1 ... 30 31 32 33 34 ... 54
Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
5/27/20 6:36 p.m.

Yesterday was hot as berk, nothing got done

 

Today was hot as berk's foyer, so I did the intake manifold after work and then took off because the Fit has air conditioning and the Batcave does not.

 

Whatever intake manifold this is cuts like Velveeta compared to the Racing Beat manifold, which had the hardest aluminum I have ever experienced.

 

Channel is cut between the two injector air bleed holes, drill a hole centered in the channel, drill a 15/64ths hole down to intersect that first hole, press in a piece of 1/4" brake line, and a carb manifold is modified to supply air to the fuel injectors!

Problem is, I drilled it 1/4", so the brake line is a rattle fit.  I put a very slight flare on the end and then pressed it in, with some epoxy to hold it in place.

 

A lot people used to eliminate the air bleeds because they got scared by a vacuum hose when they were little or something.  I have found that they are super important for getting the idle as smooth as possible with minimum richness.  You will never idle a rotary at stoich but these are worth quite a bit.

fidelity101 (Forum Supporter)
fidelity101 (Forum Supporter) UltraDork
5/28/20 9:59 a.m.

if mcmastercarr doesn't have the hardware I need - this place is good for odd end metrics:

 

https://www.belmetric.com/?zenid=daes93kkpd1c0sj4pehn01tdm1

 

even their logo is close to MMC. 

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
5/31/20 6:29 a.m.

In reply to fidelity101 (Forum Supporter) :

There is an Ace Hardware across from where I work that has the hugest selection of nuts and bolts and screws as I have ever seen.  If they don't got it, it don't exist.

 

Updates...  So I had it my head to want to try to properly characterize my injectors, which are modified Bosch units that were claimed to be 1200cc/min. I never could get them to idle or cruise well, and it would run lean or rich with air temperature changes, a sign that the deadtime was way off.  After too much thought, I decided to switch the firmware (MS2/Extra) to incorporate AFR table, then change the AFR table to see if the calculations were undershooting or overshooting. 

BUT, and this is where the "too much thought" comes in, this would only be accurate if I had the SIZE correct in the calculations, so I would have to run the AFR switching at different pulsewidths too, so I could not only get the "zero point" correct but also the "gain".  This would be remarkably difficult to do in a stationary environment as it seems like at any kind of idle or free rev the injector pulsewidth stays the same, no matter the RPM.   But then, I realized that I could run the injectors wide open in test mode into a large (60ml) graduated syringe that I had, to measure flow directly.  Perfect.

 

So, I spent too much time rigging up a way to hold the injectors in the rail (where DID my giant bag of zip ties go?) and did some math.  I want to know cc per minute.  A cubic centimeter is a milliliter, conveniently enough.  If I ran the injectors for 26 milliseconds on a 25 millisecond interval, that is 100% duty cycle guaranteed.  4 pulses would be a tenth of a second, 40 pulses would be one second, 120 pulses would be three seconds.  Multiply the result by 20 and the result would be cc per minute.  I set up the injector rig in my right hand, set up the laptop on the wiper cowl so I could click buttons with my left hand, and tested an injector.

To my surprise, the injector had a beautiful spray pattern instead of just squirting like a hose like a lot of huge injectors, so it foamed up in the syringe and half of it ran out all over my hand in the engine bay.  Oops.

Okay, let's cut it down to 60 pulses and we will multiply by 40.  I was not thrilled with this as the less fuel is measured, the more any measurement error would be multiplied, and I was not exactly using a tightly calibrated device.  But as it turns out it was good enough.

First injector, 24ml, repeated twice.  24ml times 40 is 960cc/min, significantly less than advertised even taking into account that I am running at 2.5 bar pressure and not 3 bar like injectors are rated at.  (Because that was how Mazda rolled in 1982 when they started injecting rotaries)

 

Second injector, 22 ml, repeated three times, and switched back to the first injector to verify that I was measuring THAT one correctly.  22ml times 40 is 880cc/min.  That's roughly 8% less flow.  THAT is why I could never get it to idle right, or had to make it run what seemed like amazingly rich under cruise to keep it from stuttering and lean-missing.  Because of the long primary exhaust system, I could measure only one rotor, and I probably had the wideband in the tube fed by the larger of the two injectors.

 

Well, berk.

 

Time for another hunt.... and found my good old GSL-SE injectors from way back when I started playing with injection in 2008 and picked up a GSL-SE engine complete for like $200.  (Because it was 2008 and they were worthless)  Cleaned them up as best as I could, decided to try to characterize them too because turd polishing knows no bounds.  Man these old style injectors are huge compared to modern pencil injectors.  Got them in the rig (where ARE my zip ties?), holding the rig in my right hand, turn on the fuel pump with the mouse with my left hand, and it blew one of the injectors out and started hosing pretty blue gasoline everywere.   And I am very right handed and can't mouse with my left.

 

Fine, whatever, I won't characterize them.  Open the garage door, set up a fan to blow as many fumes out as possible, go outside to clean up with GoJo and bottled water...

Patrick (Forum Supporter)
Patrick (Forum Supporter) GRM+ Memberand MegaDork
5/31/20 8:08 a.m.
Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
5/31/20 8:31 a.m.
Patrick (Forum Supporter) said:

Step 1 buy this

 

https://grassrootsmotorsports.com/forum/cars-sale/82-rx7-3500/173131/page1/

 

step 2

 

step 3 profit!

 

That looks a lot like my '81.  Which also has coilovers, a Dell'Orto on a 13B, 13x8 wheels (and 13x7 and a couple sets of stock ones)

 

You know what else my '81 has?  A full interior, and air conditioning, which were the two main reasons I agreed to get it.  (The previous owner took me aside when we were both at Frostburg in 2016 and basically told me that I was going to get a car)

 

Speaking of which, I MAY have found a good pair of 12A rotors, and thus will be able to assemble a good stockport 12A for it.  Because I want a stock 12A.  Stock drives nice, stock works like stock. 

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
5/31/20 8:08 p.m.

Told you it was close to the water pump.  The bottom of the TPS is about a ziptie's thickness away from the water neck flange.  I am very happy that I don't have to do it twice!

 

 

Spent a lot of time trying to get it to start.  The tune is "all jacked up".  Even with req_fuel at 5ms instead of a calculated value of 6.5, I had to pull the VE table down by HALF to get to run only eye-wateringly rich.  I have a lot of work ahead of me.

 

The AEM wakes up and responds much faster than the Innovate did, but I think I connected the MS to a data line instead of a voltage output.  Need to address that.  Also need to get some spade terminals so I can connect the TPS.

Bent-Valve (FS)
Bent-Valve (FS) HalfDork
5/31/20 10:06 p.m.

I have all your zip ties!

laugh

Seriously,  I needed the laugh I got while reading your injector misadventures. 

I would spray myself with gasoline too. Most likely worse than you did.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/1/20 7:41 p.m.

 

Invented a new tune while on lunch at work based off of some extrapolation from yesterday, then went to the 'cave, wired in the TPS, and it fired and ran smoothly!

Did some road tuning.  Rather a lot of road tuning.  There was a spectacular collision on Route 10 that better make the paper, looks like a dump truck tried to do a rail slide on a bridge's barrier and got stuck halfway.

 

I do not like autotune, but it has its uses.  I am less than thrilled with how it ALMOST gets you close and then sort of gives up.  Going to do it old school tomorrow, do a datalog on my way to work and look at histograms.

 

idles at 800 now, sometimes.  Idles at 1100 sometimes.  Not sure what  is up with that.

Clutch does not disengage fully, hydraulics feel good.  Transmission has anti-synchros in 1st, 2nd, and 3rd.  It has had a very hard life. There is a transmission and clutch and throwout fork in the back of the car.

Drivetrain vibrations are much improved.  Not perfect, but I have more important things to concern with.

Like, for instance, the floor, which makes all sorts of scary popping and creaking noises.

And the 5" tach stopped working.  I think it finally simply died.  I do not do in car video anymore so it is not really necessary anyway.

Patrick (Forum Supporter)
Patrick (Forum Supporter) GRM+ Memberand MegaDork
6/1/20 8:10 p.m.

Good thing I brought you like 5 transmissions last year

Stampie (FS)
Stampie (FS) GRM+ Memberand UltimaDork
6/1/20 9:16 p.m.

In reply to Patrick (Forum Supporter) :

I still remember sleeping under someone's trailer at the Challenge and hearing a near by conversation half conscious "Know anyone that needs a RX7 transmission?"  I woke from my nap and walked them over to you cause I knew if Pete needed one then a few more wouldn't hurt.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/1/20 9:29 p.m.

In reply to Stampie (FS) :

And I am extremely grateful!  I have been known to process them at a phenomenal rate.

 

To be honest, while my transmission hoard is comforting, I also find it limiting, because I have two RX-7s and I want to Duratec swap one of them and Chevy 3400 the other, and both options preclude the use of an RX-7 transmission

wvumtnbkr
wvumtnbkr GRM+ Memberand UberDork
6/2/20 2:46 p.m.
Pete. (l33t FS) said:

In reply to Stampie (FS) :

And I am extremely grateful!  I have been known to process them at a phenomenal rate.

 

To be honest, while my transmission hoard is comforting, I also find it limiting, because I have two RX-7s and I want to Duratec swap one of them and Chevy 3400 the other, and both options preclude the use of an RX-7 transmission

Let me know when you get to the 3400 swap.  I can help.  I have made all of the mistakes and ended up with an extremely reliable and basically maintenance free drivetrain.

 

This was in a 2nd gen,  but I have worked on 1st gens a bunch too.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/2/20 8:30 p.m.

TIL that I can still get the trans out in ten minutes, but I can't swing the new trans up and throw it in like I used to.  I had to heave it up on my shoulder-ish first, after taking a moment to accept my limitations and swallow my pride.

 

But between those two events, I also changed the clutch.  Flywheel looks fine.

New clutch disk's pucks are significantly smaller than the friction surface.  Hope this does not pose problems in the future.

 

the Plan was to use this with a stock pressure plate.  I only ever used an ACT pressure plate with this flywheel, because they came as a package deal.   The other TIL is that whatever brand pressure plate I had would not fit!  The area of the plate that rivets to the retention strap was too large to fit in the little cutout in the flywheel.

Stock-like on left, ACT on right.

 

so the ACT went back in, and I hope I did not screw myself by using the stock pivot ball/clutch fork instead of the reinforced ones that I had been using, because the ACT is stiff enough to cause problems.

 

The clutch verdict, by the way, is really smooth engagement, not switchlike at all.  Still has a short engagement window but it is linear.

sevenracer
sevenracer Reader
6/2/20 8:35 p.m.

ha... I was the guy looking to unload the transmissions!  Hopefully, they are of some use to you.

 

Pete, I probably have a set of usable 12A rotors that you can have.  the earlier style, not the later 12A rotors.  Not 100% sure of the condition of the bearings.  Let me know if you need them and details you need to confirm condition.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/2/20 9:15 p.m.

In reply to sevenracer :

Thank you, I definitely could use them!  I have counterweights for both rotor weights so either way is fine.  Technically I NEED to have the early heavy rotors since the car is an '81.  (I acquired something like seven 12A cores and they were ALL junk, even the one that allegedly had only 30k on it)

I forget the exact numbers, but the key factor is the width of the apex seal slots.  Basically if they are the same width at the top and the bottom they are good to go.  They usually are hammered apart because 12As run forever and those 3mm seals are heavy and beat the snot out of the rotors over the years.  Goopy makes/made .003" oversize apex seals, what they really need to do is make .006 and .010" oversize.  The engine that originally came with the black car was so worn out that you could fit an apex seal spring next to the apex seal in the slot.  So the engine was just all beat to hell.  Chrome worn through to the steel liner in places. 

sevenracer
sevenracer Reader
6/2/20 9:30 p.m.

Ok, I'll take a look at what I have tomorrow or Thu and get back to you.

 

Also - I have a vague recollection of posting this some time ago.  But I knew a guy who built a bunch of 12A race motors, who told me you could hammer the rotors to close up the gap in the apex seal slots.  Could tap a small wedge in too if you needed to open them up.

 

All the core motors I have gotten seem to have ok rotors and really worn out rotor housings.

Stampie (FS)
Stampie (FS) GRM+ Memberand UltimaDork
6/2/20 9:33 p.m.

In reply to sevenracer :

Sorry I didn't remember it was you but I was half asleep.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/2/20 9:42 p.m.

In reply to sevenracer :

Closing the gap up always concerned me because you'd have to do it extremely evenly, and do it in equal amounts from both sides so the corner seal doesn't bind on the apex seal.  And you'd have to make sure you were not closing up the corner seal's bore.

 

If there's a guy who can do it reliably, I have many sets of rotors he can use as cores.  OTOH I acquired a 12A built by a "race engine builder" (was told his name, I forgot it almost instantly.  It was NOT Dale Drummond) that turned out to be an amazing pile of junk.  Worn out rotors, used apex seals, flaky chrome rotor housings, gouged side housings... and the engine was built in 1996, when 12A parts were still available!

EvanB (Forum Supporter)
EvanB (Forum Supporter) GRM+ Memberand MegaDork
6/3/20 6:20 a.m.
sevenracer said:

ha... I was the guy looking to unload the transmissions!  Hopefully, they are of some use to you.

 

If those were the ones that I gave you then they went from grpb in Chicago, to Columbus, to Charlotte, down to the Challenge then up to Cleveland to Pete, who has a bunch of other rotary stuff from grpb.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/3/20 8:02 a.m.
EvanB (Forum Supporter) said:
sevenracer said:

ha... I was the guy looking to unload the transmissions!  Hopefully, they are of some use to you.

 

If those were the ones that I gave you then they went from grpb in Chicago, to Columbus, to Charlotte, down to the Challenge then up to Cleveland to Pete, who has a bunch of other rotary stuff from grpb.

His engine and carburetor are in the '81.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/3/20 5:24 p.m.

The shop welder stops feeding below 7 (out of 10) which reinforces my opinion about Snap-On anything.  Plus Mazda steel is difficult to weld for whatever reason.

 

 

Will probably hold off on the other side for now.  Remembering how much I hate doing any kind of bodywork.  Plus there was all that excitement where I set the inside of the car on fire.

 

Appliance epoxy makes anything look less bad!

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/3/20 8:07 p.m.

Floor is still popping.  At least, the odd popping sound is still there.

Realized today that the AEM gauge is reading differently than Megasquirt!  It's off by .3 of a ratio.  The 30-300 was not an option in the drop down menu so I chose one that seemed to agree.  I must have chosen unwisely.  The wideband is grounded at the same bolt as the Megasquirt so that is probably not it.

 

For now, I'll just adjust my air/fuel targets by .3 and go from there.  The tune is 98% there, before going for economy, so I'm tempted to just say berk it.

 

wvumtnbkr
wvumtnbkr GRM+ Memberand UberDork
6/3/20 9:44 p.m.

Mazda sheet metal is hard as berkeley to weld for some reason.

 

Says the guy burning in New cage bars and fixing some rust spots on a 2nd gen.

 

It's like this spot okay, that spot made out of tinfoil.

Pete. (l33t FS)
Pete. (l33t FS) GRM+ Memberand MegaDork
6/4/20 9:49 p.m.

 

Was registering for OVR PE1 and PE2. and as a joke changed my "sponsor" to E. N. D. because the official color of the car is "Evil Black" and this car just will not die.

 

I could still recite from memory the part that is totally 100% not taken from Master of Puppets

fidelity101 (Forum Supporter)
fidelity101 (Forum Supporter) UltraDork
6/5/20 8:28 a.m.
wvumtnbkr said:

Mazda sheet metal is hard as berkeley to weld for some reason.

 

Says the guy burning in New cage bars and fixing some rust spots on a 2nd gen.

 

It's like this spot okay, that spot made out of tinfoil.

oh yeah its terrible and difficult to weld too, i've been there... often

1 ... 30 31 32 33 34 ... 54

You'll need to log in to post.

Our Preferred Partners
3nu3lFw7HJRvbtn6SkPnoX2HURBQhX97CMR8UaVV1Uxat7ZOMmgyWLZKPv2o7gut