options request : post your ideas

[5.3]

would love to see 0-60 timer and 1/4 mile in the scanner options ,looks prity easy to add

[cmitchell17]

I want to see a 0-60 time and a quarter mile calculator too. It seems like it would be easy to calculate distance and then you already have speed.

Then who would need a dragstrip and I wouldnt see why it would not be a perfect way to test changes and tuning changes.

And then have a seperate window that would graph the delievered torque and calcualted horsepower. I guess all you would need to know is your weight.

[redhardsupra]

actually getting acceleration from speed is nowhere near as simple as basic physics would have you believe. the biggest problem stems from resolution. if you have your speed in mph, you will only see a gain if you make at least one mph improvment. if you're going for small changes, you wont see it. if you have two intervals where the speed happens to be like 81.1 and 81.9, you'll see 0 acceleration, which is just not true.
if you want to measure acceleration, i recommend using a single
gear pull in a range where you dont have traction problems and where air resistance isn't significant either. so for a normal 6sp fbody 2400-6000rpm in 3rd works well. take a bunch of measurements, know what your average time is, and then make changes, see how much that time changes. this method works. in one long night (27 runs!) i managed to drop a time in that interval for my friend's car from 10.2 to 8.7secs. this is the best 'area under the curve' measurement i found. you can also divide it up into intervals so you know which changes where made biggest improvment or loss.
oh, another trick: for stuff like that measure least PIDs you can, so you can increase the sampling frequency.

[joecar]

High resolution is the key...

Power = m * v * dv/dt

dv/dt requires high resolution (high sampling rate) of v in order to be accurate; and mulitplying by v makes it worse.

For example, the G-Tech runs at something like 100+ samples/s...

Our PCM's Fast mode 10 frames/s just doesn't give enough resolution for accurate differentiation/integration.

If you log just one pid (VSS) in Dynamic mode, you may get 40 frames/s, accuracy will be better, but it's still not enough (not to mention you won't have many other pids).

[dfe1]

Just a thought-- At some point additional "features" get in the way of the original purpose of a program. FlashScan is a scanning/tuning package not a vehicle data aquisition program. Some of the accelerometer-based systems currently available do a fairly good job of computing 0-60 and quarter-mile times and they're reasonably priced. I think there are much more important scanning/tuning issues to be addressed, especially since GM now seems to think that every vehicle needs a unique ECM and/or operating system. I think Paul and Ross are doing a kick-ass job and we should minimize diversions that take time and focus away from tasks that have much higher priorities.

[mr.prick]

the point of this thread is to give the manufacturer some feed back from it`s costumers and maybe help with future improvement that users want.
and maybe point out some things that are lacking .

http://forum.efilive.com/showthread....ht=problem+7.4

[redhardsupra]

well, if we're going broad ideas, then i want scatter plots, trendlines, and histograms in bar form, not just numbers. also charting any arbitrary functions on top of data would be nice, along with plotting distributions of residuals between data and the fitted model.
being able to display more than one piece of data in a cell would be great. you do it already on the charts, where i can see current/min/max/avg, if i could do that in the histograms that'd really make my life easier.
color coding cells based on some other variables would be good too. ie cells with a large std dev of data in that cell should be highlighted to identify trouble spots.
and i want a pony

[joecar]

Here's what I would like...

2 independent sequence id's for file name use:
tun filename: xxxx_<seq1>
log filename: xxxx_<seq1>_<seq2>

where the <seq1> in the log filename tracks the <seq1> in the tun filename,
and <seq2> resets back to 0000 whenever <seq1> goes to a new value.

i.e. so when you save a tun with a sequenced filename, the subsequent log filenames will match the tun filename.

[bK]

I use the colour coding (MIN to MAX) in the maps in Scan tool to ensure that I've hit all the cells enough times. When driving looking at colours is much better and quicker then trying to read the numbers.

Is there a way to tell the Scan tool that if the cell count is above a certain number it's reached the MAX? That way with the pic below I could set the number to 100 and will be able to see the smaller cell counts with better colour contrast.

[mr.prick]

keep the ideas coming maybe someone is watching