Random Thoughts... [Automotive Edition]

So late last month, I had a total derpstorm.

Was working on replacing the engine in the Legacy and got fed up with stupid Subaru EJ engines and their stupid head gaskets, timing bullshit, comparative expensiveness and general leakiness and explodium fragility.

On the other hand, I am absolutely IN LOVE with the FB engine family and it's FA cousins. I have two, and know people with plenty more. As long as you put oil in them, they just run. Period. There's probably a million miles of driving between those cars, and not one engine problem to speak of. Meanwhile, in EJ land, MTBCatastrophicF is something like 150k miles.

So I bought a rekt 2015 Impreza with some pocket money ($2k), and a shedload of parts ('bout $600) to fix said Impreza's engine (which was a bit more heavily damaged in the collision than I was expecting - notably the crank pulley took a hit, sheared the woodruff key and exploded, sending shrapnel into the timing sump, and also boogered up the threads on the crank.

The plan is to take the FB20 and jam it into the hole the EJ22 came out of and do a complete wiring transplant, retaining literally everything including the emissions gear.

There are two exceptions:
1) I'll be building an electronic box to masquerade as the Transmission Computer because I'll be retaining my manual transmission and I bought a CVT Impreza (oops). $50 would buy me a manual ECU, but this is more interesting
2) I'll be building another electromechanical box to masquerade as the gauge cluster, spinning motors and translating CANbus signals to analog lights and whizzy speedometer cable shenanigans (maybe, I might just keep the mechanical speed sensor in the trans and ignore the wheel speed outputs) to retain the factory gauge cluster (the remaining data will be passed to the stereo for display, probably).

I figure this will, at the end of the day, cost LESS than a professionally rebuilt EJ of ANY description.

I'm half tempted to follow this project up with an '07 Impreza (the best looking car Subaru has ever built), retaining the CVT, and developing this into a repeatable process or kit, and then shipping it to a friend of mine in California to see if we can ram it through the CARB Engine Change process, and then marketing this shit.
 
Then explain to me how was I able to get through deep snow with 4LO but not 4HI?

Low range is a torque multiplier but it also significantly limits the speed the tires can turn, this gives more traction because the engine isn’t trying to turn the tires at 5mph at idle, instead they are trying to go 1mph or less which causes less pull on the grip between the tires and the ground. This keeps the tires from spinning and is why you were able to crawl through deeper snow.

I think that makes sense.
 
There are two exceptions:
1) I'll be building an electronic box to masquerade as the Transmission Computer because I'll be retaining my manual transmission and I bought a CVT Impreza (oops). $50 would buy me a manual ECU, but this is more interesting
2) I'll be building another electromechanical box to masquerade as the gauge cluster, spinning motors and translating CANbus signals to analog lights and whizzy speedometer cable shenanigans (maybe, I might just keep the mechanical speed sensor in the trans and ignore the wheel speed outputs) to retain the factory gauge cluster (the remaining data will be passed to the stereo for display, probably).

I think you will find that 1 is far more difficult than it is worth to get into a working state. 2 isn't going to be trivial either - especially when you can get plug-in displays that can replace the factory cluster, see the Cresta done by Mighty Car Mods for an example of a modern one.
 
Low range is a torque multiplier but it also significantly limits the speed the tires can turn, this gives more traction because the engine isn’t trying to turn the tires at 5mph at idle, instead they are trying to go 1mph or less which causes less pull on the grip between the tires and the ground. This keeps the tires from spinning and is why you were able to crawl through deeper snow.

I think that makes sense.
That's more or less what I tried to say above but BCS didn't wanna believe me :p
 
I think you will find that 1 is far more difficult than it is worth to get into a working state. 2 isn't going to be trivial either - especially when you can get plug-in displays that can replace the factory cluster, see the Cresta done by Mighty Car Mods for an example of a modern one.
Eh, I've got a couple months to dick with issue #1 and gather data for #2.

I've got the electronics and software chops to pull it off, it's just a matter of how well the CANBUS chatter lends itself to reverse engineering.

I'm not planning on emulating a full under-all-conditions drive cycle for #1, I'm just gonna do the bare minimum necessary to escape the 4500RPM neutral rev limiter, which SHOULD be something like:
1) Engine starts.
2) User shifts out of park and commands paddle shift mode. TCM complies (probably tie this to the parking brake)
3) User commands first gear (again, we'll probably just assume parking brake off, first gear!). TCM complies.
<fake transmission outputs for speeds following engine RPM locked to the "first gear" ratio, if the ECU cares for some reason>
4) User shifts into park (again, probably hijack the parking brake).

Plus any necessary "Yes everythings fine" chatter.


And if I can't do that, I can build a box that pretends to be the transmission. The TCM communicates with the actual transmission on 3 binary inputs (up, down, torque converter lockup) and 3 PWM outputs (torque converter RPM, intermediate RPM, output RPM). Torque converter lockup can (probably) be ignored and torque converter RPM becomes engine RPM, and everything else is just a smoothed table lookup of "gear" versus engine RPM. If TC lock can't be ignored, it's just a matter of fudging the variables downward by a factor the computer finds acceptable.

And if I can't do that, I'll give LKQ $50.

As to the gauge cluster, I'm not particularly worried about that, but that SHOULD be substantially easier because the cluster doesn't actually emit any data in normal operations, just sink it, and if the damned VSS. I hate the look of every single aftermarket dash ever built by mankind.

And if I can't do it straight off the proprietary canbus traffic, the ELM serial protocol is stupid easy, and all the data I actually need for my idiot lights and gauges is exposed to readily accessible standards-compliant PIDs.
 
TC RPM isn't engine RPM, it's the RPM of the impeller inside the TC, IIRC. TC lockup cannot be ignored for most modern ECMs for a number of reasons (fueling map, throttle curve).

The aftermarket dashes you can get now can be just an LCD panel that you can configure however you want.
 
Low range is a torque multiplier but it also significantly limits the speed the tires can turn, this gives more traction because the engine isn’t trying to turn the tires at 5mph at idle, instead they are trying to go 1mph or less which causes less pull on the grip between the tires and the ground. This keeps the tires from spinning and is why you were able to crawl through deeper snow.

I think that makes sense.

Low range increases your crawl ratio, but cannot ever increase available traction. If that were true, we could just install a 100,000:1 crawl ratio and have infinite traction all the time, drive up the side of buildings.

If you are losing traction at idle in 4HI, you are high-centered on something! Slow tire speed is important when crawling obstacles and you want to do so in a careful fashion with better throttle control, but in snow on flat ground it's of no benefit whatsoever. If one makes it through a snowdrift in 4LO that they couldn't do in 4HI, it's because they picked a better line or packed the snow down on their previous attempts - the state of their transfer case did nothing to change their tire traction, as it physically cannot.
 
I have been asked an interesting question that I had an opinion on, but I am not really sure about. I will reserve my opinion for the moment.

What is the BEST way to store tires and wheels? On the tread as long as they are rotated occasionally, or on the side? And why?

I have also seen racks that store them on the side, but have a spacer to keep them from resting on the side of the tire itself.
 
About 4 months.
 
Low range increases your crawl ratio, but cannot ever increase available traction. If that were true, we could just install a 100,000:1 crawl ratio and have infinite traction all the time, drive up the side of buildings.

If you are losing traction at idle in 4HI, you are high-centered on something! Slow tire speed is important when crawling obstacles and you want to do so in a careful fashion with better throttle control, but in snow on flat ground it's of no benefit whatsoever. If one makes it through a snowdrift in 4LO that they couldn't do in 4HI, it's because they picked a better line or packed the snow down on their previous attempts - the state of their transfer case did nothing to change their tire traction, as it physically cannot.
What CJ and I are trying to say is that it's not the amount of traction you have but ability to rotate tires more slowly and allow them to maintain what traction they do have. Of course there is a limit, if you are sitting on ice with near as makes no difference 0 traction you won't go anywhere.
P.S. In the specific case of the Xterra you also can't lock rear diff in 4HI.

@GRtak on the side is likely safer as the load is spread over a larger contact patch to decrease possibility of a flat spot. Additionally tires are already somewhat flat on the side so it being slightly flatter won't impact drivability.
OTOH I think that if you inflated them to max and have no other load 4 months wouldn't be enough to cause any kind of flat spots considering the low amount of pressure that a wheel + tire has vs wheel + tire + whole car.
 
What CJ and I are trying to say is that it's not the amount of traction you have but ability to rotate tires more slowly and allow them to maintain what traction they do have. Of course there is a limit, if you are sitting on ice with near as makes no difference 0 traction you won't go anywhere.
P.S. In the specific case of the Xterra you also can't lock rear diff in 4HI.

And again, that's great when you're crawling rock in the summer. Nobody crawls a snowbank because that's not how you make progress in deep snow - it's very similar to mud in that you want to maintain wheelspeed and momentum, unless you're running Iceland-spec flotation tires. I wheel snow and ice on the regular - if you try to idle your way through a snowfield you just....stop.
 
Just went to the new car expo.

Toyota trucks and SUV have impressively shitty visibility. The 4Runner TRD Pro has a front blind spot that I estimate to be 20 feet long - and that's for a 6'4" tall guy sitting with his head brushing the headliner. The Tundra is ridiculous, it has TWO fake hood scoops stacked on top of each other. I could not see the rins of the minivan parked 30 feet away across the bow if that battle ship. The hood was probably 10 inches taller than it needed to be and almost half a foot thick.

The Tacoma was terrible and the interior was garbage. Ford and GM both had it beat in design and materials. Yes, that GM.
 
8" of snow overnight, turned to freezing rain by morning. Pretty much the only vehicles on the road are plow trucks, although the roads are covered in packed snow. LSD and snow tires making the commute this morning easy and painless. "As you would expect, I've brought the right car."

fEA2svC.jpg
 
Settle a debate. What would you rather drive through a major snow storm: FWD manual with an LSD and snow tires (Acura TL) or a true 4x4 with all-seasons (Grand Cherokee)?
FWD with snows. Tires are more important than drive type.
 
Top