1 2 3 ... 8
Knurled.
Knurled. MegaDork
1/22/18 8:06 p.m.

I, er, had been meaning to start this thread last year.  But then Photobucket went TU and when we eventually got Photoberkit, I found that I couldn't upload my images there due to size restrictions.

But that's, literally, so last year!  The weather got not-bitter-cold and I made it out to work in the garage.

 

 

Leftover from the 2 day rallycross last year.  This may constitute beer abuse!

 

First I scratched my head at the '81 RX-7.  Its exhaust is all plumbed in, with a Magnaflow muffler mating the 12A exhaust manifold to the ??? exhaust system.  I do not want a header on what is supposed to be a pleasant car.

 

But here is the problem:  I want to use the 12A's choke cutout switch, which threads into the water pump housing.  The FC 13B water pump housing has a coolant temp sensor there with a wildly different thread, and it does not appear to have enough material in its boss to tap it out to choke switch size.  I rather LIKE the 12A cars' automatic manual choke.  We'll see.

 

But enough of that!  On to the 1984 car!

Knurled.
Knurled. MegaDork
1/22/18 8:15 p.m.

Enter one (1) complete GSL-SE rearend, and a large piece of cardboard.

 

 

Get out a square and a frozen felt pen and let's find flange face and center for the pinion flange:

 

And the axle ends:

 

 

And mark locations for the Watts pivot, the lower link centerlines, and the spring seat centerlines.  End up with a barely legibly marked up piece of cardboard:

 

 

Now let's drop an anchor on it!

 

 

Subjectively, the 9" is actually lighter than the Mazda rear, although it doesn't have the brakes mounted on it yet!  Really must get a scale and soon.

Knurled.
Knurled. MegaDork
1/22/18 8:27 p.m.

Line it up evenly over the axle centerline/face marks I made on the cardboard:

 

 

1/2" wider per side.  Just as planned.  With the stock width rear and 40mm offset 15x6 wheels and 185/65 tires, I'm rubbing the inner fenders every now and then.  I wanted a little more width to center the tops of the wheels better when cornering.

 

And the reason above all other reasons why I wanted a 9", moreso than the gear availability, moreso than the cheap cheap take-off diffs available on eBay, even moreso than the availability of ultralight parts:


 

With a "Thunderbird" flange modified to work with the Daytona pinion housing, and machined lightly to accept the Mazda driveshaft, it is only 3/8" longer than the Mazda diff.  This means I will be able to use a stock driveshaft!  This is KEY because a custom driveshaft is stupid expensive for imports, and I go through them on a regular basis.

 

However, I made a slight error...

 

I did want the pinion offset to the left relative to stock.  Stock, the pinion flange hits the right side of the tunnel.  I incorrectly had measured the stock rear to have a centered pinion, when in actuality it's about 15mm off to the right.  So my 1/2" offset to the left means I am around 1 1/8" off from stock.  I may need to do some minor floor surgery, maybe.  The pinion is also a lot lower than a Mazda so it may all work out okay.  For certain, I will have a ton more exhaust room under the car, the driveshaft currently rubs the rear presilencer sometimes.

Knurled.
Knurled. MegaDork
1/22/18 8:45 p.m.

I did make a significant error elsewise.  Last year, or maybe two or three years ago, when I was working on the bare housing at Evan's place, I carefully measured a 3 degree nose-up angle and welded the link mounts in so that the mounts were centered under the rearend.  I want to be able to muck with instant center without having to reset the pinion angle, you see.  When I was doing this, I mis-measured where to put the mounts, and the reason for this whole escapade with laying out centers on cardboard:

 

Driver's side:  Way outboard.

 

Passenger side, inboard slightly:

 

 

I'm at an impasse here.  I did buy two new link mounts.  But part of me says fix the left side to match the right, and part of me says to fix the right to match the left.  I wanted parallel.

 

I did deliberately mount the 3rd link a half inch off to the left, so that axle torque drives the right side down slightly.  This helps offset the driveshaft trying to lift the right side up.  I can adjust how much effect there is on top with different width rod end spacers.  I can't do that on the lower links because I have UMI joints that need to spacers to work with these mounts.

 

So, when the going gets tough, we... go do something else.  Let's mount those spring perches!  Enter two Allstar clamp-on spring perches, with the clamp-on part removed with an angle grinder.  I only took pics of one side.  First we center it to where it was located on the Mazda rear with the square:


 

Then we check it for level.

Make the noise!

 

Then we fire up the MIG and see how badly we can mung this up with flux-core because we're too cheap to buy an argon tank yet:

 

 

Not too bad, actually!  I only got the tops so far.  I don't want to lift it from the template until I know what I want to do about the lower links.  After I move it, I'll weld the legs to the axle tubes for a little extra overkill.

 

And that is all for today!

jfryjfry
jfryjfry HalfDork
1/23/18 7:02 a.m.

Seems like you know the right answer is to move both and since your other options involve moving one, I'd say suck it up and just do both - once you do one the second will be faster. 

Besides, if you only do one, if you're like me, you'll be kicking yourself later for not just doing them both while you were at it. 

Just my $.02

fidelity101
fidelity101 UltraDork
1/23/18 8:28 a.m.

ive been waiting for you to make a build thread for the variety of things you do, this 9" is quite interesting.

Knurled.
Knurled. MegaDork
1/23/18 7:42 p.m.

In reply to jfryjfry :

Well, I don't necessarily WANT them to be parallel...

I mean, I think there are advantages to having all three links converging on one point in the chassis.  The complicating factor is that I will not be able to move the passenger side link mount outboard without a LOT of work since it is already at the edge of the larger diameter section of tube.  Move it outboard less than a bracket's width worth and it will be cockeyed, without a lot of fiddling that frankly I don't want to do.

But then I found a plan view image of the SN197 and its lower links are wider at the chassis than the axle.  And moving the driver's side mount would be a lot easier.

 

What I need to do, is find my dang copy of Puhn's book and go over the geometry advantages and disadvantages.  All rear links will be rod ends so there should effectively be no deflection.  (Deflection of the Panhard bushings, for example, would give the SN197 some lateral-load rear understeer, which is confidence inspiring)

 

I want something easy to drive.  I am not a good driver, so I need to account for that wink

 

Today:  Expected temps in the 50s, left warm gear at home.  Was dropping near freezing when I left work.  Nope.  Went to Summit instead, saw nothing of immediate or value except a 25 foot roll of 5/16" fuel line and some clamps that I need for various things.  Valiantly passed up the $25 pair of 4x4.5 bolt pattern 15x8 stock car wheels because they were 2" backspacing.  I forget what backspacing I would need (converting from backspace to offset and back is tricky because backspace takes the rim lip into account and offset does not) but I know it is way more than 2".  I think it's 4 or 5 inches.

 

Knurled.
Knurled. MegaDork
1/25/18 6:21 p.m.

This rear is built with the axle length for a 60" wide centered-diff on the right side, and a 58" centered diff on the left.

 

Part of me says, this thing is all berked up, just suck it up, go another $1500 in hock, and buy a 58" CENTERED rearend housing from Speedway with full-floater hubs like I wanted in the first place, ALUMINUM hubs, and aluminum as much as possible.  Cut losses, do what I want to begin with.  Then throw the rest of the crap in the RX-3.

 

Part of me says, put THIS bodge in the RX-7,  get the awesome floater for the RX-3.

 

Part of me says, holy E36 M3, I own an RX-3.  And two RX-7s, and a 300hp AWD techno-wizard masterpiece with comfy seats and kickass ICE and da Bremboez.  Why am I not giddy with excitement?  Is this just what it takes to get me to normal?

 

Anyway, Bosch 1200cc injectors for the '84 are on order.  After the recent unpleasantness with the RC injectors that died and resultant re-installation of the 680cc injectors that I used to run before I made some significant upgrades, I found that I have a horrible lean misfire that brackets the torque peak, which made me sort of park the car out of frustration.  At work today , Da Boss casually came up and mentioned that we got a whole set of 1200cc injectors from a supplier because a customer burned a piston and didn't want to muck around with having the old injectors cleaned.   (Actually, he kind of opened the Fedx package and waved them in my face, because he knows I am impressed by high performance stuff) Knurled says, hey can you order me two more, since Da RX-7 needs more fuel?

Hmm, maybe I should just cancel the order and buck for the set of "bad" injectors and send them out for inspection/cleaning, instead.  I only need two, not six.  Hey, these build threads are useful besides self-promotion!

Knurled.
Knurled. MegaDork
1/28/18 8:29 p.m.

I MAY have figured out how I am going to laterally locate the axle.

 

The housing is already set up for a Panhard, and even on the correct side, but due to the fuel tank, shock absorber, muffler, and exhaust pipe locations, I can't fit a chassis side mount anywhere near where it would have to go.

 

The Watts will be tricky but doable.  The problem is that the Watts mounts to the front, and the driver's side link runs over top of the pumpkin.  Not only is a 9" a lot larger pumpkin-wise than a 7" but it also sits a lot closer to the right, getting in the way of the bellcrank.  As a bonus, Mazda stuck the bellcrank pivot 40mm above axle centerline and I want it a lot lower than that for handling reasons, which puts the top link further into the pumpkin.

 

After much measuring, figuring, measuring some more, and head scratching, I think I have a solution.  I will mount the bellcrank on axle centerline, slightly more forward and about 15mm further to the right than Mazda did.   The upper side of the bellcrank will be extended 80mm (need to verify this) so the link will clear the pumpkin at full compression.  There are hard limits at work here because  the Watts is already the first point of interference (with no pumpkin in the diff, anyway).  I'll install bumpstops, currently have none.  The next problem is the 3rd link is mounted as high as it is on the axle specifically so that it will clear the Watts link at full compression.  If need be, I will mount the 3rd link higher, but I'd rather not if I don't have to.

 

Since I don't have any spare bellcranks, it looks like I'm going to have to buy a new one to hack up.  Mazdatrix sells a bellcrank and bellcrank reinforcement as a package for $150 or so.  Urk.  Add it to the pile of receipts.

 

 

Knurled.
Knurled. MegaDork
2/1/18 6:28 p.m.

Injectors are on order.  Wussed out and went with brand new known factor injectors.  1200cc/min is, interestingly, 114.3lb/hr, and 114.3 is also the wheel bolt circle when hammering a metric number on something designed in SAE.  (Well, Imperial, actually, I think)

114.3 lb/hr at 43.5psi is 104 lb/hr at 36psi, which is important because I am using a GSL-SE rail and Mazda gave it a 2.5 bar regulator.  So that is 208 lb/hr available fuel, which at .5 lb/hp/hr BSFC, means I would have enough fuel for 400hp maxed out, or 300hp without burning out injector drivers and still have the ability to flood out the engine at 10,500rpm, which is my official "oh berk dawg, don't go any higher" RPM.  I've seen over that and I am glad I wasn't datalogging because I don't want to know.  Engine speeds with a comma in them are not places you want to hang around very long.

New Watts bellcrank and reinforcement have also been ordered from Mazdatrix.  It's so nice to buy a bunch of expensive stuff with the knowledge that I will have to hack it up and modify it before it is usable.  And hope I get it right the first time.

 

I also dismounted the tires that wouldn't hold air from the RX-3 wheels.  They are rusty and rusty and rusty some more.  One of the tires would not dismount.  The bead would reseat immediately after using the bead breaker.  I got fed up and said, dammit tire, I am not ASKING, and used a Stanley knife to cut the 165B13 Goodyear's sidewalls off of the tread.  That is when I found out it had an inner tube.  So I finished making a tread patterned belt, peeled the inner tube off, then dismounted the two sidewalls from the extremely rusty wheel.  Need to do something about that rust.

Knurled.
Knurled. MegaDork
2/1/18 6:54 p.m.

Earlier, Evan aka Han taught me the ways of license plates as units of measure.  I felt the need to properly quantify the license plate, or lp, so as to standardize its use.  To this end, I acquired a genuine, government issue, regulation sized, official License Plate, and quantified it.


 

Officially, I will define one lp as exactly "a tad under 309 millimeters".  It appears to be exactly 12 inches and some hash marks, but my eyes get lost trying to count if that's a quarter or three eighths or whatever, and the inch is defined in terms of the metric system anyway, no doubt as a thank you to our French friends who helped us out.  And given that the soil where I was born and raised and currently live used to be part of France, I feel very well at ease in defining this unit of measure in metric as a further thank you.  (My home land is very nice.)

 

 

This is measuring a Harbor Freight 2-ton engine hoist.  From the tip of the slidey thing to the bolt that holds it in is 780mm, or 2.524lp.  I had originally measured it as 2.5lp.  Clearly, the license plate method of measurement is superior in its easily accessible granularity!

 

 

Similarly, from the slidey bit's outermost usable bolt to the pivot bolt is 770mm, or 2.492lp.  Realistically, do we really need to know down to the hundredth?  It's about two and a half license plates.  All we care about is what dimensions are wildly different between one hoist and another, we're not trying to set valve lash or anything.

 

 

 

 

wvumtnbkr
wvumtnbkr UltraDork
2/1/18 7:41 p.m.

Most engine speeds involve a comma..  like 1,000 rpm.

 

I get it though, 10k is baller!

Knurled.
Knurled. MegaDork
2/2/18 7:11 p.m.

In reply to wvumtnbkr :

Style guide sez you don't use a comma for single digit thousands.  1000 no, 10,000 yes.

 

Today:  NOTHING! ABSOLUTELY NOTHING!

Knurled.
Knurled. MegaDork
2/5/18 7:16 p.m.

Around 11am today, I went to my workstation and found that someone put a UPS shipping package on my desk that was covered with Mazdatrix stickering.  Hooray!  The shipping estimate was for the 7th.

 

Around 2pm, Da Boss comes up to me while I was sorting some stuff at my toolbox and brandishes a pair of Bosch pencil type fuel injectors.  Hooray!  My cost was $94 a piece, including a whole bunch of adaptors that I may or may not need.  They have the correct EV1 connector but I still need to look up what the actual latency specs are, as they literally just came as a pair of injectors in a vacuum pack.

 

Went out to the Batcave to start mocking things up.  No photos, and anyway Photoberkit needs about four attempts per photo before I can get one uploaded, and i'm still uploading reference images that I took this weekend of my latest acquisition.  And the stuff I was doing was not very photo friendly.  My plan is to mount the bellcrank on center instead of 40mm high.  Because the 9" pumpkin sits so much further to the right, however, the lateral positioning of the main pivot will be crucial for keeping it from fouling on the pumpkin.  I crawled under the '84 and measured the lower Watts-to-pinion shaft (a hardpoint I've marked out) to be 100mm.  I then marked that out on my measurin' cardboard, put the bellcrank up to the 9"  with the lower mount inline with that mark, and did a good amount of eyeballing.

 

I am definitely going to have to do some grinding on the pumpkin if I don't want to make the Watts links adjustable.  I can make them adjustable, but I'd prefer not to.

 

After I hemmed and hmmmed for about ten minutes, I put my new pivot bolt (GM thought it was an LS crank damper bolt) where I determined that I probably wanted it to be, and discovered that my welder wouldn't feed.

 

It is too cold to weld in the Batcave.

 

I made a few ugly looking tacks, at which point my fingers were numb from the cold, and I said to myself, self, if you're not going to get any useful work done on this rearend, and your hands are too cold do to anything but send pain signals to your brain, that is a sign that you should be at home.  So I went home.

EvanB
EvanB MegaDork
2/5/18 8:03 p.m.

It's much easier to upload photos to the forum as you write the post.

Knurled.
Knurled. MegaDork
2/5/18 9:09 p.m.

That doesn't help me cross post to other forums though.  And forums on the tablet sucks, and I still have yet to buy a new digital camera since my $50 GE is too scratched up to take good photos.

 

Photobucket is whining at me to give them $99 and they will allow hotlinking again.  I think I spent that much originally when I started with them, so I might go for it.

Knurled.
Knurled. MegaDork
2/8/18 6:55 p.m.

I have given up on trying to make the bellcrank work with the stock links.  So, I will make the stock links length-adjustable!

 

I think the easy button, with "easy" defined as "not terribly expensive and not requiring a ton of research", is to splice in some cut in half threaded sleeves, and install adjusters.  Here's the idea:  The Watts links are .689-ish OD.  I can get some AFCO 5/8-18 threaded tube (LH thread one end, RH thread the other) in nominally .875 OD and .095 wall thickness (which is close enough to .689 ID)  for $13 each from Summit.  Cut that in half, cut some out of the middle of each Watts link, splice the halved AFCO links over the Watts link stubs, and now we have left and right hand female threaded rubber bush ends for $26.  Allstar makes a 5/8-18 double adjuster that is also $13 each.  So for $26 a link or $52 total, I can adjust the axle centering as well as the bellcrank position.  This makes life a LOT easier since the bellcrank position is going to be a major PITA to gt right, and I feel it won't be get-rightable without changing the link lengths anyway.

 

 

Knurled.
Knurled. MegaDork
2/12/18 8:32 p.m.

Went to Home Depot and got myself some more "camping gas" bottles for my mini heater, and some 2"x1/8" flat stock, and got down to bidness.  Got two hours of work in, left at 8:30 mostly to keep from annoying the neighbors.

 

Let there be heat!

 

I measured the actual difference laterally between original Watts pivot and where I eyeballed the positioning of my new point, and it is 40mm coincidentally enough.  So when I hack and slice the links, the right link will need to have its adjuster at its midpoint with the link 40mm shorter, and the left link will need to be 40mm longer, although that side will need to be more flexible in length since I will also be modding the bellcrank.

 

But first!  We need to flip this housing over and finish welding that bolt in, and cut off some of those ears.

 

 

Mmmm.... spattery.

 

At this point I tried to think of a decent way to reinforce the bolt's mounting.  Failing that, I decided to just throw metal at the problem.  I measured here, measured there, then found that my NiCad 19.2 battery was down on charge, and my Lithium battery was completely dead because the Lithium-capable charger died.  Ugh.  While I had the NiCad thrown in the other charger to get a little bit of oomph so I could drill my flat stock, I used the recip saw to cut the left side link mount off.  I probably won't grind th old weld remains off, as a reminder to be more careful in the future.  That took a long time and two blades.

 

Battery was charged enough to run a bit through and then the step bit to open the stock up to 5/8".  Measure the height of the bolt, stick the stock in the vise, tap it over until the leg is the head's height away from the vise jaws, slip it over, and we get this:

 

 

Weld in the middle, tap tap tap, weld some more, tap tap tap, and thus the ends get contoured to the axle housing. 

 

The bolt head gets welded to the underside as well for belt and suspenders.

 

 

Mmm.... spattery.

 

The next bit to do, that being closing up the top and bottom, is going to involve a lot of fussing and grinding, and I didn't want to start on that at a quarter after eight, so I left it here for the day.

 

My goal is to get this thing ready for paint by Wednesday-Thursday-ish, since the temps will be up into the "paint might actually dry" region.  Still have no idea what color to paint it.  White makes the most sense, but there'd be little contrast to the aluminum bits.  Black offers the most contrast but it doesn't stand out otherwise.  I'm probably going to GO with white because I have a bunch of it, and the wheels are white anyway, but there's a nagging desire to go with some ridiculous lurid green or orange or something.

 

White shows cracks and leaks better, I'll go with white.

Knurled.
Knurled. MegaDork
2/14/18 7:07 p.m.

Yesterday and today were kind of off days for a couple reasons.

Yesterday I did finish boxing in the Watts pivot superstructure:


(don't ask me why this is vertical, 'cause I thought I had the tablet set to remain in landscape mode all the time)

And then I shuffled a bunch of stuff around to scientifically prove, PROVE, that 3mm seal rotor housings and 2mm seal rotor housings were the same, because I set them up against each other on dowels and they wee identical, and gosh dang it someone on the Internet was wrong !!!

 

And it turned out the person who was wrong was me.

 

 

Well, hell, if I'm wrong about something, I'll admit it.  I sure would like to know why when I did this in 2011 everything lined up perfectly.  Full thread is here for more pictures and engine buildy stuff.

 

Tonight, I spent a good amount of time straightening up and organizing and then I moved teh RX-7 around backwards in the garage, starting it for the first time since November or December or something.  Despite the battery having been on charge for 24 hours, it cranked kinda slowly, and then it grumble-brapped itself to life.  No video because who cares.

I'm at a standstill because I don't have a good way yet of supporting the vehicle high enough to work under it comfortably for the kind of stuff I have to do.  I did try sticking my cherrypicker on the rearend and pulling it up, but it bottomed out against the hatch sill before I got it much higher than I could have gotten it with the floorjack anyway, and then I still had the suspension compressed.

 

Reference photos, this is my current rear suspension setup as seen through the hatch, with the cover boot removed.

 

 

Note the distance from my 3rd link to the upper Watts arm.  The 3rd link's height on the rearend was determined by a clearance issue with that arm at a certain suspension position.... and I am going to need to raise that Watts link 40mm or so relative to current in order to clear the 9" pumpkin, I guesstimate.  (Because I also moved the Watts pivot down 40mm, I need to extend the one end of the bellcrank 80mm)

 

Also notice that the chassis is sort of cracking and falling apart on the driver's side there.  Need to reinforce that better somehow.

Knurled.
Knurled. MegaDork
2/15/18 8:25 p.m.

Today: Scrubbed uglies.  Meaning, I got intimate with my rear end, using a three foot pole and a wire brush.

 

Then I bought yet another pair of Craftsman jackstands, figured out how to jack the back of teh RX-7 up with the engine hoist in such a way that I could lower the boom agiain.  It involved some unsafe things like a floor jack under the cherrypicker, so no pics.  Don't try this at home.

 

Old and busted is almost ready to come out.  Brakes removed, links loosened, etc.

Knurled.
Knurled. MegaDork
2/20/18 8:06 p.m.

 

Confused?  You will be, after the next episode of Tales from the Batcave!

cghstang
cghstang Dork
2/20/18 9:37 p.m.

This thread needs more comments from the gallery.

I love this car and all the scented brap.

Knurled.
Knurled. MegaDork
2/24/18 6:02 p.m.

I have decided some rules for my posts in this thread.

1.  No chooch.

2.  No skookum.

3.  No biscuits, although tea is acceptable.

4.  There is NO...  rule 4.

5.  There will be alloted time to get the funk out.

 

Knurled.
Knurled. MegaDork
2/24/18 6:41 p.m.

Progress:

Bent and split/leaking, twisted axled, open-diffed to save the axles, grinding brakes because GSL-SE calipers are NLA from Mazda or any remanufacturers and the calipers are fully berked, Old and Busted out:

 


Some dumbass installed a rigjob fuel swirl pot/surge tank in such a way that the only way to access the upper Watts link nut is with an open end wrench.

 

Lower Watts link is 33mm center-center.

 

 

What the orientation of the Watts is, as seen from the front of the car.  Note stylishly untied leather, insulated Chucks.

 

 

Old and busted, meet new hotness.   Note how much further to the right the Watts pivot is.  This will need to be addressed.

 

 

Caliper brackets are welded directly to the backside of the axle bearing housings, so the T-bolts will now be too short to go through them.  I made some overlong T-bolts by hacking a point off of four 1 1/2" long 1/2-20 Grade 8 bolts.

 

 

Here is my Plan for making the Watts links adjustable as well as altering their length.  I accidentally deleted the pics I took of the raw materials, but what we have here is .875" OD, .095" swage tube from AFCO that has been cut in half, and 5/8" double adjusters.  I figure that 10mm of threads showing per side is a good starting point, as this is roughly midway between the adjuster being buried on the short side, and one diameter of threads engaged in the tube on the long side.  So this gives me +-20mm of length adjustment.

 

 

ID of the AFCO links is perfect match for the OD of the Mazda links!

 

 

Measure, cut, cut, a little filing, and we have a lower link that is 40mm shorter:

 

EvanB
EvanB MegaDork
2/24/18 10:33 p.m.

New axle looks pretty skookum...

1 2 3 ... 8
Our Preferred Partners
X30xAj8Y9M1E73Hb8v3NCknZ5ENwbNcmvnYVB68EV3EXMgO0Ad8XYmM6UKxOBqA3