Roman's 4GR V6 Carina discussion thread

[shrike]

@Roman did you track down a 1mz crank?

[Roman]

A friend in Auckland has a 2MZ for me, just have to organize a time to grab it. 
Been sick with flu last few weeks though so not much going on. 
Still waiting on an ethrottle motor to turn up before I can carry on with much else.


 

[yoeddynz]

On 20/06/2025 at 09:37, shrike said:

Ive heard that V motors sound better with an X pipe.

And flat motors can benefit too as I have found out   I figure its always going to be dependent on many other parts of the puzzle but if I was to muck about with another Mazda v6 I'd be doing two twins all the way to the back with an x pipe included into the sysytem. My Viva v6 merged just after the gearbox into a single all the way back and I now feel that cost it a lot of the nice V6 sound it could have had. Certainly muffled it down 

@Roman is there room for you to run two pipes back?

[Roman]

Yeah I am thinking later on I will probably do an X pipe and 2x pipes to the back. 
All of the good sounding 2GR MR2s are X pipe and 2 pipes to the back.
However, in the fairly immediate future I just need to consolidate what I'm doing and get the car to a drivable state. 
The current exhaust is good enough to get the car onto the dyno and so on. 
I'm keen to put down a baseline number, then put big cams in, then see how it's going. 
I've been very lucky that Stu has loaned me his tig setup over the last long while, but he needs it back to make some progress on his own projects.
So no more exhaust stuff for now.
So I'll be putting a tig setup into the budget at some point. Super 5000% glad that I persisted with it and can now weld alright. 
Invaluable skill.

[440bbm]

That is very over square. Is there anything with a longer skirt to help with wear? motorcycle pistons are commonly very short like that compared to automobile types.
do you know how short it needs to be/shape required to clear? 

[tortron]

22 minutes ago, 440bbm said:

commonly very short

have you seen thr crf450 pistons? they are literally just ringlands with a boss for the little end 

[shrike]

12 hours ago, Roman said:

Yeah I am thinking later on I will probably do an X pipe and 2x pipes to the back. 
All of the good sounding 2GR MR2s are X pipe and 2 pipes to the back.
However, in the fairly immediate future I just need to consolidate what I'm doing and get the car to a drivable state. 
The current exhaust is good enough to get the car onto the dyno and so on. 
I'm keen to put down a baseline number, then put big cams in, then see how it's going. 
I've been very lucky that Stu has loaned me his tig setup over the last long while, but he needs it back to make some progress on his own projects.
So no more exhaust stuff for now.
So I'll be putting a tig setup into the budget at some point. Super 5000% glad that I persisted with it and can now weld alright. 
Invaluable skill.

Would you need custom bearings if you grind down the crank journals for more stroke? 

Actually now I think about it you only need oversized bearings when you grind it down if your using factory rods. So I guess if your making custom rods you'll add the extra length there.

[Roman]

4 hours ago, 440bbm said:

That is very over square. Is there anything with a longer skirt to help with wear? motorcycle pistons are commonly very short like that compared to automobile types.
do you know how short it needs to be/shape required to clear? 

Ahh so yeah, I actually wanted something with as little skirt as possible, and the piston pin as high as possible.
When your rod ratio number goes higher, there is less side loading on the bore so skirt isnt so important. 
If you have a 1.5 rod ratio, when your crank is at 90 degree position.
The rod is pushing the piston on an 18 degree angle into the side of bore. 
If you have a 2.3 rod ratio, its only 12.2 degrees.

@shrike Will just stick with normal 2MZ crank geometry for starters on that other motor I think. But I think there's some maximum oil shear speed or something like that. Which might mean that my big end or main bearing sizes are both too big for lots of rpm.
Not sure yet. As per usual though its probably the crappy heads which will give up first.

[440bbm]

13 hours ago, tortron said:

have you seen thr crf450 pistons? they are literally just ringlands with a boss for the little end 

yes and there are also reason why they are a complete piston change with the top end and valves after like, 100 hrs. 
 

but yes single piston doing alll the work opposed to 6, may not be so bad. 

 

[440bbm]

9 hours ago, Roman said:

Ahh so yeah, I actually wanted something with as little skirt as possible, and the piston pin as high as possible.
When your rod ratio number goes higher, there is less side loading on the bore so skirt isnt so important. 
If you have a 1.5 rod ratio, when your crank is at 90 degree position.
The rod is pushing the piston on an 18 degree angle into the side of bore. 
If you have a 2.3 rod ratio, its only 12.2 degrees.

@shrike Will just stick with normal 2MZ crank geometry for starters on that other motor I think. But I think there's some maximum oil shear speed or something like that. Which might mean that my big end or main bearing sizes are both too big for lots of rpm.
Not sure yet. As per usual though its probably the crappy heads which will give up first.

yah cool you have done the math as expected. just wonder if you have mathed longevity or factors to that where as bike motors are not ( so much ). 
i spose you're not worried for what you are building. not like it's a daily.

[Roman]

If the motor lasts longer than an oil change, its a success at this point 

Haha but yes, definitely building for a good time not a long time.

Although bikes generally do have a really nice short stroke length, often their rod ratios can be quite low because they're also trying to make the motor as physically compact as possible.
Like I was surprised to see that a lot of the high rpm bikes have a rod ratio of say 1.6 or 1.8 or something. 
But it's just to keep the motor smaller and lighter I'm guessing.
 

[shrike]

1 hour ago, Roman said:

If the motor lasts longer than an oil change, its a success at this point 

Haha but yes, definitely building for a good time not a long time.

Although bikes generally do have a really nice short stroke length, often their rod ratios can be quite low because they're also trying to make the motor as physically compact as possible.
Like I was surprised to see that a lot of the high rpm bikes have a rod ratio of say 1.6 or 1.8 or something. 
But it's just to keep the motor smaller and lighter I'm guessing.
 

Ducati 916 and 848 are also 94mm and a road bike so tend to do a few kms, I know of at least one 848 thats done 100000km without a rebuild

New pistons seem expensive though, 2nd hand might be better pricing

[fuzzy-hair-man]

It's a 2GR and I'm not all that fussy about engine sounds but this sounds clean and F1 like @22.40:

 

[Stu]

Can I just say fuck that's excellent progress! Can't wait to hear all the dorts

[Roman]

Yeah! Real exciting. 

Good to have some confirmation that needing things like 4x different PID values for each of the gears is "normal".
Hopefully doesnt take too long to sort out. 
But this is definitely one of the benefits of getting some road tuning done initially.
Figure out all these types of things and get them sorted first, so hopefully when it comes to dyno time I can just progress straight to throwing a rod out of the block 

Also this motor is VERY touchy on first throttle press. So it's great to have the ethrottle setup, will pull a bit of sensitivity out of the first half of pedal press.

Will be awesome to have cruise control working again too, but I need to find where my switch ended up...
 

[h4nd]

Neat! good jerb, that Flu bashed that crap out of me a coupl'a months ago.

Jiggles: I'd have guessed you'd want to back off the PID 'i' term https://en.wikipedia.org/wiki/Proportional–integral–derivative_controller,

but that seems to contradict issue with the drift in the off/idle position? Most PID loops I've seen seem to end up only PD, or just a tiny bit of i (maybe for long term sensor drift?).

I'd be inclined to drop an actual 'scope on the feedback input to the ECU, looking for noise provoking the PID response.

 

Looking forward to a bell curve of data-points for typical rod-throw RPM.

We better start a RomanDayf slush fund to encourage RPM-abuse, and coffee and lamingtons for sustenance whilst changing blocks.

[Roman]

Ahh yeah so two different situations

 

The first is the exhaust side pulleys, these currently have NO PID action going on at all, but the positions are drifting. So something mechanical.
But one of the guys said if you add a minimum duty cycle amount to the solenoid, oil goes a little bit somewhere extra (or something) and then it helps a bit.

Then the intake side is where one is waving around, and needs PID tuning. 
Exhaust side might need PID adjustments too, but not sure yet.

So one quirk of this ECU. It can log straight to the laptop at 40hz, which is why I've got the crappy zigzag looking logs.

Then I can log at 100hz to a logfile in the ECU. Ooorrrrr I can log stuff at 200hz by sending it over canbus.

Since it's a bit of a prick to try and tune oscillating PID when you can barely see the waveform, I'll setup my dash to show some results from 200hz data and get it dialled in. Unless of course, its only a simple change and it all starts behaving, in which case I shant bother. 

 

[h4nd]

Did I just figure out you said you're going to program your car dash as an oscilloscope? 

💓

 

[Roman]

You know it! 

Thankfully there are some other good variables that you can view/log/transmit. 

Like the duty cycle of the solenoid. So you can get some feedback without needing to hook up an actual scope.

I do have an 8 channel USB scope though, but no point in spending an hour using that, when I can spend only 200 hours writing code to view it on my dash...

[h4nd]

Oh man: car dash needs to have another mode, spectrum analyser for the car-audio ( / exhaust audio)

https://www.instructables.com/Arduino-Spectrum-Analyzer-on-a-10x10-RGB-LED-Matri/