Forum banner

1 - 18 of 18 Posts

·
Premium Member
Joined
·
86 Posts
Discussion Starter #1 (Edited)
Here is a pictorial guide to synchro modification and output shaft assembly. I google image search car problems before I do anything else, as pics of the actual work can be a bit more handy.

Some notes:
Order both replacement Input Shaft Bearings (ISBs) before doing this along with an input shaft seal. They are not expensive and I've found the bell housing bearing is always bad, even if it's not making noise yet. The "back" bearing is usually not bad, but usually not as smooth as it should be either. Replace both and the seal to be sure.

The bell housing side input shaft bearing removal can be performed by driving it out with alternating strikes (around the perimeter of the bearing) with a punch or similar. The "back" ISB is a slip fit part and can be removed by placing the input shaft into the bearing then "wiggling" the input shaft back and fourth while pulling away from the bearing. It will just pop right out.

Don't do this without a factory service manual. Read through the procedure in the manual a couple times in the days leading up to the surgery. Then you'll be familiar with everything before you even get in there.

A trick for getting the input shaft disassembled quickly and without damaging the gears:
Remove the output shaft nut (it is reverse threaded).
Place a block of wood (2x4 or similar) on a concrete floor.
Hold the shaft "threads down" over the block of wood about 6" to 1' above the block.
"Drop" the shaft threads into the block. Allow the output shaft assembly to fall at the rate of gravity, guiding it with your hands. You do NOT want the shaft to hit the concrete floor).
Repeat this 4 or 5 times and suddenly the gear stack will just "drop" off the shaft and onto the block of wood.
Carefully carry all of these parts onto your very clean workspace and lay them out in order and in the correct orientation on a soft cloth.

For assembly of the output shaft, all you need is a section of pipe that slides over the shaft for a couple of the gears that press on and the bearings. Aluminum pipe section is great because it is softer than steel, but steel will work too. It's handy to round these bits up before you have your transmission apart if you have a friend who has an assortment of them.

Don't pry on the gears to get them apart and don't hammer on them to get them back together. You don't need gear pullers for this job.

Refer to the pictures for the orientation of the ball bearings on the output shaft. Not also that the conical washer that goes on before the nut, goes on like a hat, pointing upward in the center. The Honda manual indicates this as well.

I had a replacement friction disc ready for this car, as it had 265,000 miles on it at the time and was still on its original clutch. I was not sure what to expect on the pressure plate or flywheel, but guessed that they would be "good enough" not to need resurfacing. I was shocked to find that they were essentially perfect. Factory machining marks were still present on the face of the pressure plate, and the friction disc itself had 90% life left as per the measured thickness referenced against the Service Manual. I put the new friction disc in anyway (an Exedy part) and could not have been happier with the results. The marcel springs in the Exedy disc gave it a lot smoother engagement without making the clutch pedal excessively ambiguous. Either way, your clutch is probably fine unless the ISB seal has failed and oiled it down. See pics below regarding wear of the pressure plate, flywheel, and friction disc.

When you have the transmission out of the car, it is a terrific time to fix the shifter linkage bushings if they are frozen onto the shift actuators (as was the case on both my Insights and my friend's which we did this mod on as well). To free up the bushing, save yourself a bunch of time and just cut the rubber bushing off of the center bronze sleeve. Take a flat blade screwdriver and just "punch" it down against the bronze sleeve several times in a radial pattern. You can then lift the rubber donut off and get a pair of vise grips on the bronze sleeve. Penetrating oil and some twisting back and fourth will free up the sleeve. You can then wire wheel/sand the linkage and clean out the inside of the sleeve. Put a big dollop of heavy grease like ball joint grease on the parts and reassemble. Slip the rubber donut over the sleeve that you cut it away from, then install the linkage, teflon washer, and steel washer, along with a new cotter pin. The washers and cotter pin will hold the rubber donut in place even tho you've cut it away from the bronze sleeve. I use a ton of heavy grease on the rubber donut when I'm putting it on as well. Wipe off the excess once it's assembled.

I've done two of these ISB/Synchro modifications now, with one being completed in a single day and another being done over the course of a few weeks on a project car. Fellow ICer Rick Hall donated a transmission (failed ISBs, bad synchros) for the "one day" swap. I performed the ISB replacement and synchro mod on the donated trans while my friend worked on getting the old trans out (failed ISBs, good synchros). Only takes about an hour to mod the trans if it's already out of the car, so I finished that, then helped get the old trans out. We then popped the rebuilt one in and buttoned everything back up. Took about 7 hours end to end and we were taking our time. Budget for 12 or so if you've never done it before.

These first five pics show the synchros before and after the modification. The so-called "double cone" synchros on 1st and 2nd gear consist of two brass rings with three interlocking tabs and grooves. These tabs and grooves keep the two brass rings rotationally fixed so that both synchro surfaces (the one on the outer ring and the one on the inner ring) control the position and alignment of the small teeth you see on the gears. There is a steel ring that sits between these two brass rings and engages with the driven gear (1st or 2nd in this case) via splines of its own. The ring with the teeth is called a "blocker ring" because it sits in between the driven gear selector (a splined collar that slides longitudinally along the input shaft to select 1st or 2nd gear) and the so-called "dog teeth" of the driven gear itself, blocking the selector from sliding over the driven gear. The blocking process, where the sliding selector pushes first against the blocker ring, then slides over it and engages with the driven gear, is the brief (depending on how hard you are shifting) window of time that the transmission has to synchronize the speeds of the driven gears and the selector ring/input shaft. Note also that both synchro rings are locked in phase with the driven gear selector. They are allowed to float 1/2 tooth forward and 1/2 tooth back, so that when the gear selector ring slides over them, the ramp angle on their little teeth will transfer the axial load into the driven gear via the clutch grooves, causing the driven gear to synchronize in speed with the driven gear selector. Once synchronized, the remaining force of the selector ring against the oily brass ramps allows the blocker ring to index perfectly (rotating up to 1/2 tooth one way or the other) with the gear selector ring grooves, allowing the selector ring to slide over the blocker ring and onto the, now synchronized, dog teeth of the driven gear. This is how synchromesh transmissions work.

The "double cone"-ness of the Insight (and K Series transmissions) was done to reduce 1st and 2nd gear shift pressure and to reduce synchronization time by increasing the total synchro clutch area and subsequent "bite" (this pressure is a function of how hard one has to push on the stick and how long one has to wait for these parts to become synchronized, which then allows the selector ring to slide over the blocker ring). Using two cones with a steel separator ring effectively doubles the clutch surface area, which speeds synchronization time and extends the life of the clutch grooves (on account of there being twice as many of them). This is especially a "thing" on 1st and 2nd gear as the driven gears are largest on the lowest gears, so they have the highest inertia and require the greatest synchronization energy (you are synchronizing all rotating parts back to the clutch disc, and in the lowest gears the clutch disc has the highest torque multiplication). 1st and 2nd are also the most commonly used gears, so making that shift feel better has a large payoff. Some MFRs do 2nd and 3rd double cones, which perhaps makes more sense (I'd rather have a slick shifting 3rd than 1st). The Toyota E153 is an example of "a bajillion" synchro cones, especially later models which have some triple cone synchros. Honda could have learned about double cone synchros from Toyota. See notes on Honda's double cone problems (at least in the early 2000s) in a post below.

Check out the pic here of an E153 double synchro setup (double cone on the right): http://s36.photobucket.com/user/vip09/media/MR2/20140531_145523.jpg.html

Note the mechanism Toyota uses to lock the inner and outer cones together. This takes us to the five images below.

In Pic 1 you can see the tab on the blocker ring with a score mark in it.

Pic 2 shows both 1st and 2nd rings (with opposite damage marks on the tabs as they "face" each other on the shaft and therefore see forces from opposite sides of the tab). Note the toothless second clutch rings behind the blocker rings in this pic.

Pic 3 shows one of these toothless clutch rings stacked on top of one of the blocker rings (hard to tell in the pic). The tabs from the blocker rings have chipped away the groove on the secondary clutch ring. This is the result of Honda's design flaw (see post below).

Pic 4 shows a modified blocker ring (no tabs) next to a stock blocker ring.

Pic 5 shows both blocker rings after modification. A coarse sanding drum on a dremel works perfectly for removing the tabs.
 

Attachments

·
Premium Member
Joined
·
86 Posts
Discussion Starter #2 (Edited)
Shots of 1st gear and 1st gear synchro assembly.

Picture 1: Here you can see the 1st driven gear ("driven" because it is driven by the input shaft and resides on the output shaft). Notice the "dog teeth" just outside the series of holes that are drilled in the face of the gear. These dog teeth are what the selector ring slides over to transfer torque from the driven gear into the output shaft. When your transmission grinds, it is the face of the selector ring chewing on the face of the dog teeth. All gears in a synchromesh transmission except for reverse are meshed all the time, so you never really "grind the gears". Also in this pic, we can see the plane jane center section of this gear (there's no cone shape to it). Some transmissions have "triple cone" synchros, and they achieve this by putting a cone surface on the driven gear (single cone transmissions have their cone surface there as well), providing one clutch surface on the gear itself, then two on the steel collar. One, Two, Three. These are just double cone tho, which we'll get to.

Picture 2: I've slid the inner synchro ring over the 1st driven gear. With the tabs ground off the blocker ring, this ring now simply acts as a spacer. It will rotate with the driven gear now, and its clutch surface serrations won't do anything. I'm not shedding any tears about this tho as all those phenomenal shifting B and D series transmissions from the 90s worked exactly this way, and my Insight gets regular compliments on how smooth it shifts from anyone who drives it (esp now that the shift linkage bushings are freed up).

Picture 3: Now I've placed the steel cone over the inner synchro ring. This cone has three steel tabs that stick down into those holes drilled in the face of the 1st driven gear. The cone is therefore locked rotationally to the 1st driven gear, but can float axially. This is how the double cone mechanism works. When you shift, the selector ring pushes the blocker ring (which we'll get to in the next pic) over this steel cone, which itself is pushed into the inner synchro ring. The two brass rings are spinning at the speed of the selector ring. The steel cone is spinning at the speed of the 1st driven gear. The sandwiching pressure causes them to synchronize. This is a "double cone" synchro because there are two clutch surfaces at play, one inside the steel cone, and one outside it. As noted above, if the inner brass ring had serrations on the inside, and the gear had a cone machined on it, it would be a triple cone synchro. Note that the inner cones are less effective than the outer ones as they are smaller in diameter, therefore there is a law of diminishing returns at play here.

Picture 4: In this picture, I've placed the modified blocker ring over the steel cone. Note that without the tabs, it no longer engages the inner cone. This blocker ring is now the only synchro for 1st gear. Fortunately for us, this is the outer cone, so it's the most effective (highest torque component on the gear on account of being at the maximum diameter of the two cones). Also, if your transmission has been grinding forever, this cone hasn't been doing anything, so the clutch grooves on it will be good as new!

Picture 5: Here I'm installing the small "spring" over the outer blocker ring. This spring is simply a hoop of wire that slip fits over the blocker ring. When you push the shifter, the shift forks push the selector ring against this spring, which has to flex inward before the selector ring can slide over the blocker ring to engage the dog teeth on the 1st driven gear. This spring sets the amount of force that is required to shift, which is important as this force is what squeezes the synchros against the steel cone, allowing everything to synchronize before the selector ring and driven gear become familiar with each other.
 

Attachments

·
Premium Member
Joined
·
86 Posts
Discussion Starter #3 (Edited)
1st and 2nd gear selector assembly and 2nd gear synchro and driven gear. Third gear ready to go on as well.

Picture 1: Here I've placed the gear selector hub on the output shaft. Note that this hub engages the output shaft via splines. It is locked rotationally onto the shaft. 1st and 2nd gear float on ball bearings, as they are only locked to the shaft when they are selected (a mechanism we will get to shortly). Note the small "teeth" on the synchro blocker ring that engage the selector hub (dead center of the pic). This tooth locks the synrcho rings to the output shaft rotationally, while still allowing 1/2 tooth of fore/aft rotation. This is done so that, once the synchronization event is completed, the driven gear (1st in this case) can rotate a small amount to index its dog teeth properly with the selector ring. If the synchro rings did not have a small amount of rotational play relative to the selector hub, you would have to apply enough shift pressure to overcome the synchro friction (which has just gotten everything moving at the same speed) in order to clock the dog teeth in line with the selector hub, so that it can fully engage.

Picture 2: Here we see the 1st/2nd selector ring. Note the groove around the outside. This is where the shift fork sits. The teeth on the outside of this ring are used by the reverse gear mechanism. Note that these teeth are straight cut, which is why reverse is so noisy. Also, note the splines on the inside of the selector ring.

Picture 3: I've placed the selector ring in place over the hub. You can see how it slides over the hub (and subsequently over the teeth on the synchro blocker ring and dog teeth on the 1st driven gear, if we press down hard enough to compress the small radial spring on the blocker ring).

Picture 4: Here is 2nd gear, all assembled with its modified synchro cone assembly and everything in place. Simply slide this down on the output shaft over the 1/2 selector hub.

Picture 5: The 3rd driven gear. Pretty boring as the 3/4 synchro assembly is on the input shaft, which we haven't had to touch. This is done for space reasons, as well as the fact that 3rd is very close to 1:1 gear ratio, and 4th and 5th are even taller, so the larger diameter gear is on the input shaft. The larger diameter gear is a better place to put the synchro mechanism because it can be made larger diameter and will therefore apply more synchronization torque for a given amount of shift effort. Things have reasons. Usually they make sense.
 

Attachments

·
Premium Member
Joined
·
86 Posts
Discussion Starter #4 (Edited)
Assembly of the 3rd, 4th, and 5th driven gear, along with bearings and washer.

Picture 1: Tapping 3rd gear into place on the output shaft "stack". An aluminum pipe is especially nice as it will be softer than the steel.

Picture 2: 4th gear in place.

Picture 3: This side up for 5th gear.

Picture 4: Output shaft bearings installed. Be sure to get the circlip groove the correct way (towards the end of the shaft) on the second bearing.

Picture 5: Place the washer on so that the cone is pointing "up" with the highest point being the center.
 

Attachments

·
Premium Member
Joined
·
86 Posts
Discussion Starter #5 (Edited)
Nut installation, clutch wear, and new clutch disc.

Picture 1: Nut installed and staked (should end up exactly where it was before disassembly). Note again that this nut is REVERSE THREADED. If you've been working on your car for 8 hours straight, it is easier than you'd think to forget this and be baffled as to why it won't go on.

Picture 2: This is the factory original pressure plate after 14 years and 265,000 miles. You can still see the machining grooves. I didn't bother replacing this...

Picture 3: Factory flywheel, also in spectacular condition. A few tiny hot spots, but I didn't touch this either. The pilot bearing was still in good shape, as will be the case if you replace the input shaft bearing before things get too bad. As the ISB dies, the baby little pilot bearing is left to deal with all that input shaft play, and will basically get destroyed.

Picture 4: Factory friction disc after the same 265,000 miles. Phenomenally little wear (something like 10%). I already had an Exedy disc on hand, or I wouldn't have bothered replacing this. I am not sure what material is embedded in the substrate, but the only material I've seen be so easy on the pressure plate and flywheel is kevlar.

Picture 5: Exedy replacement disc. You're not supposed to replace just the friction disc without surfacing the flywheel and replacing the pressure plate, but having owned $500 EA82 based Subarus that weren't worth the price of machining the flywheel, and putting $20 dollar friction discs in them with no trouble, I wasn't worried about the "mighty" Insight 3 cylinder being "too much" for this type of clutch job (the EA82 was similarly pathetic in its torque, but those cars weighed a lot more). Also, the phenomenal condition of the pressure plate and flywheel left me with little to worry about. As noted above, the marcel spring in the Exedy friction disc has a bit more travel than the OE Honda disc, and this has made for exceptionally smooth engagement, which is a pleasant change from the moderate chatter and slightly aggressive bite of the stock clutch. I've driven many ceramic puck clutches (no marcel springs, no sprung hub) when I was younger and have no desire to deal with chattery clutches anymore, especially in a car that does 0-60 in over 10 seconds.
 

Attachments

·
Administrator
Joined
·
11,729 Posts
Great Pics

Tycho.

Can you add some more text detail about the synchro tabs, what you ground off etc, what with etc.

Perhaps some labels on those pics if you can do it/get time, we do appreciate them anyway.

Is that a before and after pic?

Perhaps also pics of the actual isb and the back bearing you are talking about.

Shifter linkage bush pics etc if you have them?

Thanks again.
 

·
Premium Member
Joined
·
86 Posts
Discussion Starter #7 (Edited)
Notes on Honda's double cone synchro design flaw:

As the clutch material (serrated faces in this case) on the synchro rings wears, a small amount of additional axial (along the shaft) play will develop in the synchro/cone stackup. This may only amount to an extra 0.020", or 0.5 mm in extreme cases, but within the limits listed in the service manual.

Note that wear in the synchro clutch material does not contribute to movement of the components along the shaft, such as the driven gears and selector hub, as these are located by machined features in the shaft itself.

Since the tab/groove engagement on Honda's design is only 1.5 to 2.0 mm deep, any extra play in the synchro stackup will allow these two rings to float around within each other just a little bit more, even if only to slightly different alignments with each other, as the clutch serrations are on the inside faces of these parts. If there is a harsh shift with a minor misalignment or without full tab/groove engagement, the tabs on the blocker ring will chip away the square edge of the grooves on the secondary synchro ring. Eventually, the tab chips away enough material that it no longer stays in the groove and it will wedge itself against the chipped "ramp" that has formed on the secondary synchro ring (see the first set of pictures). This actually turns the synchro stack into an "anti-synchro" that absolutely guarantees the transmission will grind on every shift. There are two reasons for this.

1. Since the two brass rings are wedged "apart" from each other, they cannot be compressed together over the steel cone that sits between them (which the clutches are designed to "squeeze" from each side when shifting).

2. The blocker ring is clocked within a half tooth of the selector ring so that its small ramps, which taper to a point at the center of each tooth, will allow the selector ring to rotate it ALONG WITH the now synchronized driven gear into alignment as the shift completes (as the selector ring finishes sliding over the blocker ring and onto the dog teeth of the driven gear). When the two brass rings become wedged apart, they can no longer use this ramp to ease the transition onto the driven gear because the driven gear is still rotating at whatever speed it was before (it is not synchronized as per point 1 above).

Using only OE Honda MTF and shifting gently may reduce the likelihood of the double cone system failing, but as it is essentially a design flaw that failed to account for any variation in clearances that will be seen in regular duty, it's still more or less delaying the inevitable. Honda suffered this same issue on a whole ton of K series transmissions as well, and K owners didn't buy their cars (RSX-S, etc) to baby the shifts all the time. Nor did Honda set the powerband of the K20A/A2/Z3 in such a way that nice gentle shifts would deliver maximum performance. There are plenty of multi-cone Toyota transmissions such as the aforementioned E153 placed behind the 3SGTE or the C60 in the assorted 2ZZ-GE cars that don't exhibit these problems. If you look at how their synchro rings lock together you can see it is done in a more robust way. I haven't looked inside a more modern double/triple cone Honda transmission, but I would guess that they look a bit different.

Model Year 2000 (or thereabouts) was a massive time for Honda, with the end of the B and D (and F, if you don't consider the F20C to be an F series ... I sort of regard it as a K) and the launch of the K, the R, the L, a new Civic chassis, the Fit was Go (with worse mileage than the Civic), the Insight and the IMA system were launched, etc. It was the first time they put multi cone synchros in their "normal" transmissions too. I'd chalk this up to being a more mechanically challenging version of the leaky roof rail/wet seat belt thing that we all know and love (and that first gen Fit owners know and love too). Or the EGR valves with the reversed potentiometer "rake" that destroys the linear potentiometer. Lots of new designs and new suppliers with lots of new cars platforms with brand new engines. All at once. In this case tho, having a transmission fail behind a 1L 3cyl engine, it's a little bit more embarrassing than having a synchro die on a K that's being wound out to its factory 7900 RPM redline.
 

·
Banned
Joined
·
1,541 Posts
Thanks for a very clear write-up.
 

·
Registered
Joined
·
11 Posts
Great writeup. Los of interesting details about this problem.

Replacing just the clutch disc worked great for me too. Also got the gears apart by the dropping method. Reassembled gears by tapping with plastic-faced hammer then tapped bearings into place with a socket- didnt need to use pipe sections. Those were tips by one of the gurus on an earlier thread. Amazing that no special tools were needed. Thanks to all for these postings without which i would not have attempted this!
 

·
Registered
Joined
·
35 Posts
First off, Thank you for the write up, made this one easy.

I did pick-up a transmission which was told had a bad ISB and starting to grind 2nd. Once apart, you could feel the roughness in the ISB. I went to remove the tabs on the syncos, they appears perfect -but removed regardless.

I put it all back together and can get 3/4/5/R all to shift fine. 1/2, I cannot shift on the bench - I have verified the assembly is correct. Is this due to the spring wire that at least on the bench, can't get the selector gear to deform it enough?

Thanks in advance!


+++ ok, got it. pulled the cluster apart and kept on reclocking until it worked.
 

·
Super Moderator
Joined
·
1,790 Posts
You probably don't have the reverse gear installed correctly on the first gear/second gear synchro hub. The reverse gear has 3 pairs of teeth that are probably not clocked correctly on the hub.

See the service manual page 13-49 left lower side of the page, illustration.

If you can't shift it by hand on the bench, take it back apart.

HTH,
Scott
 

·
Registered
Joined
·
35 Posts
I am using the DL'd pdf, page 104 of the manual section. I did figure out that reverse gear needed to be clocked properly. Shifts fine on the bench now.

Thanks.
 

·
Registered
Joined
·
10 Posts
Thanks for the detailed write-up!
I'm about to rebuild my transmission to fix the grinding problem. I got my hands on a new set of synchros. Would you recommend I keep the old ones and do the mod , or replace them with the un-modded new set?
 

·
Premium Member
Joined
·
86 Posts
Discussion Starter #15
Thanks for the detailed write-up!
I'm about to rebuild my transmission to fix the grinding problem. I got my hands on a new set of synchros. Would you recommend I keep the old ones and do the mod , or replace them with the un-modded new set?
If the old outer syncros have good ramps, I would re-use them and do the tab-grind procedure as outlined above. On my car that has the single cone mod, I don't even think about hard shifts, because there's nothing to break anymore! On my car that still has functional double cone synchros, I am conscious of those little fragile tabs every time I downshift from 3rd to 2nd.

My recommendation - grind the tabs off and be done with it.
 

·
Registered
Joined
·
10 Posts
Thanks for the advice! I modified the syncros and got everything back together. The entire setup shifts fine on the bench, but having trouble driving the shafts when anything besides the 1st gear is engaged. I didn't experience this prior to disassembling. Any idea whether this would resolve itself with extra torque or whether something is likely misaligned?
 

·
Premium Member
Joined
·
86 Posts
Discussion Starter #17
Everything should spin real nice and shifter engagement should be smooth after the mod. As far as the rest of the transmission knows, nothing has changed. Make sure the half shafts are pressed back together all the way and none of the bearings are binding. I didn't have issues with this on the transmissions I've done, so can't offer any specific help there.
 

·
Registered
Joined
·
10 Posts
My problem was with the middle ring of the first syncro which got wedged between the two others, blocking the shafts from turning. Put everything together, and it's running great! Sometimes a little stiff shifting into first from stop, but that was expected.
 
1 - 18 of 18 Posts
Top