Old Fashioned Tuning - Re: [Gmecm] EGT tuning (was BIG CAM TUNING)

[Jay Vessels]

Hi there!

Another DIY tuning method (with some upfront cost) is to log 
accelerometer data.  This gets you the same information in reverse; the 
benefit is that the accelerometer can be used for lateral measurements 
and braking time (though braking could be retrieved from high-res VSS 
data assuming no wheel lockup.)  I like logging spark events, since the 
interface to a PC would be pretty simple (I'm thinking an opto-isolated 
connection to the parallel port, and a bit of code to log a timestamp 
each time you get a spark) and once you gather the rough data, you can 
use your favorite existing software (Excel, Matlab, etc.) to manipulate 
the data any way you want.

It's funny -- when all we had was a distributor with weights and 
springs, and maybe a vacuum can, it didn't take much to get one dialed 
in as good as you could; make a weekend out of the "test and tune" at 
the local 'strip and you're there.  Carbs have many more knobs to turn 
but, like chip burning, they're not that easy to get right *quickly* 
once you get past gross changes like jets and needles.

EFI is a funny thing -- you can get *real close* really quickly, and 
it's forgiving in that sense; it's what makes the cars run so well even 
when the engine's out of tune or the system's been transplanted out of 
its native element.  However, there are a good order of magnitude more 
knobs available.  To get it *right* means burning a lot of chips, taking 
careful notes, and thinking about what the vehicle is trying to tell you.

Ghosts of Grumpy rattling around here,
Jay Vessels
1982 Chevrolet S-10 Sport, 2.8V6 TBI
1984 Chevrolet S-10 Blazer Sport, 2.8V6 carb. (for now)

bcroe at juno.com wrote:
> 19 Jun 2005  DV Fagan <dvfagan at yahoo.com> writes:
> 
>>Why not go to a drag strip, put a time only slip on it
>>and make repeated runs advancing the distributor 2
>>degrees per run until the et's fail to decrease.  
> 
> 
> That assumes your advance already has the ideal curve.  
> You might improve acceleration in one area while making 
> it worse in another with a net  zero change that way.  
> 
> Instead I'd log period between sparks, which is 
> the reverse of sparks per sec.  This will of course 
> decrease as you accelerate.  Subtract the latest 
> period from the previous one at each rpm to get the 
> the first derivitive, Acceleration.  Plot A vs rev/sec.  
> NOW move your whole curve forward and back a couple 
> degrees, and plot those on the same graph.  You 
> may see that a couple degrees more advance is 
> better at some rev/sec, a couple degrees retard is 
> better at other rev/sec, no change other places.  
> Now reshape your advance curve for the best of 
> each to be optimum everywhere.  
> 
> Likely taking the periods something like 20 sparks 
> at a time will be more practical.  
> 
> The same problem comes with just varying fuel 
> pressure.  Bruce Roe
> 
> Then
> 
>>at least  you will know what max ignition advance
>>should be. It will be a fun evening and it shouldn't
>>cost a lot.  On another evening you could increase
>>fuel pressure with a  dozen quarters placed in the
>>regulator one at a time in a like manner.  No
>>programming and little cost (a gasket set for $6).  Or
>>you  could reprogram the ecm 90+ times as some on this
>>list claim they have done and spend years and
>>kilobucks on chip programmers, erasers, ALDL
>>interfaces, software, a laptop and PROM emulators. 
>>It's your choice.
>>Dennis 
>>
>>--- Jay Vessels <jay at vessels-clan.com> wrote:
>>
>>
>>>Hi there!
>>>
>>>A gas analyzer would be the best bet but it's not as
>>>fast as a WB so 
>>>you're probably going to only use it for a dyno with
>>>a brake.