EvoScan help! Timing + Knock count
Originally Posted by AutoXer
How do you come up with 14*? Looks closer to 11 or 12 to me.
5 KNOCK counts
14* Ignition Timing
Code:
xxxxxxx RPM 27.7 6000 14 58 1384 81 101 0 12 255 27.82 6281 15 100 1604 80 100 5 14 255 27.95 6594 15 100 1604 80 99 4 14 255 28.07 6406 16 100 1604 80 99 12 10 255 28.17 6031 17 100 1604 80 100 12 8 255 28.27 5750 17 100 1604 80 100 11 6 254 28.36 5875 17 100 1604 79 100 11 7 254 28.47 5938 16 100 1604 79 100 10 7 254 28.59 6031 16 100 1604 79 100 10 7 254 28.72 6125 16 100 1604 79 100 9 9 254 28.85 6219 16 100 1604 79 100 8 10 253 28.96 6281 16 100 1604 79 100 7 10 253 29.06 6375 16 100 1604 79 100 7 11 253 29.17 6438 16 100 1604 79 100 7 12 253 29.28 6531 16 100 1604 79 100 7 12 253 29.4 6594 16 100 1604 79 100 7 13 252 29.49 6656 15 100 1604 79 100 7 13 252 29.6 6719 15 100 1604 79 100 7 14 252 29.72 6781 15 100 1604 79 100 7 14 252 29.83 6844 15 100 1604 79 100 6 15 252 29.95 6938 15 100 1604 79 100 6 16 252 30.07 7000 16 100 1604 79 100 6 16 251 30.18 7063 16 100 1604 79 99 6 16 251 30.3 7094 16 100 1604 79 99 6 17 251 30.42 7188 16 100 1604 79 99 6 17 251 30.52 7219 16 100 1604 78 99 6 17 251
Originally Posted by TTP Engineering
6281 RPMS
5 KNOCK counts
14* Ignition Timing
5 KNOCK counts
14* Ignition Timing
I concentrated on this section:
Code:
28.47 5938 16 100 1604 79 100 10 7 254 28.59 6031 16 100 1604 79 100 10 7 254 28.72 6125 16 100 1604 79 100 9 9 254
10 Knock Counts with 7*
9 Knock Counts with 9*
Using the 3 knock counts per degree of timing pull that I've heard and partially verified... I get 10*, 10* and 12*.
Fist and foremost i just wanted to say sorry to Scott for even creating this thread. I had no idea this is what would come of it, and you know i didn't intend for this to happen, scott. Like Malibu said, no other tuner has requested that datalogs be taken down, and i didn't expect it. Again my apologies.
Guys, i could care less about legal rights, and implecations, etc. I care about respect, and Scott asked the logs be removed, for his personal reasons, so who cares; so i did just that, not because i don't know my rights in this forum, but because i respect EVERYONE to a certain degree. It just so happens that Scott has a ton of my respect, and he has all the right in the world to ask anything from me, as many of you do, and that's that. At the point when he requested i take them down, i had already got the answers i was looking for, so i no longer cared.
So baring this in mind, would it be safe to say that 1-3 counts of knock from 6000 rpms through redline isn't due to noise? So also baring in mind that my ECU was pulling no timing, are these considered "safe counts?"
Sorry Scott but i had to ask.
Scott not meaning to discredit you in any way, but this is actually back on topic of this thread, and i have to ask because you can see why this would worry me.
My log actually shows the exact same timing at the exact same RPM point, is this a problem (meaning 14*@6218RPMs w/ 3 knock counts)? I am assuming you meant it should be reduced on that specific car, and that each car is different; but i had to ask.
+1
Guys, i could care less about legal rights, and implecations, etc. I care about respect, and Scott asked the logs be removed, for his personal reasons, so who cares; so i did just that, not because i don't know my rights in this forum, but because i respect EVERYONE to a certain degree. It just so happens that Scott has a ton of my respect, and he has all the right in the world to ask anything from me, as many of you do, and that's that. At the point when he requested i take them down, i had already got the answers i was looking for, so i no longer cared.
Originally Posted by malibujack
also, a knock count of 3 at 2500rpm, is SUBSTANTIALLY different than a knock count of 3 at 6000rpm, this is where those knock filters come in, they set the threshold for the intensity and volume of different frequencies. A knock signature at 6000rpm becomes harder to discriminate from noise, because the background noise is so much greater.. Thats why low counts are important to watch, but its more important to watch how the ECU responds to it.
Sorry Scott but i had to ask.
Originally Posted by TTP Engineering
*Edit- actually looking at it further, it appears you have 14* @ 6000.
I would reduce timing to about 10* there.
I would reduce timing to about 10* there.
My log actually shows the exact same timing at the exact same RPM point, is this a problem (meaning 14*@6218RPMs w/ 3 knock counts)? I am assuming you meant it should be reduced on that specific car, and that each car is different; but i had to ask.
Originally Posted by nj1266
What was posted was NOT a tuner's map. What was posted was a log and a poorly executed log at that. I saw the second log and commented on it, before it was removed. KOevo started his log at 3200 rpm and ended it at 6800 rpm. That is not a good log that will allow people to steal TTP's tune from. A good log starts at 2500 rpm and goes all the way 7400 rpm. We do not even know under what conditions KO conducted his log. Was it done on a flat smooth surface? How many logs did he do before this one that he posted?
No one even asked KO where he got his tune from. No one even cared until TTP came into this thread and made an issue from it.
No one even asked KO where he got his tune from. No one even cared until TTP came into this thread and made an issue from it.
Last edited by KOEvo; Aug 24, 2006 at 09:01 AM.
Originally Posted by KOEvo
So baring this in mind, would it be safe to say that 1-3 counts of knock from 6000 rpms through redline isn't due to noise? So also baring in mind that my ECU was pulling no timing, are these considered "safe counts?"
Sorry Scott but i had to ask.
Scott not meaning to discredit you in any way, but this is actually back on topid of this thread, and i have to ask because you can see why this would worry me.
My log actually shows the exact same timing at the exact same RPM point, is this a problem (meaning 14*@6218RPMs)? I am assuming you meant is should be reduced on that specific car, and that each car is different; but i had to ask.
Sorry Scott but i had to ask.
Scott not meaning to discredit you in any way, but this is actually back on topid of this thread, and i have to ask because you can see why this would worry me.
My log actually shows the exact same timing at the exact same RPM point, is this a problem (meaning 14*@6218RPMs)? I am assuming you meant is should be reduced on that specific car, and that each car is different; but i had to ask.
Different octane of fuels are being used.
I would venture to say that your car is run richer than his also. All of these factors come into play.
Originally Posted by MalibuJack
What if I told you that I was writing a logger that could auto-fill a grid of cells that looks like a timing map after doing driving and pulls? What if I told you I could also auto-fill the AFR values in the map.. It will also trace the cells in your actual map.. I'm just working on the comm routines to get a MUT Stream then all of these things will come along..
Originally Posted by MalibuJack
My whole point is tuners need to spend their time making customers happy and not policing forums for maps and logs that are now legally their customers property.
The upside for those like Scott of TTP (who certainly deserves credit for his work) is the fact that no two EVOs are exactly alike (as we've seen from the knock-count discussion here), and two cars with otherwise identical performance hardware will oftentimes respond somewhat differently for a variety of reasons. This means that a custom tune for one EVO isn't necessarily going to be optimized for another. This keeps the door open for an accomplished tuner to earn a wage for custom tuning services (via dyno, road testing, etc.), and IMO, this is where the real value lies. This contrasts with monetary value of generic emailed reflashes however, which will soon be reduced to zero (if not already).
Originally Posted by AutoXer
There is something strange going on with the first 6 rows of this log. First the RPM goes up, then down, then back up. The first row also shows a low TPS and a rather low MAS Hz value. Maybe he shifted, then the clutch slipped? I ignored this portion of the log.
I concentrated on this section:
10 Knock Counts with 7*
10 Knock Counts with 7*
9 Knock Counts with 9*
Using the 3 knock counts per degree of timing pull that I've heard and partially verified... I get 10*, 10* and 12*.
I concentrated on this section:
Code:
28.47 5938 16 100 1604 79 100 10 7 254 28.59 6031 16 100 1604 79 100 10 7 254 28.72 6125 16 100 1604 79 100 9 9 254
10 Knock Counts with 7*
9 Knock Counts with 9*
Using the 3 knock counts per degree of timing pull that I've heard and partially verified... I get 10*, 10* and 12*.
Originally Posted by TTP Engineering
Yes, it is car specific. Judgement is being made based on logs. His knocks, yours does not.
Different octane of fuels are being used.
I would venture to say that your car is run richer than his also. All of these factors come into play.
Different octane of fuels are being used.
I would venture to say that your car is run richer than his also. All of these factors come into play.
Somehow, I doubt that MalibuJack is going to start asking for money for his tools. However, if a 'Donations' tab appears somewhere, I would also be very inclined to send something his way. I'm sure most people on this forum would do the same.
As far as reconstructing a tuner map from a scan, now that we have a good estimate of calculated LOAD, it would be very easy to make an Excel macro to do this in about 20 minutes. It may not hit every single cell, but you'll get the important ones.
But, then MJ goes and puts all these other, much cooler ideas into the mix, like automating the process so that you can do a scan and have the car build it's own tune. I picture a scenario where the user enters in how aggressive they want the tune to be (ie. choose your safety margin), then does a series of pulls, then the programs updates the tune and you just flash it in.
I was talking to a GM engineer a couple weeks ago. He said that there are homemade aftermarket programs for GM stuff (smallblock V8s and such) that are now actually more advanced than the tools the GM Powertrain engineers are using. He described exactly what I mentioned above... essentially the software watches knock averages and AFR over several runs, then creates an optimized custom tune for that car. You just copy the recommended tune values into your ECU and flash it. I have no doubt Jack and company will have something like this working by the end of the year. It really is a very cool time to own an EVO.
As far as reconstructing a tuner map from a scan, now that we have a good estimate of calculated LOAD, it would be very easy to make an Excel macro to do this in about 20 minutes. It may not hit every single cell, but you'll get the important ones.
But, then MJ goes and puts all these other, much cooler ideas into the mix, like automating the process so that you can do a scan and have the car build it's own tune. I picture a scenario where the user enters in how aggressive they want the tune to be (ie. choose your safety margin), then does a series of pulls, then the programs updates the tune and you just flash it in.
I was talking to a GM engineer a couple weeks ago. He said that there are homemade aftermarket programs for GM stuff (smallblock V8s and such) that are now actually more advanced than the tools the GM Powertrain engineers are using. He described exactly what I mentioned above... essentially the software watches knock averages and AFR over several runs, then creates an optimized custom tune for that car. You just copy the recommended tune values into your ECU and flash it. I have no doubt Jack and company will have something like this working by the end of the year. It really is a very cool time to own an EVO.
Originally Posted by ST
Is this what you call professionalism? I even asked you directly if you would guarantee me a better tune...I'm all ears and have my wallets open, but I never got one iota of a response. =/
Originally Posted by ST_on_NorCalEvo
So if i took my car to TTP, would you guarantee to make me more hp?
Last edited by razorlab; Aug 24, 2006 at 11:24 AM.
Thanks for the kind words.. The truth is the only time you may see a donation tab is to help keep the aktivematrix bills paid.
Originally Posted by Rob W.
Somehow, I doubt that MalibuJack is going to start asking for money for his tools. However, if a 'Donations' tab appears somewhere, I would also be very inclined to send something his way. I'm sure most people on this forum would do the same.
As far as reconstructing a tuner map from a scan, now that we have a good estimate of calculated LOAD, it would be very easy to make an Excel macro to do this in about 20 minutes. It may not hit every single cell, but you'll get the important ones.
But, then MJ goes and puts all these other, much cooler ideas into the mix, like automating the process so that you can do a scan and have the car build it's own tune. I picture a scenario where the user enters in how aggressive they want the tune to be (ie. choose your safety margin), then does a series of pulls, then the programs updates the tune and you just flash it in.
I was talking to a GM engineer a couple weeks ago. He said that there are homemade aftermarket programs for GM stuff (smallblock V8s and such) that are now actually more advanced than the tools the GM Powertrain engineers are using. He described exactly what I mentioned above... essentially the software watches knock averages and AFR over several runs, then creates an optimized custom tune for that car. You just copy the recommended tune values into your ECU and flash it. I have no doubt Jack and company will have something like this working by the end of the year. It really is a very cool time to own an EVO.
As far as reconstructing a tuner map from a scan, now that we have a good estimate of calculated LOAD, it would be very easy to make an Excel macro to do this in about 20 minutes. It may not hit every single cell, but you'll get the important ones.
But, then MJ goes and puts all these other, much cooler ideas into the mix, like automating the process so that you can do a scan and have the car build it's own tune. I picture a scenario where the user enters in how aggressive they want the tune to be (ie. choose your safety margin), then does a series of pulls, then the programs updates the tune and you just flash it in.
I was talking to a GM engineer a couple weeks ago. He said that there are homemade aftermarket programs for GM stuff (smallblock V8s and such) that are now actually more advanced than the tools the GM Powertrain engineers are using. He described exactly what I mentioned above... essentially the software watches knock averages and AFR over several runs, then creates an optimized custom tune for that car. You just copy the recommended tune values into your ECU and flash it. I have no doubt Jack and company will have something like this working by the end of the year. It really is a very cool time to own an EVO.
Originally Posted by Rob W.
Somehow, I doubt that MalibuJack is going to start asking for money for his tools. However, if a 'Donations' tab appears somewhere, I would also be very inclined to send something his way. I'm sure most people on this forum would do the same.
As far as reconstructing a tuner map from a scan, now that we have a good estimate of calculated LOAD, it would be very easy to make an Excel macro to do this in about 20 minutes. It may not hit every single cell, but you'll get the important ones.
But, then MJ goes and puts all these other, much cooler ideas into the mix, like automating the process so that you can do a scan and have the car build it's own tune. I picture a scenario where the user enters in how aggressive they want the tune to be (ie. choose your safety margin), then does a series of pulls, then the programs updates the tune and you just flash it in.
I was talking to a GM engineer a couple weeks ago. He said that there are homemade aftermarket programs for GM stuff (smallblock V8s and such) that are now actually more advanced than the tools the GM Powertrain engineers are using. He described exactly what I mentioned above... essentially the software watches knock averages and AFR over several runs, then creates an optimized custom tune for that car. You just copy the recommended tune values into your ECU and flash it. I have no doubt Jack and company will have something like this working by the end of the year. It really is a very cool time to own an EVO.
As far as reconstructing a tuner map from a scan, now that we have a good estimate of calculated LOAD, it would be very easy to make an Excel macro to do this in about 20 minutes. It may not hit every single cell, but you'll get the important ones.
But, then MJ goes and puts all these other, much cooler ideas into the mix, like automating the process so that you can do a scan and have the car build it's own tune. I picture a scenario where the user enters in how aggressive they want the tune to be (ie. choose your safety margin), then does a series of pulls, then the programs updates the tune and you just flash it in.
I was talking to a GM engineer a couple weeks ago. He said that there are homemade aftermarket programs for GM stuff (smallblock V8s and such) that are now actually more advanced than the tools the GM Powertrain engineers are using. He described exactly what I mentioned above... essentially the software watches knock averages and AFR over several runs, then creates an optimized custom tune for that car. You just copy the recommended tune values into your ECU and flash it. I have no doubt Jack and company will have something like this working by the end of the year. It really is a very cool time to own an EVO.
it would be nuts but it's not totally easy to do... that's effectively a stfs ltfs system but for WOT. this is also similar in spirit to vishnu's smart system... so we're combining ideas that already exist in our fantasizing... but it would be nuts. just wonder how hard it is to do
Well the program isn't THAT nuts yet.. It really just reconstructs a map from the log as its being collected (right now it does nothing more than digest through an EVoscan log) But the advantage of the reconstruction is its using the corrected values after the ECU pulls timing for knock events.. I'm at the moment only adding the feature to the program so you input your injector scale value, injector voltage latency, and it looks at the IPW and AFRMAP and whatever else is used in that load calculation, and puts in a new load column (not the mislabeled map sensor one) with the load value..
Next stage would be to read a ECU Map file, and be able to step trace through it while highlighting the map cells.. or do a rebuild which will fill the table with values from the log..
No autotuning here, eventually when all the comms are done, I can use AFR from a wideband and AFR from the map, and correlate the difference and then make adjustment recommendations to hit your target values. but this is a ways off since the tool is just a simple log converter at he moment, it just combines a few data items to create a time offset to correctly load into data log lab.. Then I just added the load calculation from existing data, and add it to the log.. Then in data log lab you can either create a distribution graph or graph the cells over time so you can easily match the load site/rpm to your map.
I'll post the new info as I have something in the existing thread for my tool.. I'll also dedicate a section to it on Aktivematrix so that I can focus on development in the next month.
Like I said, once I get the comms working, I'll no longer have to rely on Evoscan and it can progress into a more standalone tool.. Once I get that working I'll focus on some of the more advanced features.
Eventually it'll be able to gather much of this information directly from your map associated with the logs.
Next stage would be to read a ECU Map file, and be able to step trace through it while highlighting the map cells.. or do a rebuild which will fill the table with values from the log..
No autotuning here, eventually when all the comms are done, I can use AFR from a wideband and AFR from the map, and correlate the difference and then make adjustment recommendations to hit your target values. but this is a ways off since the tool is just a simple log converter at he moment, it just combines a few data items to create a time offset to correctly load into data log lab.. Then I just added the load calculation from existing data, and add it to the log.. Then in data log lab you can either create a distribution graph or graph the cells over time so you can easily match the load site/rpm to your map.
I'll post the new info as I have something in the existing thread for my tool.. I'll also dedicate a section to it on Aktivematrix so that I can focus on development in the next month.
Like I said, once I get the comms working, I'll no longer have to rely on Evoscan and it can progress into a more standalone tool.. Once I get that working I'll focus on some of the more advanced features.
Eventually it'll be able to gather much of this information directly from your map associated with the logs.
Originally Posted by TTP Engineering
The other shop in question that stole a Dynoflash map was and is NOT an evolutionm.net vendor.
His tunes produce some of the best power here locally. Additionally, his customer's personal best E.T. is an 11.5 @ 119 mph on the stock 05' turbo. He is also a licensed Ecutek dealer, and has been tuning dsms for sometime.
Besides that he is an honest and straight forward person, who works on his customers cars well beyond normal work hours, while charging them a flat rate to ensure they are satisfied with the output of their cars on his dyno.
There are many different ways and many similar ways to arrive at the same results, but to accuse Paul of "stealing" because his map resembles dynoflash's is a huge jump of assumptions when you don't know Paul and you don't know under what conditions that map came into existence.
Fair Enough Andy... There is no solid proof. Nevertheless, you cannot argue how strikingly 95% similar the maps are for ign timing.
Originally Posted by EFIxMR
I think to accuse Paul of stealing a Dynoflash map is ridiculous and in poor taste. Paul is not a big internet guy, and has earned a great reputation locally with his customers, and cannot come on here and defend himself.
His tunes produce some of the best power here locally. Additionally, his customer's personal best E.T. is an 11.5 @ 119 mph on the stock 05' turbo. He is also a licensed Ecutek dealer, and has been tuning dsms for sometime.
Besides that he is an honest and straight forward person, who works on his customers cars well beyond normal work hours, while charging them a flat rate to ensure they are satisfied with the output of their cars on his dyno.
There are many different ways and many similar ways to arrive at the same results, but to accuse Paul of "stealing" because his map resembles dynoflash's is a huge jump of assumptions when you don't know Paul and you don't know under what conditions that map came into existence.
His tunes produce some of the best power here locally. Additionally, his customer's personal best E.T. is an 11.5 @ 119 mph on the stock 05' turbo. He is also a licensed Ecutek dealer, and has been tuning dsms for sometime.
Besides that he is an honest and straight forward person, who works on his customers cars well beyond normal work hours, while charging them a flat rate to ensure they are satisfied with the output of their cars on his dyno.
There are many different ways and many similar ways to arrive at the same results, but to accuse Paul of "stealing" because his map resembles dynoflash's is a huge jump of assumptions when you don't know Paul and you don't know under what conditions that map came into existence.