Romans 2005 Toyota Echo

[Toucan]

Install fourth pedal in footwell to control a linkage to butterfly? Or convert to auto an use the old clutch pedal?

[kicker]

1 hour ago, Toucan said:

Install fourth pedal in footwell to control a linkage to butterfly? Or convert to auto an use the old clutch pedal?

Something like the V-TACS pedal on the Honda Beat

 

[tortron]

1 hour ago, mjrstar said:

Fit teleflex cable to butterfly, then add a choke cable knob from a Morris minor mounted on the dashboard. No need for that vacuum fuckery.

Lucas vacuum reserve setup would be ideal

[CaMpylobacter]

have you thought of 10 helicopters

[Stu]

12v vac pump

https://www.mindkits.co.nz/store/p/8865-Vacuum-Pump-12V.aspx

 

 

[Roman]

^Might try that vac pump @Stu! If not for the valve, could be good for just running on the booster so I dont need to tap into one of the runners. 

I found my graduated cylinder so its time for some more fire hazard-ish testing, to try characterize the injectors a bit better. 




So the first test is just to find the CC rating of the injector. 
This is simply turning the injector on full blast, counting the amount of time it's open for, then seeing how much fuel came out. 
So in the ECU I am turning on the fuel pump ahead of the test, then turning the injector on and off using the ECU too. 
So I can look at the logs and find out the exact time period it was open for.

So in this case: 


These injectors are rated at 305cc at 300kpa, however I'm not sure what my fuel pressure is with the returnless setup. I thought it was 325kpa. 
However using an injector pressure calculator, it looks like it's probably 360kpa ( 52psi)


The next important part is working out the injector deadtime. 

If an injector could turn off and on instantly, you'd find that spraying 1 time at 10ms would deliver exactly as much fuel as 2 x 5ms. 

But each time the injector is commanded to open, there's a physical delay until fuel can first come out. 

So 5x2 does not equal 1x10! 

As a visualization lets imagine each ball of fuel represents spray time.
First instance is how long you tell it to open for.
The second result is how much fuel you end up getting.


As you can see, by spraying different quantities and time periods that should theoretically give you the same amount of fuel, the results will vary.

My first test ends up spraying at 10,000 times at 8ms per spray. Because it's not really feasible to measure just 1 or two sprays at a time.
Then second test is 20,000 times at 4ms, then the third is 40,000 times at 2ms. 

The results are as follows: 


With some Excel fuckery, we can figure out how much time is missing from each individual spray event, by trying to balance the results. 

There's probably a smart maths way to do it, but basically I just adjust the deadtime number up and down until all 3 of the columns are the same height.




So for yellow 2ZZGE injectors, deadtime is around 0.515ms at 12.6 volts. 

Ideally you do this same test at all of the different voltages, but this at least tells me the ballpark of where the motor will run at 13.1-13.6v which are the operating conditions.
My existing deadtime values were too high, around 0.7ms at that same voltage before.
This existing number was from guesstimating by adding 10% fuel to the running engine, and then seeing if the results changed by more or less than 10%. 
Then adjusting up or down until it sorta matched. Not a perfect method but gets you somewhat close.

Now it should be able to adjust itself with closed loop lambda or temperature corrections a little more accurately. 
It also means it takes fewer iterations to correct the fuel map, if it's off target when testing new runners or whatever.

The end result of all this is that my fuel table will end up changing shape a bit.
But once it's done this time it should stay pretty good even as altitude or whatever changes.

[yoeddynz]

Pacman graph 

[ajg193]

Easiest way to calculate your dead time is just graph it and fit linear trend line to see intercept - remember to divide flow by the multiple of pulses 

This run was 323cc/min

[ajg193]

Seeing how alternators seem to be a pretty decent drag on the engine, can you program an output on your ecu to turn off the alternator above 90% throttle and 3000 rpm?

Would be interesting to see if it makes a measurable difference on your virtual dyno

[mjrstar]

I think bmw (maybe others too) do this a bit to improve fuel economy.

It kicks the alternator in when the engine on the overrun or when battery voltage drops to a certain level.. 

[Yowzer]

Most vehicles do this nowadays. Which is also why they have such big batteries, reduce alternator output while cruising and ramp them up on deceleration.

[RUNAMUCK]

My mums vitz gets used fuck all, and subsequently I have to charge the battery periodically.

[Roman]

Ahhh yeah I guess that miiigghhht help at the drags but not going to be useful for sustained hooning. 
I guess if you do the maths, if you were drawing full 60amp from alternator, and it was outputting 14 volts. 
That's 840 watts 
Then if its something like 25% efficient at converting mechanical energy into electricity, would be 3.3kw draw on the motor. 
I guess that is a decent amount! 

However, looking at the logs from the drags, my voltage drops by quite a lot with high rpm anyway.
Maybe it's alternator rpm based reduction, maybe its belt slip, maybe it's Maybelline?



Not much interesting stuff to report otherwise at the moment, as I'm just busy trying to fix some maintenance issues.
Needs the front ball joints and a tie rod end replaced.
Some better brake pads ordered as well, as I'd like to do some trackdays this summer hopefully. 
I will try get a better radiator too, as this one has had 200,000km worth of bugs smash into the lower 1/3rd of it.
I'm waiting on parts that should hopefully arrive in the coming week.

My big goal at the moment is to try get it all ready to attend 4&rotor drags and possibly the trackday part which is about another 6 weeks away. 
It looks like I might have a work trip to Australia the same week though. So fingers crossed that gets postponed or brought a week ahead so I can see how it goes.

[fletch]

On 11/09/2022 at 23:15, Roman said:

However, looking at the logs from the drags, my voltage drops by quite a lot with high rpm anyway.

Maybe it's alternator rpm based reduction, maybe its belt slip, maybe it's Maybelline?


 

Maybe it's just electrical load. Your injectors and coils would be working hard as rpms rise, and there is only .6v difference.

At 6:20 the voltage jumps straight up as soon as the load drops.

[kpr]

depends where the ecu is feed from also, quite common to get volt drop if wiring not up to it.   likely find if measure at alternator, will be holding stable

I'll try disconnect alternator next time have something on dyno

 

 

[Roman]

I feel bad that people pitched in for a new motor for me, but now there's a bit of a lull in updates.
The dyno guy is busy for a while, and the next thing to do apart from that is... really just get to some events and see how it goes!
So I'm looking forward to that. 
There is a Pukekohe trackday in about 2 weeks time that I've got an eye on. 
4&Rotor nats would be fun for sentimental value / have never competed, but I get the feeling it's going to be a whole lot of sitting in queues and not much racing. 
Then of course oldschool drags but that's a while off yet.
I'll try find a drag comp meet where they do a proper track prep during the daytime. 

So while I've had not much to do, and since from the new motor purchase I only really needed the block.
I've now got a spare head for some R&D.

Last time with the porting the most major change I made, was downsizing the injector bump in the intake ports. 
Then the rest was just a bit of a tidy up and just minor smoothing.
I was quite conservative about removing material from the combustion chamber as I was concerned about dropping the compression ratio.
However this time I'm going to go all in, on improving air flow and not care so much if the CR drops.
If the CR does drop too much, it's still not a waste. As the head might then be useful when I eventually get some aftermarket pistons. 
Will just choose higher CR piston than I need, and the extra chamber volume will drop it back down to target.

I've cut back all of the machining ridges around the outer edges of the valves, a bit more than I did last time.
But I think I'll spend a bit more time putting a radius all the way around the outer edges as best I can.

Firstly 

I saw this picture below of a ported 1NZ head, for a turbo setup.
You can see they have gone hardcore to reduce shrouding around the valves by adding a big radius on the far side of the combustion chamber. 
I dont like that they have cut back the squish bands, as this will have dropped CR a lot. 
However it looks like they have done this to promote better flow when lift is higher.
I'll definitely try find the right shaped dremel tool to get a nice radius on the outer sides of the valves, but I wont go this extreme.
I guess I could get the head milled down slightly if I needed, but then I will lose a bit more usable VVTI range.




Then I'm also going to spend a bit more time thinning the dividers between the ports. 
Man this takes ages. 
I've been reading a lot of back and forth about whether it's good to knife edge the divider between the two valves.
However thinking about almost anything that has liquid or gas flowing around it.
Things always seem rounded at the front, and sharp at the back.  Teardrop shape.
So I think a sharp divider will likely be beneficial on the exhaust side, but a more rounded divider on the intake probably makes more sense.
Also since the dividers are quite short, the only way I could fully knife edge them would be to have an angle that's too steep. 
Will see what I can do.




If I had a flow bench setup already, I'd port each of the 4 cylinders differently and then compare them all. 
However that's not going to be doable for a long time, so I'll have to stick to a best effort approach and compare real life results.

 

[ajg193]

[deankdx]

14 hours ago, Roman said:

If I had a flow bench setup already

does this mean you are making one? or borrowing one?  

i saw a build thread years ago of a home made flow bench that worked well. 

 

[Roman]

I had a bit of a scheme for making a semi automated bench but I got a bit stuck with needing way gruntier fans to pull enough air. 
Which then meant I would need a much stronger housing and the whole thing would become big and complicated, aaaaand I ran out of effort.
Still want to make one though, one day.
A lot of the modern DIY ones use a MAF because it's an incredibly better way to go about it.

[Roman]

I found there is an online book/audio book app called Scribd, where the first months subscription is free - and David Vizard's book "How to flow test and port cylinder heads" is on it. 
So you can sign up and if you unsign in first month read it for free, might have some other interesting tech stuff on there. Will just have to try remember to unsubscribe before a month's up. Hah.
So, I read through most of this and unfortunately very little of it contained information which feels relevant to a 4 valve head.
As there's not really any testing done on things like port dividers, because, its mostly just 2 valve motors. Bummer.

However one interesting thing was that it said a valve is fully deshrouded, when the valve has reached 0.25x its diameter in lift, and has 0.2x the valves diameter in clearance from the walls. 
So scaling against a photo (which isnt even straight, meh) but can get an idea of the fully deshrouded area of the valve

""

 

So theoretically there's value in cutting back the squish pad thingies on the sides, or at least smoothing them out a bit more.
Also looks like anything around the far side of the valve towards the cylinder wall, up to edge of the cylinder is fair game.

Just to reassure myself, looking at some calcs again of compression ratio drop, I dont think I'll make any meaningful drops. 
with what I'm planning to do it might drop to 13:1 or something like that, definitely not down to 10:1 or anything. I'd have to carve out absolute shitloads to drop it by that much.
Might end up at 13:1 or maybe 12.5:1 worst case scenario. 

Actually thinking about it some more, 1 cubic centimeter is fucking heaps haha. 




Dad's CNC machine has turned up, he's slowly troubleshooting issues and getting things fixed. 
I'm thinking it would be cool to be able to get a CNC profile setup for some combustion chamber mods and smoothing out the intake bump.
As they are fairly simple mostly circular paths with a ball nosed cutter. (Simple, says person who knows nothing about it)
This way I could be sure everything is exactly the same between cylinders.

Also, have signed up for a Pukekohe trackday this Friday, exciting! Looking forward to it.