Stage 1
07-24-2013, 01:47 AM
#76
Yellow looks faster.
Thread Starter
iTrader: (1)
Join Date: Jun 2012
Location: North NJ - Utah - Arizona
Posts: 445
Likes: 0
Received 0 Likes
on
0 Posts
Hmmmm. I'm gonna bring it in for a good servicing when I can afford it. Do the injector service, and then replace the coils,plugs,wires myself.
07-24-2013, 08:18 AM
#77
Registered
Join Date: Jul 2009
Posts: 43
Likes: 0
Received 0 Likes
on
0 Posts
I did not get any proper logs. What are the things I should be logging? and the proper order to do it? This is a half ***'d 2nd gear pull. not even WOT. I'm just seeing how it works. Will get some proper logs as I learn more. Idle AFR is 15.0 ish and normal driving around 3k rpm is about 14.7 ish.
My Idle AFR is always at 18, not sure why it would be at 15, though I'm running a MM tune. /shrug
07-24-2013, 10:24 AM
#79
Registered
Join Date: Jul 2009
Posts: 43
Likes: 0
Received 0 Likes
on
0 Posts
Interesting, maybe I'm remember incorrectly. I think I'm confusing it from downshifting or coasting on the highway and releasing throttle. At any rate at WOT I reach basically just a micron below 12, which I assume for the most part is ok. Though now with my rebuild I can't go into boost for another 150 miles or so. My compression was low before, so I wonder how the dynamics will change, I'm assuming I'll need to retune anyway for some many reasons. Higher compression on rotors, different altitude and weather CA->MD as I mentioned in another thread.
07-24-2013, 11:35 AM
#81
VE will be entirely dependent on your setup. Numbers are completely arbitrary and made up until they are logged and compared with reality on your specific setup.
I merely know how to look at the hard data in front of me and tell what adjustments need to be made. There are engines where I can pull a number out of my head and it will be close. This is only because I've tuned hundreds of them. I haven't done enough rotaries to be able to know what numbers look good and what don't.
I know the COBB software very well. I know Otto cycle internal combustion very well (rotaries are still an Otto cycle engine). I bought my RX-8 because I knew absolutely nothing about rotaries up until early May when I decided to impulse buy one so I could learn new things.
The VE table can kind of fudge things on a boosted car.
Boost is basically your calc load. As boost increases, calc load increases. In those higher calc. load tables, you will end up having higher VE values (forcing the ECU to run a higher IDC to compensate for the relative drop in fuel pressure). It's a bit of a hack, but it works OK.
Honestly, for boosted cars the best option is to switch to a rising rate fuel pressure regulator that increases fuel pressure by 1psi for each 1 psi of boost (as this keeps the relative head pressure in the injector the same at all times). This isn't a small undertaking, as you basically have to switch over to a return style fuel system. This is a purely mechanical and elegant solution. It eliminates a whole bunch of variables that are otherwise impossible to pin down. Bonus points for the fuel pressure changes being instant.
It's a mod that will give you absolutely zero extra power. For reliability purposes it's huge. There are situations at part throttle where you can still have very high manifold pressure, but very low airflow rates. This will bring your calc. load way down, as it's strictly monitoring airflow. This will result in less than ideal IDC and force the engine lean. If you have a built in mechanical method to tie fuel pressure to boost pressure, it eliminates those situations.
My personal theory thus far, is that a lot of these boosted RX-8s that are popping motors despite a solid dyno tune, are due to part throttle lean conditions that were never seen on a dyno.
Difficulty depends on where you place your regulator/return.
If you totally ditch your OEM fuel rails it becomes a huge project. If you do return at some point in the engine bay before the OEM rail, it's not too bad.
I've never done this on a Mazda. I don't know the exact specifics of what's involved. I've done it on a ton of other cars though.
I merely know how to look at the hard data in front of me and tell what adjustments need to be made. There are engines where I can pull a number out of my head and it will be close. This is only because I've tuned hundreds of them. I haven't done enough rotaries to be able to know what numbers look good and what don't.
I know the COBB software very well. I know Otto cycle internal combustion very well (rotaries are still an Otto cycle engine). I bought my RX-8 because I knew absolutely nothing about rotaries up until early May when I decided to impulse buy one so I could learn new things.
The VE table can kind of fudge things on a boosted car.
Boost is basically your calc load. As boost increases, calc load increases. In those higher calc. load tables, you will end up having higher VE values (forcing the ECU to run a higher IDC to compensate for the relative drop in fuel pressure). It's a bit of a hack, but it works OK.
Honestly, for boosted cars the best option is to switch to a rising rate fuel pressure regulator that increases fuel pressure by 1psi for each 1 psi of boost (as this keeps the relative head pressure in the injector the same at all times). This isn't a small undertaking, as you basically have to switch over to a return style fuel system. This is a purely mechanical and elegant solution. It eliminates a whole bunch of variables that are otherwise impossible to pin down. Bonus points for the fuel pressure changes being instant.
It's a mod that will give you absolutely zero extra power. For reliability purposes it's huge. There are situations at part throttle where you can still have very high manifold pressure, but very low airflow rates. This will bring your calc. load way down, as it's strictly monitoring airflow. This will result in less than ideal IDC and force the engine lean. If you have a built in mechanical method to tie fuel pressure to boost pressure, it eliminates those situations.
My personal theory thus far, is that a lot of these boosted RX-8s that are popping motors despite a solid dyno tune, are due to part throttle lean conditions that were never seen on a dyno.
Difficulty depends on where you place your regulator/return.
If you totally ditch your OEM fuel rails it becomes a huge project. If you do return at some point in the engine bay before the OEM rail, it's not too bad.
I've never done this on a Mazda. I don't know the exact specifics of what's involved. I've done it on a ton of other cars though.
thanks for posting, I had come to the conclusion that the ve table would be a little fuzzy for those boosted areas and just wanted to confirm. it would be nice to keep using calc load but our system won't go past 200% for the cobb.
a RRFPR would be ideal but I am not sure how it would work with the rest of the system. currently we run around 60-70psi of fuel pressure from stock, and somewhere along the line is a High and low setting for the fuel pump. I can't remember how that is triggered. I wonder if you could have it triggered with a hobb switch instead...
07-24-2013, 02:01 PM
#82
That doesn't happen on our engines . You don't see high manifold pressures unless you have corresponding high air flow . What we do see is very high pressure pre throttle which can lead to some interesting moments mid corner if boost controller/BOV is not set up to minimise this. This will not cause a lean condition however.
I don't imagine so, as the factory ECU doesn't have a parameter for this.
07-24-2013, 04:29 PM
#91
The only time I can imagine this happening is in the instant after you close the throttle between gear changes as the engine evacuates the air from the manifold. Logs I have don't show anything that suggests it's an issue though.
Pre throttle ...............I've seen over 20psi at low load/ part throttle !
Last edited by Brettus; 07-24-2013 at 04:59 PM.
07-24-2013, 09:06 PM
#93
It relates to what I was alluding to before about high boost pre throttle in certain situations . This is part throttle/medium load as in for example .... Going up hill at medium revs/light throttle then opening the throttle a little more .
There is some load on the engine but throttle is almost closed . Boost builds up behind the throttle plate and then as you open it a little more you get a sudden inrush of air . In this scenario I can see the effect you are talking about - where the maf doesn't catch up for a split second.
I experienced this early on - I can't say whether the engine would run lean or not. But I would have noticed a detonation if one occurred.
What I did about it was connect the boost controller line pre throttle instead of post throttle as is common . What that did was significantly reduce the amount of pre throttle boost in these situations .
Last edited by Brettus; 07-24-2013 at 09:11 PM.
07-24-2013, 10:01 PM
#94
Yellow looks faster.
Thread Starter
iTrader: (1)
Join Date: Jun 2012
Location: North NJ - Utah - Arizona
Posts: 445
Likes: 0
Received 0 Likes
on
0 Posts
So I have been looking into how to search for the vacuum leak. How do you guys prefer to do it?
It seems the common way to do it is to use carb cleaner and listen for idle changes. Thoughts?
It seems the common way to do it is to use carb cleaner and listen for idle changes. Thoughts?
07-24-2013, 10:44 PM
#97
Registered
Join Date: Jul 2011
Location: San Diego
Posts: 38
Likes: 0
Received 0 Likes
on
0 Posts
20.29 3235 7.45 -0.16
20.29 3174 7.06 -0.16
20.29 3050 7.06 -0.16
20.29 2909 7.06 -0.16
20.29 2860 7.45 -0.16
20.29 2809 50.96 -0.16
20.29 3469 82.32 -0.16
20.29 4080 80.75 -0.16
20.29 4262 83.5 -0.16
20.29 4057 83.5 0.62
20.29 3938 83.89 10.76
20.29 4109 83.5 23.24
19.84 4034 83.5 24.8
17.2 4050 83.5 24.8
16.61 4098 83.5 24.8
14.99 4053 83.5 23.24
14.55 4042 83.5 22.46
13.82 4118 83.5 19.34
13.08 4105 83.5 16.22
12.79 4125 83.5 10.76
12.79 4157 83.5 8.42
12.94 4156 83.5 6.08
13.08 4176 83.5 4.52
13.23 4194 83.5 2.96
13.52 4218 83.5 2.18
Coming out of open loop should i have this high of stft?
20.29 3174 7.06 -0.16
20.29 3050 7.06 -0.16
20.29 2909 7.06 -0.16
20.29 2860 7.45 -0.16
20.29 2809 50.96 -0.16
20.29 3469 82.32 -0.16
20.29 4080 80.75 -0.16
20.29 4262 83.5 -0.16
20.29 4057 83.5 0.62
20.29 3938 83.89 10.76
20.29 4109 83.5 23.24
19.84 4034 83.5 24.8
17.2 4050 83.5 24.8
16.61 4098 83.5 24.8
14.99 4053 83.5 23.24
14.55 4042 83.5 22.46
13.82 4118 83.5 19.34
13.08 4105 83.5 16.22
12.79 4125 83.5 10.76
12.79 4157 83.5 8.42
12.94 4156 83.5 6.08
13.08 4176 83.5 4.52
13.23 4194 83.5 2.96
13.52 4218 83.5 2.18
Coming out of open loop should i have this high of stft?
07-25-2013, 10:48 AM
#98
Registered
Join Date: Jul 2013
Posts: 103
Likes: 0
Received 0 Likes
on
0 Posts
^^^ you're not in open loop quite yet.
And sure why not?
You've just gone thru the area where the SSV has likely open and the STFT is compensating for that ..
2% isn't bad as far as I can tell following this thread.
Remember there is Rpm vs Load and Thottle Duty cycle vs rpm and a EXIT Rpm, plus I bet other tables we don't have access too that dictate when the transition to OL occurs.
And sure why not?
You've just gone thru the area where the SSV has likely open and the STFT is compensating for that ..
2% isn't bad as far as I can tell following this thread.
Remember there is Rpm vs Load and Thottle Duty cycle vs rpm and a EXIT Rpm, plus I bet other tables we don't have access too that dictate when the transition to OL occurs.
Last edited by EviLStewie; 07-25-2013 at 10:56 AM.
07-25-2013, 11:27 AM
#99
So I have done some Logs and put those into the Fuel VE table.. So far.. My table looks like it has a mountain rising out of it..
My first logs showed high percentage increases in some cells. But following logs they did not show that high.. I got them within +/- 2% or so for the cells I hit.. So now I have this kinda mountain raising out of the 3D graph for my log.. Do I need to now Blend them In to sort of smooth out the peeks and valleys and such...
Screen shot of what im talken bout.
My first logs showed high percentage increases in some cells. But following logs they did not show that high.. I got them within +/- 2% or so for the cells I hit.. So now I have this kinda mountain raising out of the 3D graph for my log.. Do I need to now Blend them In to sort of smooth out the peeks and valleys and such...
Screen shot of what im talken bout.
