This is how to pin-point what load cell your car is operating in!
Jul 3, 2006, 08:51 AM
#16
Evolved Member
iTrader: (8)
Join Date: Apr 2003
Location: Georgia
Posts: 2,138
Likes: 0
Received 0 Likes
on
0 Posts
Good show always,
I figured out real quick as well that the block fill method is crap. At least by doing this you can plot out your path through the map for a given run. The difficult part is those big transitions where you have to do a lot of mental interpolation to figure out the path and the smoothing, like between 2500 and 3000 rpm :-)
Other than that it seems to work like a champ so far,
I will say that until i tried EvoScan i did not see that I was getting a lot of 1 count knock, my utec does not show any knock until
the ecu hits a count of 3 I think.
Hey, on a side note, anyone know if we can use decimal values in the timing map like 6.5 rather than having to choose between 6 or 7??????
I figured out real quick as well that the block fill method is crap. At least by doing this you can plot out your path through the map for a given run. The difficult part is those big transitions where you have to do a lot of mental interpolation to figure out the path and the smoothing, like between 2500 and 3000 rpm :-)
Other than that it seems to work like a champ so far,
I will say that until i tried EvoScan i did not see that I was getting a lot of 1 count knock, my utec does not show any knock until
the ecu hits a count of 3 I think.
Hey, on a side note, anyone know if we can use decimal values in the timing map like 6.5 rather than having to choose between 6 or 7??????
Jul 3, 2006, 08:59 AM
#17
The timing values unfortunately aren't that granular.. I have to look at the table its being extracted from, but its basically a decimal conversion of a hex number, so there are few if any places you can have a partial degree value.. Yeah its a single byte integer value, there isn't any room in the "Hex value" to allow for sub-degree timing..
Jul 3, 2006, 09:12 AM
#18
Evolved Member
Thread Starter
iTrader: (17)
Join Date: Aug 2004
Location: In da streetz
Posts: 3,275
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by mad_VIII
Hey, on a side note, anyone know if we can use decimal values in the timing map like 6.5 rather than having to choose between 6 or 7??????
Jul 3, 2006, 09:17 AM
#19
Evolved Member
Thread Starter
iTrader: (17)
Join Date: Aug 2004
Location: In da streetz
Posts: 3,275
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Noogles
I have always tuned with this method. If a car can run the same amount of timing at 260 load and 220 why not run it? There is no reason to loss power in the spirit of having a pretty table.
After doing this for only a couple of days its kind of like the engine tells you how it likes to be tuned. Very weird I know but there is no other way to explain it. My car likes a lot of fuel in this heat and the timing curve I have shown above. With those two combinations it runs perfectly. Any less fuel it knocks, any more timing in the HIGH load area and it knocks. Right now she is VERY happy and so am I
Jul 3, 2006, 09:29 AM
#20
Evolving Member
Join Date: Sep 2002
Location: Florida
Posts: 383
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by MalibuJack
The timing values unfortunately aren't that granular.. I have to look at the table its being extracted from, but its basically a decimal conversion of a hex number, so there are few if any places you can have a partial degree value.. Yeah its a single byte integer value, there isn't any room in the "Hex value" to allow for sub-degree timing..
Yup that is the only draw back to the stock ecu. THat is where having a piggy back can eek out the last bits of power.
Jul 3, 2006, 09:40 AM
#21
Evolved Member
iTrader: (17)
Join Date: Nov 2005
Location: NNJ
Posts: 2,544
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by MalibuJack
The timing values unfortunately aren't that granular.. I have to look at the table its being extracted from, but its basically a decimal conversion of a hex number, so there are few if any places you can have a partial degree value.. Yeah its a single byte integer value, there isn't any room in the "Hex value" to allow for sub-degree timing..
Jul 3, 2006, 09:41 AM
#22
Great info here. For just starting to learn to tune myself, I appreciate all this good info. Now I just need someone to help me extract my maps from my ECUFlash ROM so I can actually tune my turbo'd Lancer and remove all the dangerous knock I have!
Jul 3, 2006, 09:48 AM
#23
Originally Posted by dudical26
Again, to throw in some of my Subaru tuning experience...the program that I use for Subaru's is called ECUExplorer. It allows you to enter some decimal figures. For example if you put in 4.5, it would round that to the nearest acceptable decimal number that can be converted to the hex file. This support would of course have to be built into Open Flash so that it knows what decimal numbers are ok. Do you follow or did I not explain it well enough?
Jul 3, 2006, 09:51 AM
#24
Originally Posted by AlwaysinBoost
I've found the ECU won't pull timing if there is only 1 count of knock, unless the knock is present for a long period of time. That threshold is a little less tolorant on 2 counts, meaning much less time until it pulls timing. On 3 counts I usually see 2*'s of timing being pulled or more depending on how long the knock lasts.
Anyway, in the DSM ECU, approximately 1 degree of timing was pulled for every 3 counts of knock, up to a maximum of 12 degrees of timing retard.
I would have guessed that the Evo ecu was similar, but you are saying that you are seeing different?
For example, 9 counts of knock in the DSM Ecu would pull 3* timing. 3 counts would pull 1*, etc.
Once I get logging in the coming weeks I will nail this down for sure, but I was curious as to what others saw.
But for sure, I will be creating a knock free tune.
Eric
Jul 3, 2006, 10:02 AM
#25
Evolving Member
iTrader: (6)
Join Date: Feb 2004
Location: Maryland
Posts: 442
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by AlwaysinBoost
http://forums.openecu.org/viewtopic.php?p=6085#6085
you need to be logged in to view the attachment.
you need to be logged in to view the attachment.
I've registered twice at openecu.org, but never received any email to activate my account. I have no idea why that is.
kapolani@hotmail.com
Jul 3, 2006, 12:46 PM
#27
Evolved Member
Thread Starter
iTrader: (17)
Join Date: Aug 2004
Location: In da streetz
Posts: 3,275
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by l2r99gst
I have yet to capture any logs, since I'm waiting for the SSi4 from Innovate to come out this week, so I can hook up and log my GM 3.3 bar map sensor.
Anyway, in the DSM ECU, approximately 1 degree of timing was pulled for every 3 counts of knock, up to a maximum of 12 degrees of timing retard.
I would have guessed that the Evo ecu was similar, but you are saying that you are seeing different?
For example, 9 counts of knock in the DSM Ecu would pull 3* timing. 3 counts would pull 1*, etc.
Once I get logging in the coming weeks I will nail this down for sure, but I was curious as to what others saw.
But for sure, I will be creating a knock free tune.
Eric
Anyway, in the DSM ECU, approximately 1 degree of timing was pulled for every 3 counts of knock, up to a maximum of 12 degrees of timing retard.
I would have guessed that the Evo ecu was similar, but you are saying that you are seeing different?
For example, 9 counts of knock in the DSM Ecu would pull 3* timing. 3 counts would pull 1*, etc.
Once I get logging in the coming weeks I will nail this down for sure, but I was curious as to what others saw.
But for sure, I will be creating a knock free tune.
Eric
You can still 'get by' with the tools we have now. Just the EVOScan program & Second Chances' logger are all that I used to get a completly knock free tune.
With regard to knock, what I posted above is what my logs have shown.
1 count = no timing pulled unless the knock continues for a prolonged period of time.
2 counts = no timing pulled if it happens for a VERY short time, longer then that and a degree or two will be pulled.
3 counts or more and at least 2 degrees of timing get pulled instantly.
More knock then that I don't recall how much timing was pulled because I was looking at other parts of the map to figure out which cells to modify in order to eliminate it.
Jul 3, 2006, 02:19 PM
#28
Originally Posted by AlwaysinBoost
So they finally released the SSI4... I was initially waiting until that came out but just gave up after a month or so.
You can still 'get by' with the tools we have now. Just the EVOScan program & Second Chances' logger are all that I used to get a completly knock free tune.
With regard to knock, what I posted above is what my logs have shown.
1 count = no timing pulled unless the knock continues for a prolonged period of time.
2 counts = no timing pulled if it happens for a VERY short time, longer then that and a degree or two will be pulled.
3 counts or more and at least 2 degrees of timing get pulled instantly.
More knock then that I don't recall how much timing was pulled because I was looking at other parts of the map to figure out which cells to modify in order to eliminate it.
You can still 'get by' with the tools we have now. Just the EVOScan program & Second Chances' logger are all that I used to get a completly knock free tune.
With regard to knock, what I posted above is what my logs have shown.
1 count = no timing pulled unless the knock continues for a prolonged period of time.
2 counts = no timing pulled if it happens for a VERY short time, longer then that and a degree or two will be pulled.
3 counts or more and at least 2 degrees of timing get pulled instantly.
More knock then that I don't recall how much timing was pulled because I was looking at other parts of the map to figure out which cells to modify in order to eliminate it.
By no means did I mean that I can't use the present tools to tune my car. I just wanted to use a boost gauge and boost controller, too. I never installed either because I was waiting on the SSi4. I have and XD-16 for the LC-1 wideband and I will get another XD-16 for the boost gauge once the SSi4 comes in. I will be using a GM 3.3 bar map sensor with that. That way I can log both boost and wideband data and display it on the gauges and my CarPC along with the EvoScan software.
Your findings on knock retard are interesting, though. I will definitely verify your findings once I get going here.
Eric
Jul 3, 2006, 02:44 PM
#29
I posted this in another thread but I guess it's more appropriate here:
22.5psi peak, 91 octane.
RPM -- MAS Hz -- Timing -- Wideband AFR
2513 -- 384 -- 12 -- 13.1
3024 -- 636 -- 6 -- 11.9
3501 -- 992 -- 5 -- 11.4
4035 -- 1121 -- 6 -- 11.3
4513 -- 1266 -- 7 -- 11.1
5043 -- 1393 -- 9 -- 11.2
5522 -- 1492 -- 10 -- 11.2
6002 -- 1595 -- 12 -- 11
6515 -- 1694 -- 13 -- 11
7002 -- 1731 -- 17 -- 11.1
7528 -- 1796 -- 19 -- 11
Here is the path my ECU is taking through the load columns according to my logs:
22.5psi peak, 91 octane.
RPM -- MAS Hz -- Timing -- Wideband AFR
2513 -- 384 -- 12 -- 13.1
3024 -- 636 -- 6 -- 11.9
3501 -- 992 -- 5 -- 11.4
4035 -- 1121 -- 6 -- 11.3
4513 -- 1266 -- 7 -- 11.1
5043 -- 1393 -- 9 -- 11.2
5522 -- 1492 -- 10 -- 11.2
6002 -- 1595 -- 12 -- 11
6515 -- 1694 -- 13 -- 11
7002 -- 1731 -- 17 -- 11.1
7528 -- 1796 -- 19 -- 11
Here is the path my ECU is taking through the load columns according to my logs:
Jul 3, 2006, 03:39 PM
#30
Evolved Member
Thread Starter
iTrader: (17)
Join Date: Aug 2004
Location: In da streetz
Posts: 3,275
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by razorlab
I posted this in another thread but I guess it's more appropriate here:
22.5psi peak, 91 octane.
RPM -- MAS Hz -- Timing -- Wideband AFR
2513 -- 384 -- 12 -- 13.1
3024 -- 636 -- 6 -- 11.9
3501 -- 992 -- 5 -- 11.4
4035 -- 1121 -- 6 -- 11.3
4513 -- 1266 -- 7 -- 11.1
5043 -- 1393 -- 9 -- 11.2
5522 -- 1492 -- 10 -- 11.2
6002 -- 1595 -- 12 -- 11
6515 -- 1694 -- 13 -- 11
7002 -- 1731 -- 17 -- 11.1
7528 -- 1796 -- 19 -- 11
Here is the path my ECU is taking through the load columns according to my logs:
22.5psi peak, 91 octane.
RPM -- MAS Hz -- Timing -- Wideband AFR
2513 -- 384 -- 12 -- 13.1
3024 -- 636 -- 6 -- 11.9
3501 -- 992 -- 5 -- 11.4
4035 -- 1121 -- 6 -- 11.3
4513 -- 1266 -- 7 -- 11.1
5043 -- 1393 -- 9 -- 11.2
5522 -- 1492 -- 10 -- 11.2
6002 -- 1595 -- 12 -- 11
6515 -- 1694 -- 13 -- 11
7002 -- 1731 -- 17 -- 11.1
7528 -- 1796 -- 19 -- 11
Here is the path my ECU is taking through the load columns according to my logs: