AIM Pista Question "G's"

RSS
Email

0 Kudo

Author	Comment
abuldo	AIM Pista Question "G's"		Lead	[-]
Posts: 39 02/23/09 10:18:04	Tags : None Question on the AIM Pista Unit. I was looking to create plot of Laterial and Long G vs combined G's. In looking at the data, it looks like we only have a single axis G and it appears to be lateral. Do I need to buy a second accelerometer? I had a pro look at the data and he said not a problem..." okay use the speed in a math channel to get longitudinal acceleration." However the speed signal is a little jiggly and for some reason on my computer I cannot get the filtering to smooth the signal (Which may also expain why gear calc graph is horrible - Instead of clean U Shaped Graph, I get some of the U's but lots of spikes up and down - even when I am not shifting). Anyway the thought was use a math channel to create the long G and now use a math channel to create the combined G. Ambrose Buldo - Abuldo@aol.com
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
David Ferguson		#1	[url]	[-]
Posts: 2237 02/23/09 22:05:00	So, I'm not sure what your question is? The answer to "do you need to buy a second accelerometer?" is: no. Are you asking for the formula for calculating acceleration from velocity (acceleration = derivative(velocity) ; but don't forget the units factors ...) AIM has this function as standard in their default maths -- it's called "AIM_Linear_Acc_2" if your speed is in MPH. Try clicking the "general" tab under Modify->Math Channels, and using the "Default AIM Set" button if you don't see the "AIM_Linear_Acc_2" function. If your speed signal is "jiggly", then perhaps you are logging it at too high of a rate, or too slow of a rate -- or you have it attached to a wheel that locks or spins... You might try taking a screen shot, and posting here so we can see what you are describing.
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
David Ferguson		#2	[url]	[-]
Posts: 2238 02/23/09 22:08:07	Also, the default AIM Math set has an "AIM_G_Sum" channel (this is your combined G's), however you will need to modify it to use the "AIM_Linear_Acc_2" function since it's made for "AccLat" and "AccLong"...
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
CDS Inc.dsrforum		#3	[url]	[-]
Posts: 5 02/25/09 09:17:13	My experience of 23 years in racing data acquistion is that taking the derivitive of speed to come up with Longitudinal G force is a very poor substitute for measuring it directly. This was always CDS's philosophy even when accelerometers were expensive. Now that they have come down so low in cost it mystifys me as to why some still cling to this method. Sure you can smooth the source (speed) and smooth the result (Long_G) enough that most of the time you get something that looks halfway decent, but in the process of doing so you loose so much of the detail that it is hardly worth doing IMO. If you want quality longitudinal data do yourself a favor and buy a single axis acelerometer to measure it.
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
George R Main		#4	[url]	[-]
Posts: 137 02/27/09 23:22:58	To add to Peter Kelly's statement, Long G is as "pure" of a measurement as any other. The concept of calculated Long G started with PI systems, as they tried to compete with lesser priced systems in the market place in the earlier 90's, when a two axis accellerometer was over a thousand dolllars, as Peter stated. The calculation was formed to replace the G sensor, as a cheap way of replacing a G meter. However it also involves enough "additives" and variables to the signal to make it as accurate as an engine temp sensor. With wheel lock, lifting a wheel off the ground or a suffering tire grip, the signal is by no means level with the actual measurement of G. If braking performance isn't important to you, then it will work fine for you. I would put the measurement on the same plane as calculating steering input by using the lateral g signal solely or engine rpm by calculating the wheel speed (who cares about tire growth). There's enough of a loss of information that as a driver you can't move your braking point by sourcing your info based on the calculated input, that is without knowing that you won't exit the track based on the availiable information. I have seen 3 G's of braking on calculated signals on a FF, and the driver said he didn't lock the wheels, the same car with actual Long G when measured never got over 1.5 g in the same type of corner........3g in braking is what a F1 car sees not a FF....
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
abuldo		#5	[url]	[-]
Posts: 40 03/02/09 14:24:55	I guess I got confused with my own question. What I want to build are 2 handeling charts: 1. Steering on X (Buying a steering sensor from Ellen) and Lat G on Y. If I read this correctly, looks like I'll be abe to do this without any investment other then the steering senors. 2. Classic Friction Circle with lat G on X and Log G on Y. Based on Daves Comment, looks lke I'll be able to do this if I use AIM_Linear_Acc_2 I eventually want to measure ackerman and use that to quantify Over/understeer. I don't have those notes handy, but I;ll deal wiht that after I get the two above charts built. Ambrose Buldo - Abuldo@aol.com
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
David Ferguson		#6	[url]	[-]
Posts: 2239 03/02/09 16:24:40	You've got it right, but as our friends from CDS point out, using the calculated G's isn't as accurate as having a separate Longitudinal G sensor. But you can certainly try it with what you have to see if the output and graphs will be useful to you -- just be aware of the accuracy limitations.
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote
George R Main		#7	[url]	[-]
Posts: 138 03/03/09 07:43:01	abuldo wrote: I guess I got confused with my own question. What I want to build are 2 handeling charts: 1. Steering on X (Buying a steering sensor from Ellen) and Lat G on Y. If I read this correctly, looks like I'll be abe to do this without any investment other then the steering senors. 2. Classic Friction Circle with lat G on X and Log G on Y. Based on Daves Comment, looks lke I'll be able to do this if I use AIM_Linear_Acc_2 I eventually want to measure ackerman and use that to quantify Over/understeer. I don't have those notes handy, but I;ll deal wiht that after I get the two above charts built. Measuring ackerman won't achieve a handling chart nor will it quantify over/understeer, though I think you mean measuring the steering path, which is a signal you will need to create in Aim. It is a signal equation that is already present in CDS (you just fill in the associated numbers and calibrations) and in the more expensive Motec systems. When defining a car's handling, oversteer is the easiest to see and understeer is the hardest to define. Though the "fringes" of them both are the most difficult in terms of analysis. I know that Peter at CDS spent a long time getting the right math combinations to come up with a handling graph that was usefull and easy to read and highly accurate. So good luck on your exploration. You will need to invest in a steering sensor to achieve what your looking for as the G vs G graph don't provide enough information and will be only a fraction of definition of handling. The best definition of handling always includes every movement of the driver and can be only achieved on a reliable scale with having the following six sensors on board, measuring the movements of the driver @ 20 samples per second (and never below that rate) Lateral AND Long G, Steering, throttle, Speed and engine Rpm. Though throttle/engine RPM isn't normally inclusive in a handling graph, they are vital to accurate analysis of the handling of the car and shouldn't be forgotten...
Interact My Recent Posts Message Me Connection Invite Blocking Ignore User's Posts Report Post	Reply Quote

<< Previous Topic Next Topic >>

Yuku free message boards

AIM Pista Question "G's"

AIM Pista Question "G's"

Quick Reply