posted on July 31, 2012 09:00
Project Miatabusa Part 16:
What's That Noise?
by Dave Coleman
First things first: Where the hell has the Miatabusa been?
There's a complicated string of excuses involving lots of things that have nothing to do with Miatabusas, like buying 90-year old homes, filming zombie movies and machining medical devices. (This is a complex alliance, this Miatabusa cabal...) But there was one car-related issue that contributed to this big delay: a horrible, horrible noise.
As you'll recall, when we first fired it up, the Miatabusa impressed us not only with cool vroom-vroom noises, but also with a chainsaw massacre sized oil leak and a nasty rattle noise where an oil pan bolt was tapping against the subframe. Or at least that's what we thought was making the noise...
After tearing things apart, fixing all those problems and firing it back up, we heard the same nasty rattle we thought we just fixed.
This is what it sounds like:
As you can hear, the rattle sounds different at different times. From a deep, slightly nasty background noise at idle to a sawzall-through-your-skull resonance at around 6500 rpm. The sound is dramatically worse from inside the car, where that mid-rpm buzzsaw sound is like a murder of hammers trying to bust through the floor.
When we first heard the noise at idle, we suspected a bum timing chain tensioner. After ruling that out, we hoped it was a normal Hayabusa noise. Calling in Neel Vasaveda from Apex Speed Technology, his experienced ear, ominously, had never heard a Hayabusa engine make this noise.
When we finally drove the car and heard the deathhammer, we naturally assumed it was something hitting the bottom of the car. Hence the excitement when we found an oil pan bolt hitting the subframe.
Back together with all the obvious problems solved, though, the noises were exactly the same.
Maybe that wacky wraparound header has some nasty resonance? Damping it with a (gloved) hand did nothing to the sound, nor did changing the natural frequency by removing exhaust support brackets. And despite the distinct sound of stuff hitting the floor, none of the tight-fitting exhaust was actually touching the car.
Listening to the noise from below on MD Automotive's lift it was suddenly easy to pinpoint the source of the sound to about 6 inches forward of the flywheel. Right about here:
When the engine was out for its last round of leak and rattle repairs, rocking the flywheel back and forth made a clank noise that sounds exactly like a single note of that horrible rattle. That sound is from the backlash between the primary drive gear on the crankshaft and the driven gear on the flywheel shaft. (If you don't remember how or why we have a flywheel shaft, brush up on your Miatabusa history HERE).
After eliminating every other easier-to-fix possibility, it is now painfully clear that the death rattle is coming from this gear. Here's what we think is happening:
In stock form, the primary drive gear drives the clutch basket through a sprung hub. Note the six springs buried in this clutch basket, and the three giant rivets holding the clutch basket to the sprung hub. The purpose of these springs is to isolate the transmission from the torsional oscillations of the crankshaft.
Crankshafts don't spin smoothly it turns out, but actually speed up and slow down many times during each revolution. There are three main driving forces behind these variations in speed. First, and most obvious, is because the four-stroke cycle involves a big burst of torque during the power stroke, followed by 540 degrees of not only coasting, but actually doing work (and therefore slowing down) sucking in and compressing air. Now, stacking four cylinders in a row and staggering all of their 4-stroke events by 180 degrees should even these pulses out, since one cylinder is always on the power stroke. Of course, the torque delivered during the power stoke isn't delivered in a nice even push over 180 degrees. It comes on with a bang, peaks somewhere around 15 degrees ATDC, and then tapers off. You have to stack a lot more than four of these together to add up to smooth torque delivery.
Tuesday, July 31, 2012 1:45 AM
Interesting problem. What's to keep you from flipping the sprung-hub situation around and keeping it all inside the now-empty gearbox? You've already done a ton of custom work to stiffen up that side of the engine and it would be a shame to start over!
Tuesday, July 31, 2012 5:46 AM
Wow, I read in awe. Even your mistakes are incredibly fun to read about. Keep up the great work. I'm already anticipating the next installment.
Tuesday, July 31, 2012 7:37 AM
oh wow, i feel your pain. this is the exact same issue i had when i put a WWII radial engine into a Toyota MR2. my initial design had the flywheel seperated from the crankshaft with a chain and a gear (seen here: http://www.youtube.com/watch?v=8iN_2Jhryzs ) in the end it caused mass carnage from all the hammering back and forth and i had to completely change the design.
i corrected the issue by adding a flywheel directly to the crankshaft which you cannot do here and fixing the drive system so there's a whole lot less lash in the drive. about 1:10 into this video you can see the new assembly in motion: http://www.youtube.com/watch?v=qe5A4kt04Dc of course, with an open headered 540cubic inch motor, you can forgive quite a few vibrations.
Tuesday, July 31, 2012 9:22 AM
Reminds me of two of my favorite quotes...
"I haven't failed I've only found 10,000 ways that don't work"- Thomas Edison.
"If you're not breaking stuff your not trying hard enough"- 3SX.com
Keep up the good work Dave and CONGRATULATIONS ON YOUR FAILURE! :-)
Tuesday, July 31, 2012 10:40 AM
"Put enough clever people in a room together and occasionally their cleverness will cancel each other out..." - Dave Coleman
Engineering Quote of the Century!
Tuesday, July 31, 2012 10:56 AM
What about a dual mass flywheel in place of the Miata unit? I know it would be a little farther down the line from the drive gear than stock, but it might work...
Tuesday, July 31, 2012 11:27 AM
That quote from Edison is what gets me through each and every day.
Tuesday, July 31, 2012 12:59 PM
I like your out-of-the-box thinking, but if you look at the picture of the Hayabusa's exposed bottom end on page 1, there's a bearing and bulkhead just forward of the gear, so there's nowhere to flip that sprung hub situation to. I think we'll be able to make room once the basket is cut down a bit. We have a pretty solid plan sketched up (in our heads).
Tuesday, July 31, 2012 1:02 PM
Funny thing is, I was watching your radial MR2 and all your backlash mayhem, and it never occurred to me that we would have the same issue with our smooth little engine. Being a factory-engineered gear, I had just put primary drive gear concerns away in a corner of my brain where I keep stuff that's not worth worrying about.
By the way, for those who haven't seen it, Marc's monster build thread is here: http://forums.24hoursoflemons.com/viewtopic.php?id=4861
Tuesday, July 31, 2012 1:05 PM
A properly-tuned dual mass flywheel would accomplish the same thing, but they're thick enough to require a whole different clutch solution and they add significant weight to the car. Plus, there isn't one designed for the Miata, so we'd have to adapt something from another car. Probably just as much work, in the end, as putting the factory damper back in, and the result will be heavier.
Tuesday, July 31, 2012 1:09 PM
Dave, yeah i would have never claimed an "i told you so" i never imagined that the factory gear would have enough lash for that to be an issue.
the other things that occurs to me is the stock drivetrain does not have a flywheel load when it isn't clutched in. it looks like the basket has much less load to accelerate/decelerate than even a lightweight flywheel. and once loaded the springs probably keep the loading on one side of the gear only.
with the springs back in you may have an issue that you have too much momentum and not enough load so it still happens. you may have to tune the flywheel weight to get it to stop.
not sure what the solution is, but i agree that re-installing the basket dampers back in is probably a very good start.
Tuesday, July 31, 2012 1:15 PM
what about a solution like the Toyota camshafts that transmit power from one cam to the other using gears? one of the gears is split in two and is springloaded to remove the lash.
you can kinda see what's going on here: http://repairguide.autozone.com/znetrgs/repair_guide_content/en_us/images/0996b43f/80/20/92/98/small/0996b43f80209298.jpg
that way all your power is still transmitted through your existing shaft but a 2nd layer gear just holds everything tight.
it even looks like the drive gear is already wider than the driven gear. maybe they used to have a different factory solution like this and you can snag parts from an older bike?
Tuesday, July 31, 2012 1:22 PM
Very interesting. I've seen that groove the in cam gear before but never had any idea that's what was going on!
Tuesday, July 31, 2012 1:38 PM
That split gear solution is pretty great and might be worth exploring if we come up snake eyes again. Bikes have been doing it with a sprung hub for eons. Interestingly when listening to my CBR1000RR motor (same design at the 'Busa) very carefully, there is a pretty perceptible rattle at low rpms. It goes away once the rpm's climb and the hub springs stay loaded up.
Just to get a feel for this I tried lowering the idle to 300 rpm or so and then the rattle is clear as day although never as loud as in the Miatabusa.
Tuesday, July 31, 2012 1:41 PM
here's the picture i was looking for:
notice the wear pattern is different on one side than the other?
you can probably just use a 2nd driven gear cut down to size and use the existing spring pockets to put springs in. i'm not sure how much extra space you have to work with and how much wider the drive gear is than the driven gear (if any)
Wednesday, August 01, 2012 6:18 AM
is there enough clearance in the case to undersize one or both of the gears just enough so they didn't touch and go chain drive instead of direct drive? is there a chain that could handle the load?
Wednesday, August 01, 2012 6:58 AM
a chain would only increase the backlash problem.
Wednesday, August 01, 2012 9:01 AM
Dave: Your problems are more entertaining than most people's solutions!
Pk386: Shouldn't Edison have said, "I haven't failed, I've only found 10,000 ways of OTHER people that don't work for me." ?
Wednesday, August 01, 2012 9:14 AM
^^^ HAHA Yea he did take a lot from N. Tesla and others.
Sidenote: The Toyota split gear solution is intresting. I always thought that the 'groove' in the cam gear was a channel to pick up extra oil. I never thought it was split.
Wednesday, August 01, 2012 10:34 AM
A chain reverses the rotation of the shaft so the car would then have one forward gear and 5 reverse. That's a problem.
I did poke around a bit and found that Honda used a spring loaded split primary gear AND a cush hub on their mid 90's CBR's. They ditched that setup by about '98 and just have a solid gear with a cush hub now. I guess the small amount of rattle at idle is worth the lighter and more simple clutch hub. Also, they are running about 60% more power now so they may need the strength.
Thursday, August 02, 2012 1:51 AM
Dave, given what you know now about how difficult it is to put a 'busa engine in a Miata is it worth it to you?
Thursday, August 02, 2012 9:11 AM
We won't know how hard it is until we're finished, and we won't know if its worth it until we're on the track.
Thursday, August 02, 2012 5:43 PM
Mazda KL-ZE V6 engines have the same split design zero lash gear on the camshafts. I don't think that design would work with something transmitting more than a few HP though.
Friday, August 03, 2012 2:43 PM
I saw your issue and immediatly thought of the split gear solution. A Kawasaki zx-12 used exactly that for the clutch basket. I race a turbocharged hayabusa, but I have worked on a few ZX-12's. I don't think the zx-14's use the split gear.
This link has pictures of a ZX-12 basket:
Friday, August 03, 2012 2:52 PM
Looks like a Hayabusa uas a primary ratio of 1.596 using a 52/83 tooth ratio
According to post #9 here:http://www.esportbike.com/forums/showthread.php?t=102461
A zx-12 has a 52/83 gearset as well. Certainly worth looking into.
Saturday, August 04, 2012 11:45 PM
of course it will be worth it. maybe only kind of until its turbocharged, but 11,000 rpm!
Thursday, August 23, 2012 3:05 PM
y u no update?
Thursday, August 23, 2012 3:47 PM
What? Did you read the story? This is a 3 page explanation of exactly why no update! Re-designing parts!
Thursday, August 23, 2012 6:01 PM
Dave, since we have your attention. can we get an idea as to which direction you're choosing to go to correct the issue?
Thursday, August 23, 2012 6:05 PM
We're going with the factory sprung hub. Suzuki knows best...
Thursday, August 23, 2012 11:03 PM
Awesome. I am curious if the added momentum of the system will allow that solution to work optimally like it does in the bike.
either way, good luck and i can't wait to see the results and the 11,000rpm goodness!
Friday, August 24, 2012 4:45 PM
It's not so DIY, but the simpler solution to this problem is to use the crankcase available from Hartley, the H2. Link: http://www.h1v8.com/H2.html
In the end it will probably end up less expensive that whatever engineering solution you invent then have to have custom manufactured. Sorry to burst your bubble but it's been done , and done better by Hartley. You can even get a V8 version On the other hand, you could pretty easily reverse engineer thier crankcase with a miata compatible bellhouse flange, and mounts for miata ancilliaries. Now that's an idea!
Tuesday, October 02, 2012 7:38 PM
The Hartley Crank case looks to drive straight of the crank which would probably kill the mx5 gearbox with too many revs.
How's the re-engineering going?
Tuesday, October 02, 2012 8:38 PM
Do some research. We did. The Hartley costs 5 figures and has a flywheel speed of 11k rpm so, assuming the trans could handle that, you end up needing a 6:1 rear end to have a reasonable speed range. Not viable for any home brewed street car project. The Miatabusa adapter will cost a small fraction of that and will work with many different currently popular transmission/differential combos. After we get Dave's Miata going a Porsche/VW adapter plate will be next.
Monday, November 05, 2012 5:48 PM
I know I'm way late to post on this, but I keep revisiting these articles while I wait for you to make more progress, however, I might have just had an idea.;
How about a shaft within a shaft, as in run the torque up a hollow tube to the space where the gear box used to live, then back down the center of that tube through a shaft which will act as a torsion spring? It seems that with careful of which materials are used and the diameter of the inner shaft you could get just the right amount of damping. Just an idea, I would really like to get some feed back, just to see if it seems feasible.
Tuesday, December 18, 2012 2:58 PM
Its been many months, and several new project introductions later, and still no Miatabusa update? I certainly understand motoIq's desire to stay on top of the performance scene... hell, I enjoy subies and 370s and toybarus and whatnot as much as the next guy. But the Miatabusa is so crazy, and plausible, that it has more appeal than any of those other projects you guys have been working so furiously on. When will there be some bike-powered-roadster love? Hopefully in time for Christmas? Think of the children!
Wednesday, January 23, 2013 4:27 PM
Any new developments?? We love the miatabusa! please finish it Dave
Thursday, February 28, 2013 9:35 AM
So? Is this just dead now?
Thursday, March 21, 2013 7:08 AM
Just found your description of this build, and read all through it in one go.
I would just love to have access to the resources that you have access to….
I am playing with the idea of putting a Hayabusa engine in a Fiat X1/9.
It would be a version with a 1300 cc engine, so with the bike-engine in place it would be quite “original”.
My preferred setup would be the combined use of the original gearboxes of the Fiat and Hayabusa.
Initially selecting the most appropriate gear on the Hayabusa, and then using the Fiat’s box for the normal operation of the clutch and the gear-selections.
That way I would have a diff, a reverse, and the Hayabusa’s clutch would never be engaged, so it might be able to last.
Any educated guesses from your side?
Another option would be to only use the Hayabusa’ gearbox and clutch, and use clutchless up- and downshifts (using a flatshifter pro from www.flatshifter.com).
Engaging first gear would however still require the clutch to be engaged.
With the original engine and tranny removed, and the bike-engine, a solution for reverse and another diff in place the X1/9 would probably weigh about 1700lb.
This might be a problem for the clutch, but I’m not sure howmuch of a problem.
Any educated guesses from your side?
Can’t wait to see the next update on your Miatabusa.
Oh, one more thing, I am open for suggestions for a cool name for the Fiat X1/9 – Hayabusa project….
Monday, April 01, 2013 10:24 PM
Hmmmm.... I actually really like the X1/9 having owned one in the past and I think it's a really good candidate for a Miatabusa swap but the direction you are going might not work that well. First, I will say that the stock X1/9 transmission will die very soon after you put 180+hp of Hayabusa thunder through it. Guaranteed. The original Hayabusa trans and clutch would work much better save for the issue of the street unfriendly sequential shifter and lack of reverse gear. Motorcycle clutches are really strong and can handle serious abuse. My CBR1000RR powered race car weighs 1800lbs, gets flogged in a manner you would not believe on the race track and has never had a clutch issue. And that's in hundreds of hours of 10/10ths racing.
That all said, using the motorcycle transmission on a street driven car would suck, so the best option for the Fiatabusa "GSX1/9" would be to use a VW/Audi FWD trans (passat, A4) in the rear and mount the engine north south in front of the rear axle. The 'Busa engine is so tiny it would likely fit that way with room to spare and the transmission would be up to the task all day long.
Tuesday, April 02, 2013 12:19 PM
The bike transmissions sequential shifting wouldnt be any more of a bummer than it is when riding a bike. In my estimation, of course. I personally would think that a GSX1/9 with a close ratio sequential 6-speed would be a hoot to drive. Reverse is fairly easy to cobble up.
MotoIQ Proudly Presents Our Partners: