Greg Fell

iTrader: (9)

I have created a spreadsheet that computes a value which will (in theory?) give the best shift point for a given dyno curve.

I plugged in my numbers, and was surprised to see that my shift points are higher than I expected, and was told they should be.

Could you guys plug in some of your numbers and see what you think, and how it works?

Please keep in mind this is a quick and dirty, and if it works, I will polish it up and give to anyone who wants it.

kapedogg

iTrader: (5)

I'll plug in my numbers when I get home tonight and see what it says. However, I didn't see where on the sheet it displays the best shift point. I tried clicking and double clicking on the "find shift point" box. The double click pulled up another window, but I didn't see it there either?? I'm sure I'm probably looking right over it...

Greg Fell

iTrader: (9)

I made a small change, try it now. I'm not sure this is going to work properly, but I'd like some help in finding out what to adjust in order to make it a useful tool.

RedHardSupra

iTrader: (1)

let's say, we got a fairly stockish M6, so you get 2400rpm @40mph to 6000rpm @100mph in 3rd, and you wanna optimize a shift to 4th.
you're integrating intervals for it, so you got (6000-2400)=3600rpm range, that's 35 intervals of 100rpm. so you integrate it, and let's say you average 300hp (totally arbitrary example). since all intervals are equal, you can assume that work is directly proportional to power. so for the sake of this calculations (since units are all over the place anyway, we just want some numbers for comparison's sake) we're just gonna sum up the HP ratings for each interval.

so you have 35 intervals of 300hp/interval, that's a total 'work' of 10500.
let's say you shift at 6000rpm and you 'land' at (1/1.43*6000)=4200rpm. then you run it up to 6000rpm again, so that's 1800rpm (17 intervals, at 300hp/interval) that's 5100 work units.
so total work, going from 2400rpm->6000rpm in 3rd, shifting to 4th and then going up to 6000rpm gets us 10500+5100=15600 work units.

now let's do the same, but now we wanna shift at 6500rpm.
3rd gear work:
6500-2400=4100rpm (40 intervals)
40 intervals * 300hp/interval=12000 work units
shift to 4th:
6500rpm*(1/1.43)=4500 (rounded down)
4th gear work:
6500-4500=2000rpm (19 intervals)
19 intervals* 300hp/interval=5700 work units
4th+3rd gear work: 5700+12000=17700 work units.
oh, we got us a healthy increase!
not so quick...
let's count up the intervals:
6000shiftpoint: 35+17=52
6500shiftpoint: 40+19=59

ooops... so how can we compare these things when we got different amount of intervals? 2 ways: go by arbitrary bounds (speed for road racers, distance for drag would work best) or calculate averages PER INTERVAL.
the first one is a much more complicated thing, so we'll skip it for now (i'm about to leave for breakfast), and we'll go with the interval averages:
6000shiftpoint: 15600/52=300
6500shiftpoint: 17700/59=300
see! you didn't do any more work, you didn't optimize crap, because you compared work done over two different numbers of intervals!
thus, we need to go by the hard options, and find us some more arbitrary bounds...

to be continued, now i'm off for some omlettes...

Greg Fell

iTrader: (9)

I have updated the spreadsheet some, it takes into account intervals now.

LS1 Sounds

iTrader: (11)

RedHardSupra, Spreadsheet Assassin.

AZ28DRIVER1

There is a book called "Auto Math" (I forget who the author is) but anyways a simple rule for shift points from that book is to shift at an RPM that is 10% above the RPM where you make max horsepower. So for a stock LS1 that makes max pwer at 5,200 RPM you would want to shift at 5,200 x 1.10 = 5,720 RPM.

You'll find that by doing that the drop off in RPM will be at the ideal point for where your car is making max torque.

Greg Fell

iTrader: (9)

Hmmm, interesting.

I have found that shift points to be different based on the tranny used (gear ratios), so I think the 10% thing may be a bit too vague.

RedHardSupra

iTrader: (1)

i have that book, and that rule sucks. rpms are free horsepower, and not taking the real shape of the powerband into consideration is...suboptimal?
probably later today i'll post my version of this spreadsheet, so you guys can play with the numbers and see how non-obvious answers it will sometimes give you.
the hardest part about this particular problem is the interpretation of results. since we're working with fake figures, and over a range of speeds, it's difficult to compare the results. that's why i decided to go by the averages...but you'll see for yourself later

Greg Fell

iTrader: (9)

It would be nice to have others use their dyno numbers and shift points they've found to be the best in the real world. I used my dyno numbers, and will shift at 6700 instead of 6500 next time out (since thats what my calc came back with).

HOSS99

I think there is a mistake in the calculator. I tried a few different numbers to see if the shiftpoint would change, and no matter what # I plugged in at 6800 rpm's it always told me to shift at 6800. I even entered in a 100 horsepower decrease from 6200 to 6800. The results of the program are always saying to shift @ the highest rpm entered.

Greg Fell

iTrader: (9)

Can you send me the xls file you are using? I can troubleshoot easy that way

gregorybfell@yahoo.com

RedHardSupra

iTrader: (1)

heheh, that's the non-obvious results i was talking about... start dropping it off more and more on top end, and do it again, you'll see it will eventually not give you the highest rpm. the trick is that even if the power on top drops off, you're still gaining on the next gear starting off higher. so you can't think of it in context of just one gear, it's a sum of many little areas

RedHardSupra

iTrader: (1)

http://redhardsupra.blogspot.com/200...ft-points.html

check it out, it's cool