Analysis Paralysis: Miata Based Locost-ish

I'm not great at welding aluminum, so I need to ensure I give myself the best possible chances for success.  Welding an oil impregnated oil pan, well, that's going to be a challenge.  I spent a couple nights this week scrubbing the pan with some scotchbrite, non-chlorinated decreaser and much elbow grease.  It was working, just slow, and I was having trouble cleaning the nooks and crannies.  While questioning my life choices, I decided to call in a favor.

One of my friends is an collector of fine used industrial equipment.  Mills, lathes, other machine tools, excavators, skid steers.. you get the hint.  Well, he has a 2 hp, 20 gallon ultrasonic cleaner that's been taking up space since he doesn't have a good place to use it.  That should do the trick, right? 

Right!

Before - see previous post and the below:

After 4 hours, I present one clean pan!

Great success!  I used about 1 qt of Simple Green to 20 gallons of water.  The cleaner doesn't have a heater, but the vibration heated the 50F well water to 130ish before I turned it off.  It'll probably work faster if I can preheat the water, but that's a problem for another day.

No pictures, but I modified the pickup today and ordered some 1/8" 6061 sheet so I can finish the pan.  Probably won't be able to work on it till next weekend - I need to bring the wifes car into the garage for a full brake job.

I dare say I'm getting the hang of aluminum. It's even water tight! Hopefully I have some time to turn a couple bungs for the pan this weekend.

I ended up cutting more off the pan than I wanted. The leading edge of the sump was the lowest part, so I tapered the pan back towards the rear but lost a good bit of volume as a result. An oil cooler would probably be a good idea to make up for the lost volume and maybe an accusump if I have pressure issues.

I gotta practice feeding filler rod in thru my hand vs holding it and bringing my hand closer to the weld.  Now that I have cleanliness and some technique figured out, that seems like a next logical skill to learn.  Welding is fun!

Well, that didn't work out as initially planned... That said, I didn't thoroughly plan.

I tried turning up some 1/2 npt bungs, but had a heck of a time cutting threads. So. Fine, we'll zap in some -10 AN fittings for the turbo oil drain and the pan drain.

Next up: cutting the -10 PTFE oil drain hose.... Crap, a straight and a 90 deg fitting don't work, so order up a 45 deg.  Crap, that still doesn't really work, the flex line is going to end up being 1-2", way too short.

Fine. Let's do what I should have from day 1 on the silly oil drain: order the FM kit with fittings and the nice flex hose that's surprisingly hard to find elsewhere.

While that's in the mail, let's cut off the turbo drain and tap it 3/8 npt. After playing around with the hose when it gets here, it wants to kink, so we'll escalate again: I'll turn up a 3/8 npt bung, and zap it in on an angle. That gets us to here: I'll call this a success! I definitely made it harder than it needed to be.

Next order of business: figure out how I'm going to modify the dipstick, install the pan fer good.

On the front suspension, will you be adding a sway bar?

In reply to jr02518 :

I'm undecided.  I should be able to add one similar to Caterham - but snaking it around/between the radiator and intercooler will be tough. But I have a solution...

Relatedly, unrelatedly.... I finished the oil pan, oil pickup and..... I'm checking out a ZX10R engine today for a potential change in direction.  One of the things that's long bugged me is that I'm not super excited about tuning megasquirt on a non-road going car, with boost only complicating things.  Also, 12k RPM bike engines sound racy - so there's that.

I found an '07 ZX10R from a dwarf car locally for a very reasonable price.  I think it's ~170hp, EFI, 6 speed sequential goodness, and a slipper clutch.  Downside: I'm going to need to change the differential to something with ~3.2 - 3.4:1 ratio.  Top contenders are E36 or Subaru R160. Although, I'm curious if the hive knows of other compact pumpkins available with similar ratios. 

If I go this direction, I should be able to sell my MS3 Pro and the Miata drivetrain for enough to pay for two bike engines.  No crazy idea's of making a dual bike engined Locost...

Remember how this adventure is called "Analysis Paralysis".... the struggle is real.

I did the thing.

Introducing the likely donor heart: 07 Ninja ZX10R from a dwarf car. The nifty flat bottom pan and driveshaft adapter sealed the deal. 

Like an excited kit, tearing into their Legos as soon as they get home... I just had to see how it looks.

With the pan flush with the bottom of the frame, I'll need a power bulge on the hood, or just cut a hole in it for a foam filter (we'll burn that bridge when we get there...). It's offset to the passenger side about 4.5", but I think that may work out since the driveshaft is going to be about 59" long. I think I'll need a 2 piece driveshaft with a center bearing, and all that puts the u joint angles at 4 deg.

Looking down the trans tunnel:

I guess passengers with big feet need not apply. Narrowing the footwell should be fairly easily with the tunnel only partially welded.

Besides staring at it and dreaming: time to do some planning.

• Need a header. Not sure if I want 4-1 or 4-2-1. The factory headers are titanium and I'd rather not mess with trying to modify them to fit. There's a slim chance a dwarf car header could fit, otherwise I'm fabricating my own.
• Engine mounts
• Wiring. The harness was hacked for a race car already, but I need to figure out what they did before in put power to anything. I may pick up an unmodified harness and ninja cluster.
• Clutch - my chassis is set up for a hydraulic clutch. Building an adapter to allow a clutch cable shouldn't be hard.

I think mounts will be the first step while I noodle on the other stuff.

Exposed filters are a Seven thing. Nothing wrong with it at all :)

Keith Tanner said:

Exposed filters are a Seven thing. Nothing wrong with it at all :)

I do love me an exposed filter on a 7! Have you ever had an issue in the rain with yours?

No, a bit of moisture isn't a problem. It takes more than rain to bother an intake. The driver will give up first :)

The last couple days have been spent staring at the engine, wondering, dreaming, maybe making 12k RPM sounds.

The motor mounts are going to be kind of funky, but funky could work.  To be honest, I'm not super thrilled with my idea so far, but I figure I'll start with the lower passenger side mounts first and move up from there.  

Here's what I'm thinking:

• Orange: New or moved chassis tubes, will be welded in.  I may make a couple trapezoidal plates to mount the lower mounts.
• Green: I think these are also going to be welded in, but will have a sleeve on the motor end where I'm using control arm bushing as a motor mount.
• Pink: Removable mount.  I may make the similar mount on the rear side of the motor permanent.

I know this is more mounting locations than I'll need... In theory, 3 mounts should be sufficient, but it's going to react engine torque with fewer mounts since the support angles for the drives side mounts leaves something to be desired.  Reacting all the engine loads through the passenger side seems like a bad idea, and not the "that could be fun" type of bad idea.

I should probably lay out my shifter to make sure I'm not painting myself into a corner.

Im happy to see this. I think about it occasionally when i think about the Europa im glad i didn't get.

Ha, I'm glad I sold the Europa, and kind of wish I both sold it and changed to the bike engine course on my Locost years prior.

Differential: I think I'll go with an E36 diff.  The local salvage yard has an open, 3.15:1 diff for about $100.  It's far beefier than I need, but weighs about the same as the Miata rear end.  Hopefully the ears package reasonably well - if not, I ought to be able to sell it for little lost.

The backup plan is a small Subie R160 differential. LSDs are harder to find should I decide to go that route in the future, and there's a chance the diff may not be up to the task.  I'm not putting much torque through it, but I'd hate to go through all the work to find out that I should have just taken the 10-15lb debit up front.

Driveshaft: I need a long shaft... Real long, about 59" give or take, that's fixed at both ends - so it'll either need a slip joint in the middle or a center bearing.  It'll also need it to hold together at 6000 RPM.  All of that steers me towards a 2 piece driveshaft.  BMW and Subie come to mind again, with the BMW having the benefit of a guibo to help dampen some of the drivetrain harshness.

Problem being: from what I can tell, E36 shafts are too short.  I could see if I can get a longer section, or custom half with the right end for the bike engien.  Cursory research shows E39s use a CV joint on the differential side, coupled with a bigger diff - so I'd need an adapter.

Alternatively, Subaru Outback/Legacy shaft looks like the perfect length, but will it be strong enough?  The bike only makes 80 ft-lbs torque on a good day, with peak driveshaft torque with all the trans gear reduction at 330 ft-lbs.  If the Subie is a 60:40 front-rear split and makes 200 ft-lbs, with a 3.45:1 trans gear reduction, there'd be ~300 ft-lbs torque going through the shaft from the factory.... Interesting.  I'd need to machine up an adapter to mate it to the E36 diff.

With either shaft, I'll need to convert the forward end to mate to the bike engine's driveshaft yoke. 

Am I over thinking it?  Should I just get the Subie driveshaft and diff?

Just remember that drive shaft shops can lengthen drive shafts as well as short. If the BMW fits the bill otherwise call your local drive shaft guy and ask him if they can lengthen it for

Thanks Michael, I'll call around and see what it'll cost.  The custom driveshafts I was checking out online in my caffeine induced sleep insomnia last night were eye wateringly expensive.  Hopefully the local guys are cheaper - that'd also be a good opportunity to swap out the front end to mate to the bike engine so I don't need custom adapters.

Mike ice down here if you bring them the end can generally do new joints a new shaft and balance it for $300 or so

I was wrong, E36's shafts aren't too short.  I was looking at the length from an auto, but the manual shafts are just about perfect!  I'll need to machine up an adapter for a guibo and that ought to put the engine where I want it.

Interestingly, E34 525's also have the same diff as an E36, but a more compact/adaptable mount.  They also feature a really neat tripod joint at the pinion input, great, right? Well, close.  Manual 525s have the ideal ratio, 3.23:1, but they're pretty rare.  Oh well, the E36 3.15 will have to do.  I may grab an E34 cover.

In the interest of not having to modify the driveshaft, I'll machine up an adapter for a guibo at the bike's driveshaft adapter.  I love it when you can play legos and it appears things all bolt together.  I've got a line on the differential I want, but it doesn't include the output flanges.  Time to scour FB and CL to see if there's anyone local.  I think the updated order of operations is to get the new diff mounted, cut trans tunnel, install driveshaft, machine guibo adapter (I guess I can do this first if I have the material), then motor mounts.  Woohoo, we got a plan!

E34 cover:

Hypothetically perfect E36 driveshaft

Much head scratching this week, with the objective being: lock down on a drivetrain permutation that'll work and in general, I think success was had.

Sometimes I just need to see something in person to rule in or out... And the E36 differential was just that. A buddy LS swapped his e36 and had his old open diff lying around: perfect. As soon as I set eyes on it I realized its both overkill and unlikely to fit, but I brought it home to make sure. Sure enough, it'll only fit with major surgery. 

Next up, head to the local junk yard to set eyes on an Outback rear differential. I've caught glimpses under the skirts of varies subies over the years, but never up close. It's cute! Behold: an outback 3.27 viscous LSD, compared to NA Miata and E36.

There's a chance it'll grenade, but we'll see. The viscous LSD isn't great either, but we'll burn that bridge after we roast the diff.

Please excuse the rapid fire self responses... I've been burned a couple times with failed posts.

Next chapter: connecting the dots... Driveshaft.

The snout on the Subie diff is much longer than the Miata, so now I'm on the hunt for a 50-52" driveshaft.  Rockauto was great to cross reference various driveshaft lengths. Of course, the Legacy/Outback is too long. The Impreza may be OK if I can swap out the original yoke for a flange. Cue hours of research to find U-joint dimensions and scouring the internet for a suitable flange. No dice. The backup plan is that I could buy 2 junkyard shafts to steal the flange end off one to replace the original trans yoke on the other. It'd work, but seems clunky and could be out of balance - maybe? It'd also be 2-3" too long. 

Another option is the BMW parts bin. I like the idea of a guibo at the front end of the driveshaft to potentially dampen some of the bike engines liveliness.  In keeping with German automakers traditional, there are a ton of unique parts depending on year/engine/trans/chassis/awd/phase of the moon. Eventually I found the E46 325xi auto to be close to what I want. Off to FB marketplace and now I've got a shaft.

Fire up fusion, melt some plastic over night and I have driveshaft adapters for both ends.

Now to move everything around to figure out the drive line angles and mounts. The passenger footwell is going to need to get smaller, and I may swap the front section of this shaft for one that's a little longer so there's more room for the exhaust. 

Does anyone have experience with funky angles in a 2 piece driveshaft? I'm not sure if what I've laid out will be OK or shake like hell 

The 3d printer has been busy this week. Starting to lay out the header.

I think I'll try rotating the differential so pinion flange plane is parallel to the driveshaft flange on the engine. I suppose this is when a guibo is cumbersome to work with.

Ref GRM thread: https://grassrootsmotorsports.com/forum/grm/is-this-bad-driveline-angled/159448/page1/

Lasers!

Aligning the differential and engine was getting annoying with all the tubes in the way. Prusa to the rescue again!

3d printed an alignment tool that fits my laser pointer.

Clamp mirror to the driveshaft adapter on the trans.

Fire up the laser and kill the lights!

I love the visual indicator, all I need to do is get the reflection aligned to the source and we good! Admittedly, it took some fiddling to get the laser centered since the print warped. Rotating the pinion, the pointer traced a circle. I tried dialling it best I could. But since I'm really struggling with the left-right alignment of everything, turning the pinion so the dot is at top dead center of it's arc ought to be a reasonable approximation for the vertical axis of the centerline. I'll use a level to finalize the tilt.

So the less exciting news: it's looking like the BMW shaft would require significant hacking of the passenger footwell, or sliding the engine about 2.5" more to the driver's side which puts the header exactly where the steering shaft needs to be. Decisions. I didn't really appreciate how stiff guibos are and that the first shaft needs to be almost straight off the engine. 

Sometimes I just get lucky, and this was one of those weeks. Combined diving thru the 90s Japanese parts bin, and a couple off shelf of flanges got me a driveshaft.

I started with a cheap late 90s Impreza shaft which was too long. While laying out mixing and matching Subaru flanges, I noticed the middle flange on the Impreza shaft matches the flange on the bike engine! With that, a plan is born. We'll swap the differential flange to the shaft with the center bearing, use the stock Subie mid flange, and a 4wd driveshaft that's 24-26" long that hopefully has the right flanges. Apparently the flanges are commonly used by Nissan and Subaru.

After much research, I found that Nissan uses standard 1310 u joints, and I can buy the flanges cheaply if they aren't already on the shaft I find. So, let's check rockauto for shafts that use those U joints in the right length range, available for cheap at the local junkyard.  Behold! They have a $40 90s Nissan truck driveshaft on the shelf. Unfortunately the flanges don't match what I need, but that's OK. Maybe an older Nissan/Datsun shaft would have them, but new joints and flanges are cheap.

Fast forward thru tearing the various driveshafts apart, grinding the annoying stakes out of the Subie stuff, cleaning, paint, pressing in new joints, welding diff mounts gets us here:

Next up: front diff mount, engine mounts. I'll need to get the shaft balanced at some point.

The last couple weeks have been productive, but this is also the type of building I like to do.  After staring at the car, wondering why it isn't building itself, I got about building motor mounts.

I'd planned on making the mounts bolt in, but couldn't settle on a simple, light weight option. So, let's just try not to grenade the engine right away. *Makes note, I should start piecing together an accusump....

I think I'm going to make another mount on the forward side of the cylinder head since the engine is cantilevered out a bit.  Some bike engines have mounting lugs on what'd be the drivers side in the car - if I ever swap to a different manufacturer, I'd look more closely at mounting options.

I tapped the forward lower, right mount, and modified an M20 bolt in the corresponding upper location to act as a mounting bung.

TIG welding is fun, especially when you're not in a hurry. Next up: finish printing up the header Legos, then start laying out an equalish length header. It'd be awesome to start this up this summer.

Keeping with the theme of the build, I may have overcomplicated things. But, let's not get ahead of ourselves. Here's the final iteration of the equalish length header. All the tubes are within an inch of each other, so that'll do.

One nice thing about this collector is that I can try different cylinder pairings. It's currently set up as 1-4,2-3, but I can rotate it and get 1-2,3-4 which is how Kawasaki set up their header.

Now to figure out how to make nice square cuts on the header tubes. With how visible the header is going to be, I want it to look nice. The removable table on my HF horizonal/vertical band saw has always been off square. It's fine for thin sheet, but annoying for anything bigger.  It cuts really nice in horizontal mode, but vertical is a compromise. I tried to shim the table square, but wasn't happy with how unstable it felt.

The answer is naturally a need for a new tool, right?  My dad has an imported 14" wood band saw that I dragged home to scope out what sort of speed reduction I can get with a jackshaft. Adding up the cost of everything and the fact it's still too fast for metal, and a VFD plus a 3 phase motor suddenly makes the free bandsaw pretty expensive. I mentioned the dilemma to a coworker who learned me that Delta used to make a vertical wood and metal cutting bandsaw, and a couple days later one popped up for a reasonable price nearby.

Enter this gem from 1965

It's in pretty good shape, despite the awful paint work and hacked base to fit a massive motor. The internal gearbox is pretty cool, I can get blade speeds from 80-2700 fpm. Swapping between wood and metal speed can even be done without pulling the belt off. Time to test it out while making a tube holding fixture/guide.

I think I'm going to like this saw, but it's probably time I sell some other machines to reclaim floor space. Hopefully I'll get some more time to start building the header soon.

Nice Work!

Slight progress made this week. Hope to get out again this weekend and work on another runner.