Gear Display in Analysis

[pault]

I've entered my gear ratios, final drive and rolling circumference
1.778, 1.55, 1.364, 1.167, 3.44 and 1.835m respectively. I have logged speed and RPM, but my gear display in an XY graph is wrong. It mostly shows 1st gear. I'm using all 4 gears. The last lap shows 3rd gear for the whole cool down lap (which could be correct, I can't remember if I was in 3rd or 4th for it).

Any thoughts or ideas?

Thanks
Paul

[mikee boy]

I also have the same kind of problem, but with me it's usually 4th!

I'm keen to get this working properly.

[faraday]

The feature certainly used to work.

What version of Analysis?

What are your selections in Vehicle Options for wheelspeed input?
I'm not sure what this does, but it might be confusing something

I have had successful "gears" for data with wheelspeeds and without. It looks as though changes have been made to the software, but it is premature to blame your problem on a new bug. I will check with my old data in the latest version over the next hour or so...

[pault]

I'm using version 7.3.5 of the analysis software. I don't have a wheel speed sensor hooked up yet. I didn't see a place to add this, so I might have missed some important setting.

I also assumed that I was doing something wrong as opposed to this being a bug.

Thanks,
Paul

[faraday]

My data displays properly.
The wheelspeed selections that live under the 'RPM and Wheelspeed' tab have picked up the channel names that I had previously assigned in the Variable Manager when I have the appropriate .var file loaded. These selections can't be changed within this tab screen in this case.

Thus, I can only conclude that the feature still works as intended. No evidence of a bug...

Do you have a sensible RPM trace?
If the settings that filter revs are not appropriate then you will see revs rise, fall or stay the same when they should be doing otherwise. If the trace is not quite jagged then your settings are probably wrong. I have never seen a trace, in any car or data system, that looked as clean as one would like.
Unless you know you had the foot on the clutch, were in neutral or were smokin' the bags, RPM should obviously rise and fall in a linear relationship with speed.

Did your ratios come from counting teeth?
Don't assume what you read anywhere is correct.

Drive wheel circumference can't be measured statically. The best you can do is roll the car (laden with tyre pressures as it circulates the track) for a few revs.

Set up a spread sheet or at least do a few manual calculations using the values you have entered. Do RPM/speed pairs from your calcs match data point pairs.

It is only when you have wheelspeed sensors that you can properly determine the effective circumference, but it is not a constant. While it changes very slightly with speed and load it also changes with wear and from tyre to tyre. Of course, there is some slip and speed variation between inside and out when cornering. Nevertheless, all these sources of "error" are typically small compared to the way the R-T software can get RPM wrong if you don't have the settings well chosen.

Finally, you can't assume that GPS speed is always correct. When the system is behaving properly speed is as good as spot on. Sometimes erroneous readings are obvious because of spikes or other discontinuities, but I've experienced speed traces with the correct "shape" that are certainly wrong. This has normally been less than 5MPH but occasionally over 10MPH!

[pault]

Gear ratios are correct. I read them off of the gears as I installed them into the gear box at the track.

Using math and tire diameter, I converted the GPS speed to a theoretical RPM. This value matched within a 1% the RPM from the same data point in the analysis tool. My RPM trace is ugly, but not any worse than I would expect and certainly not as nice as I would like. The RPM vs. GPS speed XY graph has 4 distinct lines on it. The lines are "fat" but clearly defined.

I would have expected noise in the gear trace near gear changes or particularly when the clutch is in, but getting just 1st gear for most of each lap is odd.

I'll rerun the graphs and post them.
Thanks,
Paul

[pault]

So I'm not using a wheel speed sensor at this point. I can see that the trick is to set the source for the wheel speed in kph on the "RPM and Wheelspeed" tab of the Vehicle Options window.

What do I set this to?
I've tried setting it to the speed (VAR_0015), but this only gives a value every so often (it peaks up to the correct value then returns to 0 for the next few ticks)

I've also tried setting it to a user defined variable, that is my speed * 1.609, this just gives me a 0. I think this is due to the order of operations (wheelspeed is calculated before the user variable).

So what do I put into that location in the menu to get the gears to work properly?

Thanks,
Paul

[faraday]

So the best you have had is a momentarily correct gear number?

Without a wheel speed sensor, I do better.
The default (for me, it seems) VARs chosen in wheelspeed are the first four frequency channels, which are empty, so they appear to do nothing. They are switched off in the Variable Manager.
I can't change them in Vehicle Options!

The ... = text entry field is empty

I smell a rat

Before you installed 7.3.5, did you have an earlier version installed?
Did you use the Race Tech... > Other > Uninstall utility

[pault]

I did have version 6 installed and at least 2 versions of 7. So perhaps there was a problem with the uninstall.

I did notice that the help files suggest that there are a lot of mathematical functions available for use in the user variables, but I only get about half of them. In particular the IF function wasn't listed.

Maybe it is worth a re-install.

Paul

[pault]

I did a complete uninstall and a re-install. I get the same thing.

I created 4 user variables that are the RPM in each gear calculated from the speed variable (VAR_0015), the gear ratios and a tire circumference. I found a good circumference by trial and error (trying to get the RPMs to match).

I've posted a XY graph here:
http://home.comcast.net/~thedrol/images/gears.jpg

You can clearly see that the RPM trace is pretty close to the RPM in gear traces. The car starts out in 4th, down shift to 2nd, down to 1st for a while, then back to 2nd, 3rd and then 4th.

I didn't show the RPM in 4th trace, so I could show the gear variable from the standard variables.

I have tried this exercise on 2 different PCs, with similar results both running version 7.3.5.

Any new thoughts or ideas?

Thanks,
Paul

[faraday]

How close to steady throttle are the accelerations in 2nd and 4th?

I'll assume that throttle is steady and state that your RPM trace is particularly horrible, but at least it is sensible.
Nevertheless, the above gears are held long enough and the rate of acceleration is not too great, so you would expect that the software would pick up a correct gear at some stage during the accelerations.



There must be a built in tolerance that has to work in favour of a result, even if its an incorrect (by one) result.
I have not applied, or experimented with any smoothing on the gear channel. Have you, perchance?

Your efforts should be rewarded with a better result.

Support... please help

What sort of car/engine/ignition system is this?
I'm going to guess it's a carbie engine so we don't have the opportunity to take an RPM signal from an injector. An alternative means of triggering is worth exploring, but I'm not confident that what you have is the source of the gear problem. Is the gear result any different with less RPM smoothing? If it is, maybe a cleaner RPM signal is the answer after all.

[pault]

Gear is something that is calculated by the analysis software right?

Or do I have to configure my DL-1 with the correct gear ratios before I start logging?

Where the 2nd gear RPM lines up with the RPM I'm switching into 2nd. At that point I'm still decelerating and the throttle is closed. Then the throttle is fully openl as I accelerate to the 1st gear corner. I'm on the brake in 2nd until I shift into 1st. In 1st, I'm modulating the throttle around the this turn. After this turn it is full throttle in 2nd, 3rd and 4th.

I realize the RPM looks bad. The original issue I was looking into was the poor autometer tach readings I was seeing while in the car.

Paul

[faraday]

You are doing nothing wrong, unless something you have done in your exhaustive efforts to find a solution has been wrong.
Uninstall undoes any of that.

The gear calculation you are looking for is purely in Analysis. If you have revs (in the standard RPM channel?) and the ratios, etc are populated in Vehicle Options there should be a result assuming the standard speed result from the internal GPS is there.
If the values are wrong, you will still get a result... albeit incorrect.

What you have pictured tells me your speed is working...
but the result makes no sense. Obviously.

Send me the run file and I'll see what happens here: rotoflow at optusnet dot com dot au.

[pault]

The files were pretty big even zipped so I posted them here:

http://home.comcast.net/~thedrol/TEST__02.zip

I guess I could have shortened the run file to a single lap. The zip file contains a vehicle file, markers and my RPM variables.

Thanks!
Paul

[faraday]

No surprises, unfortunartely Paul.
My analysis is no different to yours.

I played with RPM filtereing over a wide range and this made no difference, so I'm inclined to think that perhaps the gear calculations take place before the RPM input is cleaned up.
Perhaps Support could comment

I note that your logger is running very old firmware. The Run Manager can't tell us the serial number. What is the history of your DL1?

Perhaps a firmware update is possible and will yield dividends.

Can you contact Al Seim for advice?