Page 1 of 3

Home dyno?

Posted: Sun Sep 03, 2006 10:31 pm
by S.Bretz
Has program been made that can calculate a MS user's HP based on rpm accleration rates? I was looking into something like this, but all it needs an rpm signal....which is somethin that is already logged in a DL :wink:

http://www.charm.net/~mchaney/roaddyno/index.html


Was also think it would be cool if You could hook an accerometer up to the MS...maybe one of those coupled with software that read rpm accel rates would make a real accurate software/road dyno.

Posted: Mon Sep 04, 2006 4:39 am
by Dr.Hess
I thought of this as well. We have the data available (time and RPM). The only data points to add are the mass and final gear ratio, both constants. Just need to do the calculations and plot it.

Posted: Mon Sep 04, 2006 5:14 am
by gmt
Hmn.. nice idea! Maybe move it to megatune forum?

Posted: Mon Sep 04, 2006 5:57 am
by renns
This has been discussed in the past. As I recall the problem was the 100 rpm increment in engine speed (gone with msnsextra), combined with the fact that serial comms being asynchronous may not present data at a fast enough rate. There was some talk of calculating a 'rate of change of rpm' value within MS, and then passing that out to MegaTune.

Feel free to give it a go, though. If you dig into the MegaTune .ini file, you can likely find enough capability there to create a rudimentary 'hp' gauge with some vehicle-specific values hard-coded in the .ini file.

Roger.

Posted: Mon Sep 04, 2006 1:01 pm
by Bernard Fife
Hi,

Yeah, there are a bunch of problems with something like this. One is the rpm resolution, though one could use the MS-II or MSnS-E code.

Harder to solve problems include things like that you don't know when the wheels are spinning, if you are headed down a hill, into (or with) a stiff head wind, etc., etc., etc. So it really only tells you about horsepower as long as you are on the straight and level, on a perfect windless day, and don't lose traction. So it's pretty limited at best.

For example, if you wanted 1000 'gauge' HP, you could rev the engine in neutral and run the datalog through the calcs! :shock:

Because this hasn't been done yet, it seems most people would rather spend time making more horsepower than measuring a small amount of horsepower.

For my money, something like the G-Tech Pro accelerometer is a better bet for this purpose.

Lance.

Posted: Tue Sep 05, 2006 3:40 pm
by S.Bretz
I beg to differ.....they are about 300 dollars and would still have the same problems with "stong head/tail winds". It would bw just like a hammer, or any other tool for that matter. If you can't hit a nail on the head, then you can use a hammer....if you try to make runs up or down a hill, or use data logs when the tires are slipping, then you don't know how to use the tool.

Posted: Tue Sep 05, 2006 4:06 pm
by Bernard Fife
S.Bretz,

But they don't have the wheelspin problem (at least not for overestimating power), since they are accelerometer based. For cars with a decent amount of horsepower, wheelspin is a huge concern (and not always easy to detect).

In any case, if you think that this is a good idea, by all means do it. My point was that no-one who has looked into it carefully has bothered to make it work, to my knowledge. If you want to be the first, by all means do it.

Lance.

Posted: Wed Sep 06, 2006 9:50 am
by S.Bretz
Gtech's also use the tach signal that they pick up threw a power outlet.

Your point of spinning tires and getting on over estimated out put is errelivant as that case would fall under the guides of "not using the tool properly."

I can get the math and all that right on paper....which is ver time consuming, its just making a program to do it fro me ...its beyond me.

Posted: Wed Sep 06, 2006 10:17 am
by evolotion
i made a spreadsheet that did this as a bit of a curiosity, it allowed me to compare power runs. however it didnt have much resultion so you had to perform about 4-5 runs wiht hte same settings before changeing somehthing and running agin. and even then even undulations in the road affected the results more than say a 1 degree change in ignition timing. it was certainly better than nothing for tuneing but far from ideal.

will be purchasing one of these g-tech pro things soon as hav the money spare, they seem to do the job much nicer. and AFAIK the RPM reading is only for plotting the graph the power is measured via accelerometers.

Posted: Thu Sep 07, 2006 8:33 pm
by DARCOM
why doesent some one just add a accelerometer chip to the megasquirt .they sell the chips here ,
http://www.botmag.com/articles/maker_faire.shtml
then right code for it