Spark Bias (new ECU now there is warm and cold)

Subscribe
Dec 8, 2005 | 10:14 PM
  #1  
Bias Q:
I notice that the cold bias is 20, and the warm bias is 9.84.
In my previous ECUs, there was just one bias. 9.84


The coolant compensation has alot of 20's in it, where it would normally be 0. So that's where the cold bias works in.

The warm bias is 9.84... does that mean that I'm getting another 11 degrees at all times when the engine's warm?!~ That might explain my lean pop! lol

Thanks
-Paul
Reply 0
Dec 8, 2005 | 10:39 PM
  #2  
Re: Spark Bias (new ECU now there is warm and cold)
Quote:
Originally posted by Twilightoptics
Bias Q:
I notice that the cold bias is 20, and the warm bias is 9.84.
In my previous ECUs, there was just one bias. 9.84


The coolant compensation has alot of 20's in it, where it would normally be 0. So that's where the cold bias works in.

The warm bias is 9.84... does that mean that I'm getting another 11 degrees at all times when the engine's warm?!~ That might explain my lean pop! lol

Thanks
-Paul
From your other post, I'm assuming you're using the 7747 now and not $6E or $8D. If that is the case, the warm bias is the main SA bias table not "just" warm. It's pretty much all the time. All of the biases are subtracted out. If the main SA table has lets say 30, and the bias is 10 then engine will get 20 plus (or minus) the other SA tables (cts, startup, tcc locked). With the xdf file format you can have the main tables displayed with the bias already subtracted. Just modify the conversion of the main SA to include the bias, very easy to do. If you have questions about this let me know.
Reply 0
Dec 8, 2005 | 11:07 PM
  #3  
I have a 7747-mine.xdf and ecu.

The xdf has less tuff.

I just import the ECU in. It has the bias not taken out of the main table.

In the 42.xdf file I have, if I set the main bias to 0, it would add the original 9 to the main table. With the bias set at nine, it would subtract from the main table.

It's confusing as to why there are two. With the $6e I used to tune the bias was 20 all around.

So what you're saying is the "warm" bias set at 9.84.... is the bias for the main table.....

And the "cool" bias, is just for the coolant compensation table?



Bummer, I was hoping I was getting mass amounts of spark and it'd be an easy AE fix. LOL
Reply 0
Dec 8, 2005 | 11:08 PM
  #4  
Quote:
Originally posted by Twilightoptics
I have a 7747-mine.xdf and ecu.

The xdf has less tuff.

I just import the ECU in. It has the bias not taken out of the main table.

In the 42.xdf file I have, if I set the main bias to 0, it would add the original 9 to the main table. With the bias set at nine, it would subtract from the main table.

It's confusing as to why there are two. With the $6e I used to tune the bias was 20 all around.

So what you're saying is the "warm" bias set at 9.84.... is the bias for the main table.....

And the "cool" bias, is just for the coolant compensation table?



Bummer, I was hoping I was getting mass amounts of spark and it'd be an easy AE fix. LOL
Correct.
Reply 0
Dec 12, 2005 | 10:57 AM
  #5  
Quote:
Originally posted by Twilightoptics
It's confusing as to why there are two. With the $6e I used to tune the bias was 20 all around.

So what you're saying is the "warm" bias set at 9.84.... is the bias for the main table.....

And the "cool" bias, is just for the coolant compensation table?
As it was explained to me...there is a warm spark bias so GM could have spark retard (or negative advance) in the main SA table without actually having a negative value.

Let's say they wanted -1° of advance. With a 9° bias they would put 8° in that cell. 8° - 9° = -1°. I'm not a programmer, but I think it takes up more space in the cal to have a negative number.

That ECM was generally used in trucks so there are alot of places in the SA table where they wanted negative SA for towing and lo-po engines.
Reply 0
Dec 12, 2005 | 01:42 PM
  #6  
I don understand the idea and concept behind the bias, what was confusing was having a "warm" bias and a "cold" bias with just one compensation table.
Reply 0
Dec 22, 2005 | 06:03 PM
  #7  
It's not that having negative numbers takes up more space, they cut the granulatiry of the table in half. An 8-bit number can represent 0 - 255 when unsigned, but only -128 - 127 when signed. So instead they use all 8 bits across the table for the value, and one extra byte to house the bias.

Teeleton
Reply 0
Subscribe