Roman's 4GR V6 Toyota Carina

[Roman]

I've pulled the existing intake manifolds off the motor so I can measure everything up easily for seeing how much space there is etc.
One thing that is very critical now, that wasnt before. Is that I need to know exactly where each bank is relative to each other.
So when the pipes are criss crossing between each other in my model, that's represented properly. As it only needs to be a little bit wrong, then nothing will fit.
I have printed a few iterations of a piece like this until everything exaaaaactly lines up. 



One quirk/feature with Fusion360, and I think solidworks operates this way too. 
Is that everything works on a timeline. Your model isnt just "how it is". It starts from the first thing you ever drew, then deleted, then drew again, and so on. To end up with the final result that you see. 
Which is really cool because you can roll back the timeline to make amendments or whatever. 
What's not so cool though, is that if you have been working on the same model file for say.... 3 years with zillions of things in it. 
Every time you make a change, it needs to reprocess all of these tasks. So it starts getting really sluggy. 
So I've started a new model and am working from scratch. 
If I turn everything back on in the old one, there's a lot going on! So can understand why it's getting a bit slow.
I think out of all of this stuff, the only bits I'll still be using are the VVT cover plates and the collectors for the exhaust.


 

I'm making some reasonable progress on my new design for the manifolds.
It looks like metal printing the runners will be prohibitively expensive (about $1k for just runners) 
So will need to try something else that's more cost effective.
I am thinking that I will get some alloy flanges cut. 
Then will get the runners printed from nylon, then wrap in carbon sleeve. 
Then alloy flanges at both ends so there's no bolts tapping into plastic and so on.

The runners need to go slightly oval where they cross over between each other.
But I've got the same cross sectional area all the way through. Will probably end up looking something like this




It's been a bit of a mission trying to find a good injector position that isnt going to interfere with the throttle linkages and/or stick up too high. 
Trying to have 2x injectors per cylinder might be too hard to fit.
I might need to put it somewhere like this perhaps.
Not ideal, but arguably still a lot better than it was last time.
Ideally I want to have them at least firing onto the hot aluminium part rather than the runners.

I've got a single runner ready to go on the printer, so hopefully will be able to test fit some time tomorrow. 





One other thing. 
Although I was a bit dissapointed to see the motor make ~220hp. 
I found an old dyno graph from my beams motor that I had in this car.
But from back when it was in my SW20 with standard ECU. 
So not 100% representative of the final result with beams, but, pretty decent difference!
If the car ends up only as good as it currently is, it's going to be wildly fun to drive. 
But my nerd brain wants to see the graph keep pointing upwards, so, keep working on things and try find some improvements here or there.

My Cartest program which has proven to be fairly good in the past.
Says in its current state should run a low 13 when everythings dialled in a bit. 
So long as it beats the Echo eventually 😐

The big dip at the end of the graph here was from the roller speed maxxing out on a 5th gear run, doesnt actually dip power like that. But you get the idea.

[Roman]

The general design is mostly settled. I have just had a few test prints optimizing spacing and fine tune a few things.
The major change that I've made is making the whole assembly a lot taller than before.
I printed a few runners of my original scheme and found that I had HEAPS of room to the bonnet. 
Even with these taller runners, there's probably still 40-50mm to the bonnet at worst point. Which is great.

I'm still going to go with the idea of having the runners nylon printed, then wrapping them. Then alloy flanges glued on at each end. 
However I've also add a plate that bolts across all 3 throttles to hold them all square and true. So there's no twisting or misalignment causing annoying issues with throttle imbalance.
It'll need some bracing to support the weight of the throttles but will figure that out once it's all together and actual throttles are here.

 

The hard part at the moment is still figuring out how to add the injectors. 
So for now I have just spammed it with every possible location where it looks like they might fit. 
Ideally I'd have the end of the injector poking into metal somehow rather than the nylon printed part. 
However it's not looking very practical to do that.
My throttles should be here in the next day or two, but realistically I need a fully "working" plastic printed setup before I can get any of the parts properly made.

Option A and B are lower than the port divider so would need to be twin injector per cyl.

I am thinking that option D with perhaps a slight tweak of position. Is probably best overall. 
But I will get it all printed and see how it's going to work. (or not)

 

[Roman]

It's all go! 
I decided to sling the injectors underneath, mainly because I found a spot that will work with my existing printed fuel rails and injectors.
So it'll look a bit like this:

Then I've had a few more iterations of prints and test fits etc, and everything's good. 

I've got ants in my pants to keep making progress, so today I have: 

-Ordered the intake runners to be PA12 nylon printed out of the powder stuff.
-Ordered 2x TPSs that fit. They were cheaper to buy brand new on aliexpress, than second hand from pickapart...
-Ordered the carbon sleeve for wrapping the runners.
-Modified my baseplate and flange drawings, so that they are ready to be cut. 

Exciting! 

That's well and truly blown my car budget for the next month though.
So probably not much else happening for a while. 

[Roman]

Carbon sleeve turned up, so had a test hoon on one of the PLA runners.



This is way thicker material and a lot courser weave than I was expecting. So it was quite hard to use compared to other times I've done similar.
I am thinking maybe instead of using the sleeve. Perhaps I'll just use regular carbon fabric and try keep all of the seams on the underside. 
As I've got some very thin stuff with a much finer weave. 
I guess since I've got a bunch of PLA runners already here. I may as well try it on one of those before committing to applying on the proper ones.

Also, proper runners turned up!
MJF printed Nylon PA12.
And they look great. They have a nice texture that should be good for adhesion.
Yet again I'm blown away by the detail from the MFJ printing process.
It still feels incredible to be able to draw something on a computer, then have some real life parts pop out.
Whether printing myself, or from someone else. It's incredible.



In preparation for coating I've ahhh... waxed their bung holes. 
But need to make some decisions about how to wrap them. 
Guess a PLA trial of the normal fabric is the next step.
If using the sleeve again, I've got some better ideas for how to stop it from fraying at the ends. 
It was only 5 weeks ago that I was on the dyno, so I'm pretty stoked to have had a reasonably short turn around time on designing a whole new setup! 
But it's still a while off yet. I need to figure out how the ethrottle stuff is going to work.
The TPS sensors I bought technically do bolt on - but dont allow the throttle to open its full range which is annoying.
So I'll try figure out a little adapter to rotate them around a bit.
The only "bolt on" TPS units look to be some very expensive ones from Jenvey or another 3rd party company in Europe.
So I'll make something work that's easier to source.


EDIT:

[Roman]

I got two of the proper runners wrapped. 
I didnt have enough sleeve to do all 6, and, it wasnt working very well anyway.
So went back to regular CF fabric.
This worked alright. 
I ended up with a lot of creases from the bagging process to compress it together.
Not pretty but they are super strong now. Good enough! 



Then the absolute legend Stu has helped me out yet again.
All of the flanges and base plates are now cut. 
So I will pick them up early next week.
I cant wait to get this all together! 



From here, I will get all of the lower flanges glued on to the runners first. Then bolt them all to the base plate. 
I am printing a jig that aligns 3 of the top flanges to the correct spacing and orientation. So the 3 top flanges will slide/glue in all at once. 
I really want to avoid any possibility of the throttles being misaligned relative to each other, so hopefully that works well enough.

Another long term problem that I've had, is how to run dual ethrottle.
A friend of mine has given me a long term loan of a G4+ Thunder ECU. 
Which will be great!
There's a small amount of depinning/repinning required, but it's largely compatible with the G4+ Xtreme pinout of the first two plugs. 
He's not a member here, but, thank you very much Mike for helping me out.



Apart from dual ethrottle, it has some extra features over my current ECU which will be fun. 
However currently the focus is to just get the motor running again, and get back to the dyno with new intake and test some runner lengths.
Cant wait to see how it goes.

[Roman]

Getting there! 
I spent about 5 billion hours tapping all of the threads for the studs in the base plate and top flanges.
As there's 60 holes to tap.
You'd think I'd be good at it by then, but, I still managed to do a bunch of them a bit crooked.

Once all of the base plates were glued on, I fitted them up and made sure fuel rail situation will be okay.
Should work alright, but I will need to cut the existing brackets off and figure out something new I think.
The injector plugs are a bit tight, I might solder/pot the plugs and put fly leads on them.



Printing the jig to glue in the top flanges worked well:

 

While doing an engine bay test fit, I discovered some excellent news.
I can have all of the runners bolted on, and remove/install the entire intake as one piece.
The throttles need to go on later, but, could be worse!
This whole assembly looks huge on the motor, its gonna look cool when it's finished.



I've almost finished wrapping the runners. I gave up on the bagging process and have just been doing a wet layup. 
To make life eaiser for myself, just doing the top side and bottom side in two separate stages. 
Will need some sanding and some clear coat but should look good enough left as carbon finish.




Since the rest of it is nearly done, back to figuring out trumpets. 
I had a good measure up of the existing lengths of everything with my previous setup, then new. 

Since my runners are now a lot longer, my trumpets only need to be a lot shorter. 
It's looking like shortest length it will like will be a ~30mm effective length, longest 110mm give or take. 
So I'll try get some adjustable lengths within that range before going to the dyno again.

For starters I've just gone for a 35mm long non tapered section, then a bellmouth on it. So effective length is probably 40mm or something. 




Next things to sort: 
-Cut the 15 degree angle into the base plates on each port
-Print some little brackets which rotate the orientation of the TPS on each bank.
-Figure out how ethrottle will work now.
-Make a new tune file for the Thunder ECU, and come up with a plan for the wiring changes needed.
-Might need to buy some new injector O rings.
-Finish printing trumpets. (3 done, 3 to go)
-Fit some hose barbs into the base plate for PCV and brake booster lines, hook them up.
-Remake the AN lines to fuel rails as the spacing is narrower now.

I doubt it'll be running by this weekend, but maybe next? 

[Roman]

Quote

Next things to sort: 
-Cut the 15 degree angle into the base plates on each port
-Print some little brackets which rotate the orientation of the TPS on each bank.
-Figure out how ethrottle will work now.
-Make a new tune file for the Thunder ECU, and come up with a plan for the wiring changes needed.
-Might need to buy some new injector O rings.
-Finish printing trumpets. (3 done, 3 to go)
-Fit some hose barbs into the base plate for PCV and brake booster lines, hook them up.
-Remake the AN lines to fuel rails as the spacing is narrower now.

I doubt it'll be running by this weekend, but maybe next? 

Well, not quite running by the weekend.
I've ticked a lot off the list, but still need to get fuel stuff sorted properly.
The main thing that I've been scratching my head about has been fuel rail mounting.
The bendy rails looked like they would work okay, but the mounting tabs were in a bad spot.
I tried the straight rails, which is an easier shape to work with.
However the rails hit the runners before they could go in far enough.
Then I remembered you can buy little extension caps for injectors.
So I bought some of these and now it's good with the straight rails.
But then I've still needed to figure out how to hold them in the right place. 
It looks like hanging some brackets from the front of the throttles will work best.
I was originally planning to get these cut from flat plate, but looking at the price for printing and not too bad. 
So I've scope creeped some complexity into the shape, so my car looks more like a tower crane or something. 
(Actually can barely see any of this from the top when engine is together)

After printing 6x trumpets and fitting them up, I also decided that was a total prick to try and get the screws in and do them up when you have to try and do it from the sides. 
So I've added some holes through the trumpets so you can get a long allen key through. 
I've chamfered the holes so hopefully they dont whistle.

The little round dingus on the right hand side of the picture will be what holds the cable for ethrottle on that bank. I will need a few more test prints to figure out the right position and angle. 
Then I'll send all of the brackets off to get printed. 

[Roman]

I got it all running again with some plastic printed bits to hold the fuel rails in. 
Obviously not a good solution for when the engine fully heats up. But lets me make some progress until the other bits turn up. 
I wont drive the car but I can at least start the motor up etc.

With this new intake, a lot of the wiring is now much more hidden which is nice. So it feels like a worthwhile idea to try and get the visible wiring on top of each bank running a bit nicer. 
So I'll do that at some point. 



It idles nice and low, can get it down to 700-800rpm at the moment, and throttles arent adjust yet. 

Speaking of adjusting throttles...
Previously I made a small tube with a MAF fitted in it, to balance them up. Seemed to work fairly well. 
Now that I have a Thunder ECU to use, I have an absolutely enormous amount of IO at my disposal.
What to do, what to do... 



Since the Toyota style MAF used in the vitz etc is now quite an old part, they're at end of stock prices... $20 each on aliexpress.
I've already got one, so I've bought 5 more of them and the plugs to go with it. 
I will run a MAF on each cylinder and see what it shows. 
I'll need to validate that each MAF shows similar output. 
But will be interesting to see if it shows much of an imbalance.
I wont leave them in there long term, but it will be good for two things...
Balancing all of the throttles relative to each other. Which is a bunch of fiddly adjustment of linkages. 
Then also balancing the cam offset left to right for VVT. 

When I went to the dyno, one bank ran leaner than the other.
When correcting the fuel values, there is/was a 6% difference in VE bank to bank. Which is huge. 
If they could both match to the highest value, I guess that's potentially a 3% power gain just from adjusting a number in the software.

The cam timing doesnt match properly bank to bank, if you just base your zero point from "home position" with no cam advance.
One bank is effectively slightly advanced relative to the other, due to tolerance stack I guess.
It will be easy to balance up the bank to bank offset in software, with the MAF results. 
Since I need the airflow to be fairly smooth, my MAF test pipes need to be longer than my actual runners will likely be. 
But, like I said, it's only there to help setup the mechanical parts of the engine that are irrelevant to runner length. 
I will still take it back to the dyno as I'd like to see how power compares.

I've got 6x of these to print and they take 10 hours each. But no rush as I wont have the sensors in 60 hours time anyway. ha.




I spent a bunch of time migrating my tune to the Thunder, but havent plugged it in yet. 
The only pins that I need to move from existing are the canbus wires. Easy! 
 

[Roman]

Welcome to Daves Instrumentation Adventures

So I thought I'd better test the idea on a single runner before after I've already purchased 5 more MAFs

the black stuff is using a 3d print pen to join the 2 parts. My only red roll of filament was still on the printer printing the other runners... 




Restrictive? Probably not.



Results are decent, and it is insanely sensitive. The wobbly bits shown here are reversion from having god awful cam timing. Which smooths out once rpm gets up a bit. 
But it's looking promising as a good method for balancing throttles and the bank to bank VVT stuff. (Apart from that my bonnet wont close now) 



The horrible reversion stuff mostly happens just off idle, when rpm is 1000+ and it's steady idle it gets a nice smooth signal. So should be okay to use it to balance the throttles.

[Roman]

 

All of my MAFs turned up, so I got 3 of them on and gave it a rev up to see what happens.
It works fairly well for high rpm/load, but the pipe size is too large to show anything at low airflow. 
But low airflow is the area where I most need everything to balance.
The results were very jiggly looking, partially because my ehtrottle PID settings are still a bit whack.

So, need to get ethrottle better sorted.
Two things needed working on to get it good enough to be drivable.

Firstly, my TPS sensors needed to be on a different rotation than a direct bolt on, in order to show the full 0-100% throttle range. 
So drew some adapters like this:



So I printed some spacers from PLA which worked alright, but, wasnt really rigid enough. So the PID values had to be weird to stop it jittering a lot.
So once again, GC to the max Stu cut me some alloy ones.
He sent me a pic along with an unrelated graph that he must be half way through drawing on the paper...




My other problem is that I need to support the cables for the e-throttle motors a bit better so they dont flex when it pulls.
Easiest thing to do here was just build something into the front two trumpets. 
One side needs to be higher than the other, in order to clear the VVT solenoid. 
I've got these printed and fitted up, and now it's much better.



Then, I've just been being a shit head revving it up in the driveway on stands. 
To see how the fuel wants to change. Is the new intake is worthwhile? I hope so, because its been a ball buster to make. 
I only managed to get some results from about 8000 to 9000rpm, because I forgot the widebands reset when the engine stalls. 
But in that region the motor wants around 8% more fuel than before, which is fairly promising. 
I still need to figure out the VVT offset amount that is needed bank to bank. 
Once that is sorted, might see a 10-12% increase overall compared to before. 
Hard to say, and I'm not one to just wildly speculate 
It sounds a bit more intake-y now which is cool. 
 

I forgot that the rev limiter is set a bit higher than before, woops...
Well, oil pressure is still pretty good and nothing exploded yet at 9370 rpm




I've been struggling to get the car to behave nicely just off idle, but I think I've found some improvements. 
I've found that I can run the exhaust VVT down to around 800rpm without issues, and retarding the exhaust cam by 20 deg (which puts it on the "normal" timing home position) and it behaves considerably better.
I found my cruise control switch in the garage, so I've got this wired up again. 
I can hopefully start getting some low load areas dialled in nicely soon enough.

Keen to see how it compares on the dyno! 
I'll print some shorter trumpets as well, and then some spacers too. so I'll have a few combinations of runner length to check. 
Hopefully I'm roughly in the right sort of range currently.

[Roman]

I'm inching closer to getting things sorted. 
I spent a few nights working on unrelated wiring to get the lights and so on working. Just indicators and windscreen wipers left to sort then I can tidy it all up and tuck it away.

I've been doing some more car-on-jackstands based fuckery to try and dial in the fuel map. 
It's surprising how well you can get the ramp rate when you put it in 4th or 5th and just jam the brakes on. 
But just have to keep them spinning for a while after to let the poor discs cool down haha.

TL;DR: noises:
 

But anyway, after a few more runs, much to everyone elses annoyance probably. 
I've got some good values in the fuel table now, and it's mostly tracking on target apart from some bank to bank imbalance.

Its looking fairly promising for next dyno visit! 
Here are the old VE values from the old intake, new values, and the percentage difference.
Keep in mind that with NA motor improvements, you're usually scrambling for 1% or 2% at a time and looking for accumulative gains. 
So this is pretty amazing / shows how crap my old setup was. 




Now I appreciate that it's a stretch to expect horsepower to directly correlate with AFR changes. 
But, if wheel horsepower did improve by the same % as extra airflow now, here's where we would be looking:




I believe part of this is thanks to my new runner length giving the good bump right on the 7500-8000rpm region now. 
If I try some slightly shorter runners, I might be able to bump it over to the right a bit which would be cool! 
If it gets to over 250whp I will be pretty happy.

Here's a diagram of what my original runner length was, the two long lengths that we tried on the dyno (which were either too short or too long) and my best guess for current intake length. 

Another thing to keep in mind is that the motor might like having some more cam advance now, for higher up in the rpm range. 
So it might not be done yet! 

Exciting if any of this pans out to be true, but, we all know that dyno day can be a heart breaker. Haha






Something else that I screwed around with, was adding or subtracting 5 degrees of VVT advance from 1 bank only. 
It didnt seem very sensitive to this, not enough to make up for the AFR imbalance. 
So I'll need to do some more checks and figure out if there's any other reason why this might be happening. 

The bank that is running richer is the drivers side, which is funny because its the one with the better looking exhaust manifold on it.
Its not toooooo far off I guess ,but I'd rather try find a mechanical reason for the imbalance than just accept it and run a fuel correction on one side.
As there's possibly a few more farts to squeeze out of it yet.

[Roman]

I'd just like to get this out of the way 

FUCK YOU, V MOTOR PIECE OF SHIT, WHY ARE YOU LIKE THIS ,WHAT IS YOUR FUCKEN PROBLEM ,WHY ARE YOU SO DIFFICULT, FUCK YOU 

So anyway, 
Ethrottle. 
Easy right! 
Get an ECU that can run 2x motors, read 2x TPS then you're sorted.

Wrong. 

I drove the car up the road and it was insanely uneven left to right banks, was misfiring and popping and all sorts of bullshit.
So I've been thinking back over my setup a bit. 

If the TPS doesnt output an exactly linear signal (like 0.1 volts per 5 degrees or something) 
Then there are problems if you have two TPSs and need them to exactly match.
What makes the problem worse for me, is that I'm sweeping one "backwards" as it rotates counter clockwise. 

To get a feel for whether this is a genuine issue or not, I randomly stuffed some garage objects into the throttle on each side, then wrote down the output voltage. 
If everything is good, it should be two straight lines, just with a constant offset of 0.5 volts (or whatever)





So yeah its a bit screwy. 

So I spent a bit more time on it this morning. I figure the best option is to scale them both to 100% and 0% with some exactness, and then make a compensation table for one bank so it opens the throttle at different amounts until they match.
So started writing down some numbers, starting with something that opens them about half way. 
With the throttles opened the same physical amount, one TPS showed 40%, the other side 70%.
Even though they both show 0% and 100% correctly. Gah! Nowhere near good enough to be usable.

So my next option is likely gonna be buying the TPSs that are designed to be left and right sided, and presumably have a lot more windings on them. 
But it's another delay and another $320ish I have to pull out of my ass. Frustrating.

Also the cable setup works kinda alright, but funnily enough it seems like it worked better previously when I was using a single motor to pull both banks open. 
The problem is that the PID settings expect to be able to use the motors to pull the throttles open, but then also "push" the throttles shut. 
If I have the PID settings nice and snappy for opening, it has a bit of a fit when they're closing again, and they can only close at the rate that the return spring allows them to.

So my next idea. These throttles have an 8mm "D" shaft for the TPS or throttle pulleys to bolt to. 
I've found some DC motors that have a planetary gearbox on the front, so they are slower speed but higher torque. 
So I'll biff these onto the front of the throttles, and hopefully without any backlash etc in the system, it'll work better.





I am conscious of running out of time heading towards drag day, only about a month to go! 
If need be I'll just go back to a fully cable operated setup. 

Another thing is, the Link G4+ Thunder.
Man I would be absolutely furious if I bought one of these back in the day. 
They basically released this as a top spec fancy ECU then dropped support for it pretty much instantly. 
So some of the features particular to this model are still undercooked. 

Example, the dual ethrottle control.
The idle up table only works on one of the throttles not both. What the fuck is that good for?
Also one side can have up to 3 different tables controlling it, but the second bank only gets one. 

I'm very grateful that a friend has let me use it, but, man its annoying that some of the code is half arsed and they wont ever fix it.

[Roman]

Ahhh so the TPS issue. 
A bit more thinking on it, once the frustration faded. 
The TPS output was sort of banana shaped, and the 2nd TPS like an upside down banana. 
This reminded me that I had previously reversed the 0v and 5v signals on TPS2, so that it tracks in the direction of 0v when closed to 5v when open.
This is fairly commonly done, and isnt an issue if the output is linear or if you are running only 1x TPS.

However this is what made one of the banana shapes upside down and exaggerated the differences in the midrange. 
So I set it back, and fiddled the TPS angle a bit to match them as best I could. 
The outputs are still a bit shit, but, at least workable now:




So this results in a table for Ethrottle 2 like this




with some road testing I can see if the 2nd bank wants some small tweaks up or down to balance airflow. 
Will give it another run shortly once the laptop has charged up.

[Roman]

I fixed the fuel breather situation, and took it for a run up the road.

Interestingly the bank to bank balance that wants to match AFR side to side isnt equal to what mechanically sits the throttles open at the same point. 
So even if my TPS situation wasnt funky, it'd probably warrant having the side to side differences anyway. 
So I'm just trimming the fuel to suit ethrottle 1 results, then tweaking the shape of the ethrottle 2 curve till things match up. 
Annoyingly though if using the 2nd table for ethrottle. Then NONE of the extra features that make ethrottle worthwhile, actually work. 
Idle control, ethrottle blip, or cruise control all only apply to the first bank. Which is irritating! 

It seems that the easiest way that I can "fix" this, is to swap out the pedal position as the load axis on the ethrottle 2 table.
Swap it for "Ethrottle 1 target" 
Except, "Ethrottle 1 target" isnt something you can set as the load axis. ffs. 

What I can do though, is send "Ethrottle 1 target" to my dash over canbus, then have my dash send it back. 
I can use the canbus value as a load axis to bypass whatever reason it's not allowed. ha.
What a fuck around, I was hoping to avoid having to do janky shit like this.
But thats just V6 life I guess. haha

The car is starting to drive nicely, even though had some serious lean and rich spots on  the fuel map on first run. 
What is also funny is that on decel it kinda braps a bit sometimes when it gets to certain rpm. 
Hopefully I can sneak in another quick drive in the morning before I have to head to Invercargill for work.

I've only got a vid of it from after, idling. But you get the idea
 

 

[Roman]

I went to invercargill. It was incredibly pleasant.

Anyway, back to this bullshit 

My fuel map is based on the result from TPS 1, and then the other bank will just run rich or lean depending on how wrong its throttle angle is.
So I've been simultaneously revising my fuel map, and adjusting the left to right throttle balance too.

In this table below you can see that when throttle 2 is at "30 percent" throttle one needs to be at 19 percent to match. 
and so on. A mildly annoying problem to deal with, but I've got it nicely matched up now.



The difference in the fuel map now is absolutely incredible. It was well worth the effort to get the new intake made. 
Here's the extended rpm range of what I posted a few posts above. Gains everywhere, nothing anywhere is worse. 
I was surprised to see that even as low as 3000-3500rpm there are improvements.

There might be some mild gains from doing some VVT adjustments, and perhaps it wants a little less ignition timing now that the airmass is so much higher.
Although its not running a crazy amount of ign advance anyway.

 

I can confirm that its pretty quick now.
I am open to the possibility that the dyno results arent any where near the above estimates.
However whatever the number is. It revs out amazingly, and its plenty fast enough for me. 
I've still got issues with clutch engagement, I need to fit the bias valve for the brakes, need a wheel alignment, and a few other fiddly things.
But I'm comfortably on track for drag day. So long as nothing blows up in the meantime.
It doesnt feel like it's out of puff, it just keeps going until the limiter starts kicking in if you havent shifted by then.
Pretty stoked that it's all coming together now! 
Will get some video soonish 


 

[Roman]

I swear by now I've redone almost every single thing on this car about 5 times. 
The carina of Theseus
Anyway, 

My e-throttle woes.
The cable pull setup and these TPS sensors just isnt good enough to get the car driving as nice as I'd like. 
I read something that made me realize that these TPS sensors are not "sprung" and thats probably half of the problem.
If you put a screwdriver in and turn the center, it just stays where you left it. 
That means, if there's any play in the TPS on the shaft. Vibrations can potentially make it wiggle or whatever. 
These TPS are from/for an E36 BMW, which doesnt use the TPS for much - likely just accel fuel and figuring out when to go into idle condition. 
So its likely only looking for very large broad changes in throttle.
I'm trying to use it as the load axis for fuel/ignition, and also try get ethrottle matching to 0.1%.
Which it's just not good enough for. 
One of the problems that the above situation causes, in conjunction with my cable ethrottle setup. 
Is that sometimes when the system lands at home position. It might report 0%, it might report 1%, it might report 1.5% or whatever. 
Give it a blip on the gas and the number changes, despite landing on the same end stop each time. 
Which I think this is due to backlash and the TPS being unsprung. 

The problem is that if my idle target is 0%, and it cant reach it. Ethrottle will try "push" the system closed, which it cant do with a cable. So it just feeds out the slack of the cable.
Then when you next tap the gas, and it pulls the cable in a bit and nothing happens. It then gives it a big yank and is twitchy. 

So, sticking with what I know and have. I've decided to adapt it to the Toyota TPS sensors. 
These are sprung, and since they're designed for e-throttle. Have presumably a finer number of windings that give a more steady output down to small %. 

I've only got a plastic prototype on at the moment, but seems like the concept will work and fit well enough.



The gearboxed ethrottle motors turned up, that I can connect directly onto the end of the throttle body.
Even with the shitty TPS still in place. The performance is way way way better! 
The response is good. These have around a 30:1 gearbox ratio.
I can now significantly ramp up the PID settings and it's all a lot snappier. 
i dont like the rigid coupling, but it was good enough for testing. So I've bought some flexible ones. 
This will work a lot better, and fits nicely in space I've got available.

I setup a test routine where it cycles through different throttle positions on a timer. 
Which is handy for setting up PID but also to see if it's getting too hot at all.
After a few hours running, it gets to just under 50 deg when it's in ambient conditions with no engine heat. 
So I'll probably try to get some cooling over these to deal with engine bay temps.
But also keep a few spare motors on standby in case there are any issues.

 

One win that I've had today though. 
The car has been very hard to change gear, ever since the engine swap.
One of the things I've done, was a short shifter kit install.
I pulled the shifter off, and found that the plastic bush at the bottom was broken. 
The short shifter kit has/had a larger ball on the bottom than standard. 
So I've gone back to standard throw, and a brass bush to suit. 
Hopefully it changes gear nicely now, that was one of the last major concerns with drivability.
It will be a big relief if that's fixed.

I've also printed some trumpets which are about 20mm shorter, as seen in yellow here. 
Hopefully these will move the powerband bump slightly north and we'll see some good peak power gains next time I can get up to KPR's dyno.




On the whole, pretty good progress and my list of issues are getting smaller and less critical. 
Fingers crossed it continues on this trajectory! 

[Roman]

Quote

Fingers crossed it continues on this trajectory! 

 

It didnt 

 

Hopefully it changes gear nicely now

 

It didnt 

Swapped out the gearbox for a spare. 
New ethrottle setup is working well so far, very snappy! 

Just bleeding coolant and sorting a few similar tasks, then ready to see if box is any good. 

Looking on track for drag day, so long as gearbox is fine and no other major issues 

 

[Roman]

After a bunch of hectic late nights I got everything sorted and car on the trailer for drags. 

Passed scruitineering and ready to roll! 

 

I had a whole bunch of hectic late nights and was starting to question my sanity for trying to get to drag day. 
But I did it, and that in itself put a big grin on my face. 

Cool stuff first... Banged out some big rpm! Almost got to my initial goal of 10k rpm and nothing blew up.

The quarter mile times were slow as I wasnt trying any launches yet, and gear shifting was still a little problematic (or maybe the driver was)

But made some cool noises! 

My main problem of the day, and what made me call it a day after only 3 runs. 
Was that the ethrottle motor on one side was shutting down sometimes. 
What makes this doubly bad, is that this was the bank that has the TPS that my fuel map runs off. 
So when the throttles start to shut on that side only. The other bank runs lean and gets a bunch more timing pumped into it. Not good. 
So decided to call it quits before I damage the engine for unfun reasons. 

I suspect that the reason it had a problem, is my ethrottle motor bracket started warping.



This yellow bracket was only meant to be a dimensional test piece. 
I finished printing an incredibly solid version from PETG-CF, that finished mid day, on day before the drags. 
Unfortunatley I found that the only reason the yellow one actually fit, was because it could flex slightly. 
As it was actually 3-4mm out. So there was no way my other one would fit, and I didnt have time to print another.
So I was lucky that it still ran at all and I could get some passes in. 

For my 3rd run I tried putting  my drag slicks on, and had the launch control setup. 
But ethrottle packed up almost immediately so it bogged off the line and just died. 

I definitely need to put a better final drive ratio in. 
First gear was going to 70kph, 126kph in 2nd, then crossing the line at 8700rpm in 3rd. 
I'm not making use of any of the good gears in a J160 box! 
I dont think a 4.7 will make enough of a difference so I'll look at ordering a 5something at some point. 
It will probably pay to get some brand new driveshafts made when doing this, as the shaft speed is going to become considerably higher.

On the whole though, super stoked to have made it there, and made it back with a car that still goes. 
I've got some problems to sort, but they're all sortable. Keep plugging away it! 

[Roman]

I had the 20mm shorter trumpets on, and since I figured this would make it run leaner/richer in some areas I added 5% fuel in across the board before doing any runs. 

Then updated it after each run. 

So was fairly well dialled in by 3rd run.

The graph below is fuel table values needed to reach target, before and after. (VE)
It looks like I've traded the 7k rpm dip for one at 5k instead. 
On the whole it's probably a better length because once first gear is done, there's no reason to ever see 5k rpm again. 
Once the car is better sorted it might want to launch at higher than 5k anyway.
Once you're in the right sort of intake length range, it's interesting to see how sensitive it is to trumpet length changes. 

I've also not done any VVT amendments since having my old intake on. It will most likely be wanting some more intake cam advance. 
But another dyno visit might be easiest way to figure that out from here.

[Roman]

Coming back to exhaust noise stuff. 

From the drags, watching the video back to back compared to the Echo.
The car sounded less screamy/high pitched/ high rpm than the echo did. 
Even though it's got 50% more cyls and was revving higher! 

So looking back at my chart thingy, it paints a picture...





My motor is making the noise of a 3 cyl motor instead of 6 cyl motor. 
This is likely because my 2-1 section is unequal length. Because lazy

I'm guessing that the pulses from each bank are landing right on top of each other, and travelling to the end of the car as a big doubled up pulse.

It will be cool to see if I can split the pulses back apart and get the higher frequency happening. 
My current exhaust is rather crap anyway, for a variety of reasons.
So will redo nearly all of it soonish, and hopefully I can get a nicer pitch out of the engine with some tweaks. 

Will make the 2-1 section equal length as best I can, and then run an x pipe with hopefully 2 pipes down to the back of the car. 

Using the an audio editing program I can pitch shift the sound clip to be 100% higher frequency. 

So this is roughly what it would sound like if I can split the pulses back up