So, you want to be a tuner? Look no further.
06-28-2016, 04:28 PM
#353
08-12-2016, 01:02 PM
#354
To keep the knowledge flowing in this thread...Does anyone have thoughts on aggressive NA ignition timing, and particularly split?
In the videos, Kane mentions to not go below 10 deg split, under load. Is this just a safe starting point, or the recommended min split, period?
I ask because the MM ignition split calc spreadsheet has default values as low as 5 deg split at 0.88 calc load and high rpm.
Since I'm at higher altitude (3000 ft) 0.88 calc load is the highest column I am seeing. Would you bottom dwellers consider that high load?
In the videos, Kane mentions to not go below 10 deg split, under load. Is this just a safe starting point, or the recommended min split, period?
I ask because the MM ignition split calc spreadsheet has default values as low as 5 deg split at 0.88 calc load and high rpm.
Since I'm at higher altitude (3000 ft) 0.88 calc load is the highest column I am seeing. Would you bottom dwellers consider that high load?
08-15-2016, 03:50 PM
#356
I found 2 maps on JDM cars that had been reflashed in Japan by some tuning company over there . They had 32degrees of timing at high rpm and a 2 degree split . We put one of them on the dyno and compared it to stock . The low split map made way less power than stock from mid range right through to peak.
Last edited by Brettus; 08-15-2016 at 03:55 PM.
08-16-2016, 11:00 AM
#357
Thanks of the info, guys!
Brettus, those results from the JDM flash are interesting, but that makes sense if I'm thinking about it correctly. The trailing ignition is probably firing so early that some of the pressure developed by the trailing ignition is pushing backward on the rotor, negating some of the power developed by the leading ignition. Or, it's at least pushing on the rotor at some point where the mechanical advantage isn't optimal...?
Kane, First off...thanks a ton for your videos! I've tuned a few boosted MAP-based piston engines in the past, but your videos totally saved me from chasing my tail, in regard to MAF and injector scaling, and gave me a little more confidence in the presence of apex seals!
In one of the videos you show that it's a good idea to advance ignition timing by 1deg/rpm row at 6000rpm and up, in higher loads. Then, you mention that NA guys can get more aggressive than this, but I didn't catch if you had a definite suggestion, since the audio cuts in and out. If it's not a trade secret, would you care to comment? I've had trouble finding any good examples of timing maps for NA engines.
Following that trend, I'd be at 36* advance at 9000rpm and 0.88 calc load and above.
I'm OK with some risk to the engine, but I don't want to be so on the ragged edge that failure is guaranteed. lol
Brettus, those results from the JDM flash are interesting, but that makes sense if I'm thinking about it correctly. The trailing ignition is probably firing so early that some of the pressure developed by the trailing ignition is pushing backward on the rotor, negating some of the power developed by the leading ignition. Or, it's at least pushing on the rotor at some point where the mechanical advantage isn't optimal...?
Kane, First off...thanks a ton for your videos! I've tuned a few boosted MAP-based piston engines in the past, but your videos totally saved me from chasing my tail, in regard to MAF and injector scaling, and gave me a little more confidence in the presence of apex seals!
In one of the videos you show that it's a good idea to advance ignition timing by 1deg/rpm row at 6000rpm and up, in higher loads. Then, you mention that NA guys can get more aggressive than this, but I didn't catch if you had a definite suggestion, since the audio cuts in and out. If it's not a trade secret, would you care to comment? I've had trouble finding any good examples of timing maps for NA engines.
Following that trend, I'd be at 36* advance at 9000rpm and 0.88 calc load and above.I'm OK with some risk to the engine, but I don't want to be so on the ragged edge that failure is guaranteed. lol
Last edited by ion_four; 08-16-2016 at 11:04 AM.
08-17-2016, 06:21 PM
#359
Hybrid Greddy Boosted
I've also seen this lean spot. It only appears at 10-40% load and 1500-2500rpm. It's not really an issue as the load is not high enough to cause noticeable engine damage, it doesn't impact LTFTs or drive-ability, and it likely occurs on all stock tuned rx8s driving around on the roads today, which means Mazda haven't bothered to tune it out of their factory tune.
I believe it's caused by the gradual cutover of fueling from P1 injectors cutting to the next bank (Secondaries I think, can't remember). If you don't like the behavior you can tweak this area of your VE table (increase VE value by 10% richens the mix so effectively reduces fuel trim by 10 in that area). You'll need to make sure your changes don't change your LTFTs if you actually want this change to be an improvement. Probably not worth the effort TBH.
I believe it's caused by the gradual cutover of fueling from P1 injectors cutting to the next bank (Secondaries I think, can't remember). If you don't like the behavior you can tweak this area of your VE table (increase VE value by 10% richens the mix so effectively reduces fuel trim by 10 in that area). You'll need to make sure your changes don't change your LTFTs if you actually want this change to be an improvement. Probably not worth the effort TBH.
08-18-2016, 09:07 AM
#360
There is a still a ton I have to learn, but regarding spikes, one of the things I learned to do was look at how long they last. I have no idea how an AP decides update rates, but I found some of the spikes I was chasing were less than 0.1 second wide, and when I did the math, were for less than an engine rotation. Those are the ones I decided to worry less about.
08-19-2016, 09:11 AM
#361
Non-Savant Idiot
I've also seen this lean spot. It only appears at 10-40% load and 1500-2500rpm. It's not really an issue as the load is not high enough to cause noticeable engine damage, it doesn't impact LTFTs or drive-ability, and it likely occurs on all stock tuned rx8s driving around on the roads today, which means Mazda haven't bothered to tune it out of their factory tune.
I believe it's caused by the gradual cutover of fueling from P1 injectors cutting to the next bank (Secondaries I think, can't remember). If you don't like the behavior you can tweak this area of your VE table (increase VE value by 10% richens the mix so effectively reduces fuel trim by 10 in that area). You'll need to make sure your changes don't change your LTFTs if you actually want this change to be an improvement. Probably not worth the effort TBH.
I believe it's caused by the gradual cutover of fueling from P1 injectors cutting to the next bank (Secondaries I think, can't remember). If you don't like the behavior you can tweak this area of your VE table (increase VE value by 10% richens the mix so effectively reduces fuel trim by 10 in that area). You'll need to make sure your changes don't change your LTFTs if you actually want this change to be an improvement. Probably not worth the effort TBH.
08-24-2016, 03:16 PM
#362
I've also seen this lean spot. It only appears at 10-40% load and 1500-2500rpm. It's not really an issue as the load is not high enough to cause noticeable engine damage, it doesn't impact LTFTs or drive-ability, and it likely occurs on all stock tuned rx8s driving around on the roads today, which means Mazda haven't bothered to tune it out of their factory tune.
I believe it's caused by the gradual cutover of fueling from P1 injectors cutting to the next bank (Secondaries I think, can't remember). If you don't like the behavior you can tweak this area of your VE table (increase VE value by 10% richens the mix so effectively reduces fuel trim by 10 in that area). You'll need to make sure your changes don't change your LTFTs if you actually want this change to be an improvement. Probably not worth the effort TBH.
I believe it's caused by the gradual cutover of fueling from P1 injectors cutting to the next bank (Secondaries I think, can't remember). If you don't like the behavior you can tweak this area of your VE table (increase VE value by 10% richens the mix so effectively reduces fuel trim by 10 in that area). You'll need to make sure your changes don't change your LTFTs if you actually want this change to be an improvement. Probably not worth the effort TBH.
It's possible there is some inaccuracy in the VE map, but common knowledge on here says you shouldn't have to mess with that on a stock internal engine. If you can't feel it at low rpm/low load, I wouldn't mess with with. I agree it's not worth it
This reminds me...I recall reading that the S-DAIS tables don't do anything, but I raised the VDI opening point to 9500rpm just as a test and according to the virtual dyno, my top end seems to have dropped off. That being said, I find that comparing pulls in that software is pretty much like apples and oranges, since I have to change vehicle weights quite a bit to get values in line with "real" dyno values and to line up with successive pulls. Thoughts on the S-DAIS tables, or virtual dyno?
08-27-2016, 10:05 AM
#363
Self-reliance is OK when you're working on a model train. When you're playing with a supercharged RX-8, maybe not so much.
Utter n00b here. Cobb AP owner, have functional(ish) installs of AccessManager and AccessTuner Race. I've got issues with the AP not logging (says so right there on the faceplate, "Not Logging"), but have managed to get a laptop synced with the Cobb's OBD2 to get live logs. I also have Torque Pro with PLX's Kiwi3, but logging on that combo is a whole 'nother ball of wax.
The car runs, can be carefully driven, no sudden loud pedal stabs, and has produced a LTFT of -10 over ten miles of cautious cruising and driveway idling. The intake tract is utterly different, so a MAF baseline is mandatory. The "how" is what's got me at a standstill: I see equal arguments to scale the MAF voltage to null the LTFT, or adjust the Inj1 fueling downward(?). The second seems counterproductive.
The -10 LTFT made me sweep the install for intake leaks, and didn't find anything using the carb cleaner trick. Car idles perfectly; adjusted it upward to 950 to give it a little recovery room with the SC drag.
tl:dr; Where to start scaling MAF changes on a new SC install?
Utter n00b here. Cobb AP owner, have functional(ish) installs of AccessManager and AccessTuner Race. I've got issues with the AP not logging (says so right there on the faceplate, "Not Logging"), but have managed to get a laptop synced with the Cobb's OBD2 to get live logs. I also have Torque Pro with PLX's Kiwi3, but logging on that combo is a whole 'nother ball of wax.
The car runs, can be carefully driven, no sudden loud pedal stabs, and has produced a LTFT of -10 over ten miles of cautious cruising and driveway idling. The intake tract is utterly different, so a MAF baseline is mandatory. The "how" is what's got me at a standstill: I see equal arguments to scale the MAF voltage to null the LTFT, or adjust the Inj1 fueling downward(?). The second seems counterproductive.
The -10 LTFT made me sweep the install for intake leaks, and didn't find anything using the carb cleaner trick. Car idles perfectly; adjusted it upward to 950 to give it a little recovery room with the SC drag.
tl:dr; Where to start scaling MAF changes on a new SC install?
08-27-2016, 10:39 AM
#364
Self-reliance is OK when you're working on a model train. When you're playing with a supercharged RX-8, maybe not so much.
Utter n00b here. Cobb AP owner, have functional(ish) installs of AccessManager and AccessTuner Race. I've got issues with the AP not logging (says so right there on the faceplate, "Not Logging"), but have managed to get a laptop synced with the Cobb's OBD2 to get live logs. I also have Torque Pro with PLX's Kiwi3, but logging on that combo is a whole 'nother ball of wax.
The car runs, can be carefully driven, no sudden loud pedal stabs, and has produced a LTFT of -10 over ten miles of cautious cruising and driveway idling. The intake tract is utterly different, so a MAF baseline is mandatory. The "how" is what's got me at a standstill: I see equal arguments to scale the MAF voltage to null the LTFT, or adjust the Inj1 fueling downward(?). The second seems counterproductive.
The -10 LTFT made me sweep the install for intake leaks, and didn't find anything using the carb cleaner trick. Car idles perfectly; adjusted it upward to 950 to give it a little recovery room with the SC drag.
tl:dr; Where to start scaling MAF changes on a new SC install?
Utter n00b here. Cobb AP owner, have functional(ish) installs of AccessManager and AccessTuner Race. I've got issues with the AP not logging (says so right there on the faceplate, "Not Logging"), but have managed to get a laptop synced with the Cobb's OBD2 to get live logs. I also have Torque Pro with PLX's Kiwi3, but logging on that combo is a whole 'nother ball of wax.
The car runs, can be carefully driven, no sudden loud pedal stabs, and has produced a LTFT of -10 over ten miles of cautious cruising and driveway idling. The intake tract is utterly different, so a MAF baseline is mandatory. The "how" is what's got me at a standstill: I see equal arguments to scale the MAF voltage to null the LTFT, or adjust the Inj1 fueling downward(?). The second seems counterproductive.
The -10 LTFT made me sweep the install for intake leaks, and didn't find anything using the carb cleaner trick. Car idles perfectly; adjusted it upward to 950 to give it a little recovery room with the SC drag.
tl:dr; Where to start scaling MAF changes on a new SC install?
To then view your logs, you have to connect via PC using AP Manager, go to Advanced View and download the logs. Then, go into Access Tuner Race and under the "View" menus, select "Data Log Viewer" and find them on your PC. Not exactly intuitive. I don't think there is a way to view logged data on the AP itself, but I could be wrong.
My quick suggestions for getting started, most of which are regurgitated: Scale MAF to ~5-5.5g/s at warm idle if you're near sea level. Adjust according to altitude (I'm at 2800ft and am usually in the high 4.x range). If your MAF is reading in this range, you can probably move on to scaling the Injector Bank 1 if your LTFT is off at low load/rpm.
The happy problem you have is that your VE has completely changed with the supercharger and your calculated load values will be going much higher. So, you have to change the column labels to extend into boost. I have not done this on an RX-8, so someone else probably has much better advice, but I would extend the tables well beyond what you think you might need and make them horribly rich to start with, since I would only be guessing at how much fuel is really needed.
Do you have different fuel injectors? I have basically no clue how much headroom the stock setup has.
08-27-2016, 02:56 PM
#365
As far as logging, I also found this initially confusing. First, select all the values you want to log within the AP (logging fewer parameters yields a higher sampling rate, btw). When you are at a screen showing some live data value like RPM and "Not Logging", hit the round select button. After a couple seconds it will change to "Logging". When you are done, hit the button again and the log will be saved to its own file. Repeating will create additional files. If you're doing this, I dunno.
THANK YOU! Now, we're getting somewhere!
My quick suggestions for getting started, most of which are regurgitated: Scale MAF to ~5-5.5g/s at warm idle if you're near sea level. Adjust according to altitude (I'm at 2800ft and am usually in the high 4.x range). If your MAF is reading in this range, you can probably move on to scaling the Injector Bank 1 if your LTFT is off at low load/rpm.
The happy problem you have is that your VE has completely changed with the supercharger and your calculated load values will be going much higher. So, you have to change the column labels to extend into boost. I have not done this on an RX-8, so someone else probably has much better advice, but I would extend the tables well beyond what you think you might need and make them horribly rich to start with, since I would only be guessing at how much fuel is really needed.
The happy problem you have is that your VE has completely changed with the supercharger and your calculated load values will be going much higher. So, you have to change the column labels to extend into boost. I have not done this on an RX-8, so someone else probably has much better advice, but I would extend the tables well beyond what you think you might need and make them horribly rich to start with, since I would only be guessing at how much fuel is really needed.
MAF was logged as 5.65g/sec @ 910rpm, 105F inlet temp, and I'm so close to sea level the correction factor is below negligible. Log now shows -12.6 LTFT at idle, tapers off to -3 @ 2,000rpm, to -.9 (apparently closed loop?) around 3,400. AFR stays at or above 14.7 all the way up with slight transients to 13.5, then ~20 on throttle cut (expected).
As for throttle, load and VE table setup, it's back to Kane's videos to eak out knowledge.
Thank You for helping me get off the dime on this.
08-29-2016, 10:51 AM
#366
No prob to help you get started.
It seems your MAF is on target. I would move on to experiment with increasing the Injector Bank 1 size by 5-10% and see if that makes a desirable change in your LTFT, but you may very well just have a significant VE change due to the supercharger, even out of boost. Scaling your injectors 5-10% and staying out of high load situations won't cause anything catastrophic to happen. So, just experiment a bit. It's time consuming, but informative. Small changes won't break anything as long as you are coming from a safe starting point.
I'm assuming you have extended your tables into boost loads. That should be the first move if the belt is on the supercharger (i.e. you could make boost). I would go one column beyond what you think you will reach and make that dump fuel as a safety measure incase your calculations are off. It doesn't even have to be at max boost, if you're going to keep your foot out of it. I would also consider some timing retard in boost, and not just level off the OEM. I'm sure someone with direct experience with a supercharged Renesis could speak better to this. It might even be in the Kane videos, but I didn't pay attention to that stuff
It seems your MAF is on target. I would move on to experiment with increasing the Injector Bank 1 size by 5-10% and see if that makes a desirable change in your LTFT, but you may very well just have a significant VE change due to the supercharger, even out of boost. Scaling your injectors 5-10% and staying out of high load situations won't cause anything catastrophic to happen. So, just experiment a bit. It's time consuming, but informative. Small changes won't break anything as long as you are coming from a safe starting point.
I'm assuming you have extended your tables into boost loads. That should be the first move if the belt is on the supercharger (i.e. you could make boost). I would go one column beyond what you think you will reach and make that dump fuel as a safety measure incase your calculations are off. It doesn't even have to be at max boost, if you're going to keep your foot out of it. I would also consider some timing retard in boost, and not just level off the OEM. I'm sure someone with direct experience with a supercharged Renesis could speak better to this. It might even be in the Kane videos, but I didn't pay attention to that stuff
The following users liked this post:
Striker-7 (09-06-2017)
09-15-2016, 10:01 AM
#368
Hey guys, I'm working on a little side project right now, and I was wondering how the ECU determines which value to use in the AFR tables, or actually just in general really. Given that an RPM and Calc Load can be values between the row/columns, does the ECU shoot for a middle value? Or does it simply choose the closest value? Or does it err on the side of caution, using the higher calc load or RPM?
09-16-2016, 11:07 AM
#369
No prob to help you get started.
It seems your MAF is on target. I would move on to experiment with increasing the Injector Bank 1 size by 5-10% and see if that makes a desirable change in your LTFT, but you may very well just have a significant VE change due to the supercharger, even out of boost. Scaling your injectors 5-10% and staying out of high load situations won't cause anything catastrophic to happen. So, just experiment a bit. It's time consuming, but informative. Small changes won't break anything as long as you are coming from a safe starting point.
It seems your MAF is on target. I would move on to experiment with increasing the Injector Bank 1 size by 5-10% and see if that makes a desirable change in your LTFT, but you may very well just have a significant VE change due to the supercharger, even out of boost. Scaling your injectors 5-10% and staying out of high load situations won't cause anything catastrophic to happen. So, just experiment a bit. It's time consuming, but informative. Small changes won't break anything as long as you are coming from a safe starting point.
Scaling the primary injector has settled the LTFT / STFT into a +/- 3% range. I now have a vacuum / boost gauge plumbed in, so I can at least visually tell when this beast tries to go positive manifold pressure. It has flickered to +4 on a short acceleration run, otherwise I'm seeing steady vacuum during the 30-second data runs for injector scaling.
I'm assuming you have extended your tables into boost loads. That should be the first move if the belt is on the supercharger (i.e. you could make boost). I would go one column beyond what you think you will reach and make that dump fuel as a safety measure incase your calculations are off. It doesn't even have to be at max boost, if you're going to keep your foot out of it. I would also consider some timing retard in boost, and not just level off the OEM. I'm sure someone with direct experience with a supercharged Renesis could speak better to this. It might even be in the Kane videos, but I didn't pay attention to that stuff
So far, no changes to the fuel / VE tables while I'm doing the injector runs. I've tried finding examples of the suggested FI fuel tables, but running into 'link rot' in the forum. I'm getting a little miffed at that smug kitten (Instagram '404' image).
First pass at the "max load" table. The car is idling / returning to idle like a factory tune, even with the belt on the SC, which is why I left the low end as close to stock as possible. I'll probably have to bump the 1500rpm+ range up a lot as well, but I'll see how it reacts to slow roll-on throttle first.
Last edited by Striker-7; 09-16-2016 at 11:20 AM. Reason: TeamViewer 11 "borked", workaround info
09-16-2016, 06:22 PM
#370
I am the noob tuner, but.......
I have an NA street ported engine that makes it all the way to 118% load around 6,000 RPM.
And, there is at least one modifier in the tables that will cut the max load based on temp. You do NOT want to every outrun the load table. It responds by not believing the MAF, and keeping the last injector pulse width. This drives things lean, and can drive them lean in a hurry. If I had a cool supercharged 8, I would be asking what the realistic max load should be and adding at least 10%.
And, if the smart folks here tell you I am wrong, believe them : )
I have an NA street ported engine that makes it all the way to 118% load around 6,000 RPM.
And, there is at least one modifier in the tables that will cut the max load based on temp. You do NOT want to every outrun the load table. It responds by not believing the MAF, and keeping the last injector pulse width. This drives things lean, and can drive them lean in a hurry. If I had a cool supercharged 8, I would be asking what the realistic max load should be and adding at least 10%.
And, if the smart folks here tell you I am wrong, believe them : )
09-17-2016, 10:42 AM
#371
Can't get much more n00b than this guy... 
Been picking through the threads around this place, and stumbled into Brettus and Slash128's discussions on Calc Max Load here. I've seen graphs where all the 'max load' tables were set to 200%, and trying to wrap my head around the reasons for it.
The pic in my last post is a work in progress, hasn't been committed to the ECU. Between family issues, age and full-time employment, I don't get much time to dig into the forum for info.
Thanks for the warning!
Been picking through the threads around this place, and stumbled into Brettus and Slash128's discussions on Calc Max Load here. I've seen graphs where all the 'max load' tables were set to 200%, and trying to wrap my head around the reasons for it.
The pic in my last post is a work in progress, hasn't been committed to the ECU. Between family issues, age and full-time employment, I don't get much time to dig into the forum for info.
Thanks for the warning!
09-18-2016, 09:42 AM
#372
We found that maxing out the Calc Load table (along with the IAT and Baro tables) has no adverse affect. Brettus theory is that Mazda had a specific max curve to save the cat, which I believe is in that same thread. But no cat = no benefit so no reason to go to the trouble of figuring out the optimal curve for boost vs saving the cat. Just flatline all three at max
09-18-2016, 09:47 AM
#373
Oh and if you come across mysterious stuff about Einsteinia space / edge of the paper / stretching rulers or whatnot don't get sucked in. The only mystery at this point is how the myth perpetuated for so long 
