Stage 1
07-21-2013, 07:03 PM
#26
The variations in flow rates are why we have manifold pressure that is different from atmospheric pressure even when the car is at WOT. There are pressure variations through the entire intake tract. You have air that stacks up in places and doesn't go anywhere (hence the pressure variations).
If the engine follows a predictable path through the VE table, (such as idle, 3k cruising, then WOT for a dyno pull) adjusting the MAF curve will work just fine, as you are basically cancelling the VE table at this point.
When you start getting into transient situations, it all falls apart, as your VE is now different than what was seen on the dyno or that street tuned predictable path through the VE table.
The bitch of it happens when you radically alter engine airflow AND your MAF housing. Now you aren't working with a known foundation.
If the MAF housing is the same diameter as stock, I usually still leave the MAF curve alone. Sometimes long term trends can spot a variation between a known good VE table and a wonky MAF table.
Personally, when working with heavily modified intake tracts (such as going to a blow through MAF setup) I pull the intake off (or as much as possible, like 1 foot before and after the MAF is sufficient) and take it to a place with a flow bench. I have a small 5V power supply (made from a cannibalized desktop PC power supply) to power the MAF. Then i use the flow bench to get an actual legit correlation between MAF voltage and airflow. Dump my flow bench numbers into the MAF cal and it's good to go.
With this done, it's a simple VE remap again.
Last edited by no-coast-punk; 07-21-2013 at 07:14 PM.
07-21-2013, 07:20 PM
#27
The amount of air actually entering the combustion chamber through the rotor port isn't the same as what is passing through the MAF. If the engine had a volumetric efficiency of exactly 100% (or 1 on the VE table) the airflow would be identical at both points. It isn't though.
The variations in flow rates are why we have manifold pressure that is different from atmospheric pressure even when the car is at WOT. There are pressure variations through the entire intake tract. You have air that stacks up in places and doesn't go anywhere (hence the pressure variations).
If the engine follows a predictable path through the VE table, (such as idle, 3k cruising, then WOT for a dyno pull) adjusting the MAF curve will work just fine, as you are basically cancelling the VE table at this point.
When you start getting into transient situations, it all falls apart, as your VE is now different than what was seen on the dyno or that street tuned predictable path through the VE table.
The bitch of it happens when you radically alter engine airflow AND your MAF housing. Now you aren't working with a known foundation.
If the MAF housing is the same diameter as stock, I usually still leave the MAF curve alone. Sometimes long term trends can spot a variation between a known good VE table and a wonky MAF table.
Personally, when working with heavily modified intake tracts (such as going to a blow through MAF setup) I pull the intake off (or as much as possible, like 1 foot before and after the MAF is sufficient) and take it to a place with a flow bench. I have a small 5V power supply (made from a cannibalized desktop PC power supply) to power the MAF. Then i use the flow bench to get an actual legit correlation between MAF voltage and airflow. Dump my flow bench numbers into the MAF cal and it's good to go.
With this done, it's a simple VE remap again.
The variations in flow rates are why we have manifold pressure that is different from atmospheric pressure even when the car is at WOT. There are pressure variations through the entire intake tract. You have air that stacks up in places and doesn't go anywhere (hence the pressure variations).
If the engine follows a predictable path through the VE table, (such as idle, 3k cruising, then WOT for a dyno pull) adjusting the MAF curve will work just fine, as you are basically cancelling the VE table at this point.
When you start getting into transient situations, it all falls apart, as your VE is now different than what was seen on the dyno or that street tuned predictable path through the VE table.
The bitch of it happens when you radically alter engine airflow AND your MAF housing. Now you aren't working with a known foundation.
If the MAF housing is the same diameter as stock, I usually still leave the MAF curve alone. Sometimes long term trends can spot a variation between a known good VE table and a wonky MAF table.
Personally, when working with heavily modified intake tracts (such as going to a blow through MAF setup) I pull the intake off (or as much as possible, like 1 foot before and after the MAF is sufficient) and take it to a place with a flow bench. I have a small 5V power supply (made from a cannibalized desktop PC power supply) to power the MAF. Then i use the flow bench to get an actual legit correlation between MAF voltage and airflow. Dump my flow bench numbers into the MAF cal and it's good to go.
With this done, it's a simple VE remap again.
What I'm interested in is why you see adjusting VE as superior to adjusting the open loop fuel tables?
07-21-2013, 09:27 PM
#32
Registered
Join Date: Jul 2013
Posts: 103
Likes: 0
Received 0 Likes
on
0 Posts
Holy CRAP!!!!
Ok I'm not as experienced as Punk or Brettus but I know a +20 LTFT is jacked.
Perhaps a vacuum leak?
What are your mods again?
Ok I'm not as experienced as Punk or Brettus but I know a +20 LTFT is jacked.
Perhaps a vacuum leak?
What are your mods again?
Last edited by EviLStewie; 07-21-2013 at 09:32 PM.
07-21-2013, 09:30 PM
#33
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
Only ones that would affect this are the Greddy Catback and Mazsport mid pipe.
Maybe re flash to stock and do logs?
Maybe re flash to stock and do logs?
Last edited by Wolfe; 07-21-2013 at 09:33 PM.
07-21-2013, 09:47 PM
#38
A LTFT that high is definitely a symptom of something pretty major going on . This is mirrored with a high LTFT at cruise as well . As Evil suggested , this could be a vacuum leak .
The fact that AFRs are staying so close to 14.7 could also be symptomatic of a faulty AFR sensor . Do AFRs go rich at full throttle ?
The fact that AFRs are staying so close to 14.7 could also be symptomatic of a faulty AFR sensor . Do AFRs go rich at full throttle ?
07-21-2013, 10:12 PM
#40
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
Also, the only CEL is the one from the De-Cat.
I will also give the MAF a cleaning.
I will also give the MAF a cleaning.
Last edited by Wolfe; 07-21-2013 at 10:14 PM.
07-21-2013, 10:34 PM
#41
Disable the CEL for the missing cat. It's under Edit/Advanced parameters.
Those logs really look like you have an air leak somewhere. Notice how the LTFT goes much lower as load rises? That is because if you have a leak at a constant rate, the airflow from that leak becomes a lower percentage of total airflow as total airflow rises.
Failing MAF sensors typically read high/low across the board.
Those logs really look like you have an air leak somewhere. Notice how the LTFT goes much lower as load rises? That is because if you have a leak at a constant rate, the airflow from that leak becomes a lower percentage of total airflow as total airflow rises.
Failing MAF sensors typically read high/low across the board.
07-21-2013, 11:07 PM
#42
Those tables are really fucked up from the factory, and I don't understand why.
Those tables are a direct reading of target lambda.
The ECU does some math with airflow and VE. This tells it exactly how many grams of air per cycle are in the combustion chamber. It knows injector size. From there it's a simple calculation for how much fuel the ECU needs to inject to hit that target lambda.
(This is also why I work in lambda instead of AFR. The ECU is working in Lambda, and it saves you conversions)
The following numbers are completely made up and wildly unrealistic, but they are used for an easy example (and I'm too tired right now to math properly for real numbers)
Let's say you have a target lambda of 1.
Let's say the ECU determines you have exactly 14.7 grams of air IN THE COMBUSTION CHAMBER (after doing the MAF vs VE calcs). To hit that lambda of 1. The ECU will have to inject exactly 1 gram of fuel. The ECU knows the volumetric flow of the injectors. The ECU also knows the fuel density (unfortunately, this isn't a table that we have access to in order to change in the Mazda ECU's. This would make my life much easier on alcohol fuels.). It's a simple calculation for the ECU to determine an injector duty cycle in order to inject that exact 1 gram of fuel.
Now. Let's say we give the ECU a target of .85 lambda. The ECU does the math required on the fly to increase the IDC to hit that fuel target.
You want to re-shape your target tables TO ACTUALLY BE THE LAMBDA VALUES YOU WANT TO HIT. If your VE table is correct, you tell the ECU to hit a lambda of .85 at say... a calc load of .81 at 8k RPM's, logging with a good aftermarket wideband will show (surprise surprise) a lambda value of .85. (the factory wideband isn't super accurate under high gasflow conditions).
Now, this also explains why manipulating the target tables sort of works OK. If the VE table is wrong, and you treat the target tables as an arbitrary value instead of a real lambda value, you can force the ECU to hit the lambda value you want by forcing the targets super rich/lean.
Mazda made the target tables rich as **** above certain load values to keep morons from popping their motors when they bolt **** on ***** nilly without a tune.
Proper VE tables become VERY important once you start throwing big variations in atmospheric density at the ECU. If VE is correct, it will be able to compensate for huge altitude/temperature swings. If the VE is incorrect, and you are manipulating target tables, or MAF values, all of the assumptions the ECU begins to make for air density go right out the window.
To sumarize:
You want your MAF scale to actually reflect reality. A given g/sec in the MAF scale should 100% actually equal that MAF voltage compared to a real world airflow amount (this is why I like flow benches).
(This is also why anybody not running a blow through MAF on a boosted car is a dummy. Are you telling me that if a draw through MAF meters air at ambient temperature, then that air gets heated a few hundred degrees by a turbo that it won't have a massive change in air density actually hitting the combustion chamber? With a blow through MAF, you are actually metering those changes in air density after the turbo and after the intercooler. Funny thing how you would want to meter the air that is actually going to be used.)
You want your injector scaling values to actually reflect reality (unless you are dealing with alcohol fuels, that is a topic for another day).
You want your VE table to actually reflect reality.
If you have those three things (MAF cal, injector scale, and VE) dead bang accurate, the ECU actually knows how much air is hitting the combustion chamber (due to intake losses from our VE table). It's a simple calc to figure out exactly how much IDC it needs to run in order to throw a given amount of fuel at that air.
You want your target tables to actually be the lambda values you want to see.
If these things are all done correctly, all of those fancy ideal gas law calculations that the ECU does behind the scenes actually matter. This allows for huge swings in ambient air conditions without screwing up fuel delivery.
It also does amazing things for transient response. When you snap the throttle open, the ECU can accurately predict what the fueling needs to be. Remember, you always have a bit of latency between the MAF and the combustion chamber. It takes a bit of time for the air to move between those two points.
Yes, you can absolutely fudge the results by only manipulating the MAF cal, or manipulating target tables. Thing is, as soon as something changes, whether it be air density, or VE (because you are running a different area of the MAP than when you were fudging sensor tables) things go out the window. It also does weird things with transient response, because the predictions being made aren't accurate.
Once this is all done correctly, tuning for power becomes stupid easy. You want the engine to run a half point richer? Change your target table by a half point, sit back, watch the magic happen and see if the fueling change had an impact on power.
Those tables are a direct reading of target lambda.
The ECU does some math with airflow and VE. This tells it exactly how many grams of air per cycle are in the combustion chamber. It knows injector size. From there it's a simple calculation for how much fuel the ECU needs to inject to hit that target lambda.
(This is also why I work in lambda instead of AFR. The ECU is working in Lambda, and it saves you conversions)
The following numbers are completely made up and wildly unrealistic, but they are used for an easy example (and I'm too tired right now to math properly for real numbers)
Let's say you have a target lambda of 1.
Let's say the ECU determines you have exactly 14.7 grams of air IN THE COMBUSTION CHAMBER (after doing the MAF vs VE calcs). To hit that lambda of 1. The ECU will have to inject exactly 1 gram of fuel. The ECU knows the volumetric flow of the injectors. The ECU also knows the fuel density (unfortunately, this isn't a table that we have access to in order to change in the Mazda ECU's. This would make my life much easier on alcohol fuels.). It's a simple calculation for the ECU to determine an injector duty cycle in order to inject that exact 1 gram of fuel.
Now. Let's say we give the ECU a target of .85 lambda. The ECU does the math required on the fly to increase the IDC to hit that fuel target.
You want to re-shape your target tables TO ACTUALLY BE THE LAMBDA VALUES YOU WANT TO HIT. If your VE table is correct, you tell the ECU to hit a lambda of .85 at say... a calc load of .81 at 8k RPM's, logging with a good aftermarket wideband will show (surprise surprise) a lambda value of .85. (the factory wideband isn't super accurate under high gasflow conditions).
Now, this also explains why manipulating the target tables sort of works OK. If the VE table is wrong, and you treat the target tables as an arbitrary value instead of a real lambda value, you can force the ECU to hit the lambda value you want by forcing the targets super rich/lean.
Mazda made the target tables rich as **** above certain load values to keep morons from popping their motors when they bolt **** on ***** nilly without a tune.
Proper VE tables become VERY important once you start throwing big variations in atmospheric density at the ECU. If VE is correct, it will be able to compensate for huge altitude/temperature swings. If the VE is incorrect, and you are manipulating target tables, or MAF values, all of the assumptions the ECU begins to make for air density go right out the window.
To sumarize:
You want your MAF scale to actually reflect reality. A given g/sec in the MAF scale should 100% actually equal that MAF voltage compared to a real world airflow amount (this is why I like flow benches).
(This is also why anybody not running a blow through MAF on a boosted car is a dummy. Are you telling me that if a draw through MAF meters air at ambient temperature, then that air gets heated a few hundred degrees by a turbo that it won't have a massive change in air density actually hitting the combustion chamber? With a blow through MAF, you are actually metering those changes in air density after the turbo and after the intercooler. Funny thing how you would want to meter the air that is actually going to be used.)
You want your injector scaling values to actually reflect reality (unless you are dealing with alcohol fuels, that is a topic for another day).
You want your VE table to actually reflect reality.
If you have those three things (MAF cal, injector scale, and VE) dead bang accurate, the ECU actually knows how much air is hitting the combustion chamber (due to intake losses from our VE table). It's a simple calc to figure out exactly how much IDC it needs to run in order to throw a given amount of fuel at that air.
You want your target tables to actually be the lambda values you want to see.
If these things are all done correctly, all of those fancy ideal gas law calculations that the ECU does behind the scenes actually matter. This allows for huge swings in ambient air conditions without screwing up fuel delivery.
It also does amazing things for transient response. When you snap the throttle open, the ECU can accurately predict what the fueling needs to be. Remember, you always have a bit of latency between the MAF and the combustion chamber. It takes a bit of time for the air to move between those two points.
Yes, you can absolutely fudge the results by only manipulating the MAF cal, or manipulating target tables. Thing is, as soon as something changes, whether it be air density, or VE (because you are running a different area of the MAP than when you were fudging sensor tables) things go out the window. It also does weird things with transient response, because the predictions being made aren't accurate.
Once this is all done correctly, tuning for power becomes stupid easy. You want the engine to run a half point richer? Change your target table by a half point, sit back, watch the magic happen and see if the fueling change had an impact on power.
Last edited by no-coast-punk; 07-21-2013 at 11:11 PM.
07-22-2013, 12:24 AM
#43
I have tuned lots of turbo RX8s and have not tried to adjust Ve tables to any great degree - settling for values that are close to where i believe actual Ve is . Only play with them if it doesn't make sense do anything else . This yeilds AFRs very close to target suggesting that VE must also be close . My logic was that VE doesn't suddenly change (except when a valve like the APV opens) so why try and chase it .
Had pretty good results from doing it this way but can see your way could be superior .
Will give your method a try and see if it improves anything.
As far as Blow through vs draw through .... I run draw through but with the IAT sensor post turbo/IC . This seems to work well as i don't see any fluctuations in AFR .
Last edited by Brettus; 07-22-2013 at 12:47 AM.
07-22-2013, 03:29 AM
#45
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
Alright, I will try to get rid of that CEL Code, and Also change the temp my fans come on. I have an oil catch can on the way, and will check out all the hoses and what not thoroughly when I install it. I'm also going to re flash back to stock and do the same logs. If there is no change in LTFT, I would assume that rules out the Stage 1 tune being at fault and means a mechanical problem. At least the car does a good job keeping the AFR around 14.7 and I should be safe from being too lean. Not great that it has to do that much compensation, but at least I know now, and can do something about it. I will post the stock logs and any finds soon. Any advice with how to proceed is appreciated. Thanks all for the replies and help. Looking forward to learning more from you all!
07-22-2013, 04:40 AM
#46
You still get better response from moving the MAF itself. By just moving the IAT, you induce a weird latency into things. Your flow and temperature references aren't measuring the same air due to the several feet of piping in between. There have also been pumping losses introduced after the MAF due to your plumbing, IC and the turbo itself. This does weird things with VE (We don't have a table for compressor efficiency. Only ze Germans get that lucky). Not only are you measuring two separate colums of air, one of those is way delayed from throttle opening. Things can get a little weird during transient conditions.
With a pure blow through setup, you are back to just measuring air as it enters the engine.
I'm kind of a psychotic about transient response. My background is with rally/road racing cars. A throttle hiccup on a rally car can kill people.
Last edited by no-coast-punk; 07-22-2013 at 04:43 AM.
07-22-2013, 01:31 PM
#48
I've done lots of (hundreds of cars) AP tuning on non-Mazda platforms. Nothing is that different, outside of some of the ignition timing foibles due to the rotaryness. I'm happy to share everything I know.
The big remote tuner in this community seems to be MM. That guy was a complete **** head to me when I was ordering my AP. If I can take away a bit of his business, I'm more than happy to do it.
I got out of the tuning game long ago. I'm not here to make money.
First things first.
Only log the bare minimum. The serial bus only has so much bandwidth. The fewer variables logged, the faster your update rates.
At this point, all you are really concerned with is your STFT (Short Term Fuel Trim), RPM, Calc Load and Lambda.
If you have a stock intake (which you really should) leave the MAF cal alone. Mazda spent a **** ton of time developing that particular map.
The thing you want to start manipulating is your "Fuel VE" table.
You are looking at your STFT at a given calc. load and RPM.
If you look at your logs you will see a spot at ~4600 RPM (from your log) with a calc load of ~65. Look at your STFT. It's currently at 10.
The basics of what your ECU is doing is this:
It is looking at sensor inputs to see how much air is flowing into the engine. Estimating what the engines volumetric efficiency is to guess how much of that air is actually entering the combustion chamber. Injecting the amount of fuel needed to reach a given AFR specified in the target tables. After the fact, comparing what the O2 sensor is seeing as a result of the combustion.
Your STFT of 10% means that O2 sensor is seeing a mixture that is 10% leaner than the target AFR. It will add 10% more fuel to try and reach that AFR the next time around.
Since this is all happening after the fact, and the ECU's predictions are 10% off, it means that fueling is never consistent. This leads to weird throttle hiccups. It's like trying to keep your car between the lanes only using the rear view mirror. The important stuff already happened.
So what you want to do is to increase the value that particular load cell (closest you can find to 4500 RPM and 65 calc load) in the VE table by 10%. This is telling the ECU that the engine is 10% more efficient at inhaling air in that load cell. This means the ECU will inject about 10% more fuel.
When you are changing cells in the VE table, never change a cell in isolation unless it's a very small change. You want to blend any changes in with the surrounding cells.
So if you make your 10% change @ 4500 and 65, make a 5% change in each cell that touches it. Airflow doesn't make radical changes from one cell to another. Your mapping should reflect this.
Get your closed loop VE stuff as close as possible. This is where fuel economy and responsiveness comes from.
Once you get that dialed in, we'll move onto open loop stuff.
The big remote tuner in this community seems to be MM. That guy was a complete **** head to me when I was ordering my AP. If I can take away a bit of his business, I'm more than happy to do it.
I got out of the tuning game long ago. I'm not here to make money.
First things first.
Only log the bare minimum. The serial bus only has so much bandwidth. The fewer variables logged, the faster your update rates.
At this point, all you are really concerned with is your STFT (Short Term Fuel Trim), RPM, Calc Load and Lambda.
If you have a stock intake (which you really should) leave the MAF cal alone. Mazda spent a **** ton of time developing that particular map.
The thing you want to start manipulating is your "Fuel VE" table.
You are looking at your STFT at a given calc. load and RPM.
If you look at your logs you will see a spot at ~4600 RPM (from your log) with a calc load of ~65. Look at your STFT. It's currently at 10.
The basics of what your ECU is doing is this:
It is looking at sensor inputs to see how much air is flowing into the engine. Estimating what the engines volumetric efficiency is to guess how much of that air is actually entering the combustion chamber. Injecting the amount of fuel needed to reach a given AFR specified in the target tables. After the fact, comparing what the O2 sensor is seeing as a result of the combustion.
Your STFT of 10% means that O2 sensor is seeing a mixture that is 10% leaner than the target AFR. It will add 10% more fuel to try and reach that AFR the next time around.
Since this is all happening after the fact, and the ECU's predictions are 10% off, it means that fueling is never consistent. This leads to weird throttle hiccups. It's like trying to keep your car between the lanes only using the rear view mirror. The important stuff already happened.
So what you want to do is to increase the value that particular load cell (closest you can find to 4500 RPM and 65 calc load) in the VE table by 10%. This is telling the ECU that the engine is 10% more efficient at inhaling air in that load cell. This means the ECU will inject about 10% more fuel.
When you are changing cells in the VE table, never change a cell in isolation unless it's a very small change. You want to blend any changes in with the surrounding cells.
So if you make your 10% change @ 4500 and 65, make a 5% change in each cell that touches it. Airflow doesn't make radical changes from one cell to another. Your mapping should reflect this.
Get your closed loop VE stuff as close as possible. This is where fuel economy and responsiveness comes from.
Once you get that dialed in, we'll move onto open loop stuff.
RPM
4089
4096
4088
4088
4094
4094
4108
4092
4085
4105
4110
4107
4086
4100
4109
4106
4085
4104
4105
4104
4100
4099
4087
4078
4081
4101
4101
4086
4099
4098
4075
4074
4097
4093
4097
4080
4093
4075
Short Term FT (%)
6.08
6.08
6.08
5.3
5.3
5.3
5.3
5.3
5.3
5.3
5.3
5.3
5.3
6.08
6.08
6.08
6.08
6.08
6.08
6.08
6.86
6.86
6.86
6.08
6.08
6.08
6.08
6.08
6.08
6.86
6.86
6.86
6.08
5.3
4.52
4.52
5.3
5.3
Long Term FT (%)
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
0.62
Calculated Load (%)
26.26
25.48
24.3
24.7
25.09
25.09
25.09
24.7
25.09
24.7
23.91
23.91
23.91
23.91
23.52
23.91
23.91
23.91
23.91
23.91
23.91
24.3
24.3
24.3
24.7
24.7
25.09
25.87
27.05
27.44
26.66
25.87
25.48
25.87
26.26
26.66
26.66
27.05
If I take the first line of this log and the logged items.. i have.
RPM=4089
Cal. Load = 26.26
STFT=6.08
LTFT=.62
My VE table looks like this..
RPM=4000 and Cal. Load at .25 = 0.99 + 6% = 1.0494 is what should be in the cell? and + the adjacent sells 3%???
is that what that cell now be? and I can still blend the adjacent sells with half of 6%?? sound correct?
oh BTW I have a AEM CAI and my Maf has been scaled somewhat... But im just trying to compare the two..
Last edited by WreakLoosE; 07-22-2013 at 01:49 PM.
07-22-2013, 01:58 PM
#49
SARX Legend
iTrader: (46)
Wreakloose, you need to get your STFT to within 3%, the AEM MAF is the same as stock IIRC, are you sure you have no vacuum leaks?
no-coast, great posts.
I have been messing with this for a while now and the differences in understanding an tuning philosophies has always threw me off in terms of my tuning and understanding things on my turbocharged RX-8. Your comments on a blow thru setup make sense, I just think it is something that is tough for most to tune as it is a pretty difficult thing to do.
I was having many tuning issues (bogging rich mainly and causing spark blow out) that were difficult to get sorted and had me looking at mechanical problems outside of the tune. Until I visited Steve Kan and then he confused me even more with his philosophy but then was able to get it tuned really well(despite not being able to hit decent boost due to a shitty MBC).
I plan on replacing my wastegate actuator here soon (its very old and boost is not holding at high RPM) and retuning again soon.
no-coast, great posts.
I have been messing with this for a while now and the differences in understanding an tuning philosophies has always threw me off in terms of my tuning and understanding things on my turbocharged RX-8. Your comments on a blow thru setup make sense, I just think it is something that is tough for most to tune as it is a pretty difficult thing to do.
I was having many tuning issues (bogging rich mainly and causing spark blow out) that were difficult to get sorted and had me looking at mechanical problems outside of the tune. Until I visited Steve Kan and then he confused me even more with his philosophy but then was able to get it tuned really well(despite not being able to hit decent boost due to a shitty MBC).
I plan on replacing my wastegate actuator here soon (its very old and boost is not holding at high RPM) and retuning again soon.
Last edited by 9krpmrx8; 07-22-2013 at 02:19 PM.
