New project?

[MurrayPeterson]

Cameras usually use software to move the captured frame around inside the raw pixel sensor, but there are some lenses that have mechanical anti-vibration mechanisms.

Drones have a huge selection of anti-vibration systems as well.  e.g. 
https://hobbyking.com/en_us/vibration-and-shock-absorbing-mounts-400g-a4plus4-100g.html

[MurrayPeterson]

So a bit of smoothing and low pass filtering does wonders for the results:
https://drive.google.com/open?id=0ByeFbiNdvLbGWnAzQkVxMEtEVTQ

[Reijo]

Interesting.

[majormojo]

Murray, I've done some work on a lap timer app (iOS/Android). My use case is for endurance racing but I've dealt with some of the same issues with sensor and GPS data. If you haven't already, take a look at the Kalman data smoothing filter for time-series data. I found it useful and incorporated it into my app. There are well-tested implementations of the algorithm freely available for many different languages.

https://en.wikipedia.org/wiki/Kalman_filter

My experience with GPS data is that it's quite good for calculating speed and acceleration as that relies on calculations requiring only changes in relative co-ordinates over time. Where it starts to fail is when doing driving line analysis as that requires comparison of absolute position co-ordinates. Then you have to start worrying about all kinds of other issues like drift, sat signal strength, sat count, etc.

[MurrayPeterson]

Thanks Jim -- I'll look into that filter.  My biggest issue so far has been the sheer sensitivity of the sensors, especially acceleration.  That sucker will actually show the effect of me turning up the bass on my stereo, so road and engine vibrations can drive it crazy.  Thankfully, the unit has a pretty good DMP processor on board, so I can do some nice stuff before I even see the data.

[MurrayPeterson]

So here is some fairly readable data from one of my runs on Sunday:
https://drive.google.com/open?id=0ByeFbiNdvLbGaGpUbDNtcHJVVm8

I was amazed that I was hitting a bit more than 1G on the sweepers -- those tires sure handle the cold.

[zero10]

So here is some fairly readable data from one of my runs on Sunday:
https://drive.google.com/open?id=0ByeFbiNdvLbGaGpUbDNtcHJVVm8

I was amazed that I was hitting a bit more than 1G on the sweepers -- those tires sure handle the cold.

Which tires were you on?

Also, interesting to see the slalom migrate, the first right hander was 0.8g, then 0.72g then 0.7g, while the lefts went higher.  Is that the tires rolling under the rim and the car starting to hop I see at 17-20s?
Looks like there really wasn't a lot of room to brake in a straight line, but it's neat to see how consistent you are getting hard on the brakes then easing off as you dial in the steering.

[MurrayPeterson]

I was on new, unscrubbed RE-71Rs.

The slalom acceleration shows a driving error -- I have other runs that look like perfect sine waves, including the throttle and car rotation rate (gryo).

There were no scuff marks that even came close to going off the tread blocks, so I don't think the tires were rolling over all that far.  However, that rough plateau was definitely where the car had transitioned into understeer.  I need to sort out tire pressures and sway bar settings to eliminate that problem, since it is costing me some serious time.

[PedalFaster]

that rough plateau was definitely where the car had transitioned into understeer.  I need to sort out tire pressures and sway bar settings to eliminate that problem, since it is costing me some serious time.

Somewhat off-topic, but I wonder if what's happening there is that you're hitting your front bump stops, which is in turn "shocking" your tires and causing them to break loose.

[MurrayPeterson]

Somewhat off-topic, but I wonder if what's happening there is that you're hitting your front bump stops, which is in turn "shocking" your tires and causing them to break loose.

So that would suggest that a stiffer sway bar setting would work?  I am soooo confused

[Reijo]

I was getting understeer (as well as some other spots) there also but I felt that was more because of the changing off-camber of the pavement that seemed to induce it.  For me it was more that the front winters were more worn than the rears ... both of which were old ... Just thinking that your front tires hit that first and then your rear ... sort of like going off a jump.

How was the car elsewhere?   e.g. understeer-wise?


R

[MurrayPeterson]

Car was perfect except for that understeer.  However, it happened on every sweeper, independent of camber.  It might have been due to cold pavement and tires, but it did seem to be consistent.

[PedalFaster]

Somewhat off-topic, but I wonder if what's happening there is that you're hitting your front bump stops, which is in turn "shocking" your tires and causing them to break loose.

So that would suggest that a stiffer sway bar setting would work?  I am soooo confused

Maybe? Normally a stiffer bar would induce more push, but if it kept the front off of the bump stops, and hitting the stops is really what's inducing the death push currently, then the stiffer setting might still result in a net improvement in front-end grip.

Only one way to know for sure: test.

[Terry Johns]

Car was perfect except for that understeer.  However, it happened on every sweeper, independent of camber.  It might have been due to cold pavement and tires, but it did seem to be consistent.

As I mentioned Murray back when you got the car, fitting a big front sway bay on the Miata is a backwards step, you give away all the front end grip and as you've experienced the car washes out. I would suggest you refit the stock front bar and fit a 17.5mm rear bar, run the rear dampers as little softer and then get a trick alignment that compensates for the rear being a little loose. My NC is almost totally neutral, all most zero understear. Did you uprate the bump stops, very important on a Miata. The car has a huge amount of body role, trying  to eliminate the body with the shocks will only take away grip, the car needs to role and lean on the bumpstops as it creates massif weight transfer to the weighted side. As you have discovered this isn't an S2000. No need to reinvent the wheel here lots of stuff on line.

Conventional wisdom says FWD Big Front, AWD Big rear bar. The Miata is none conventional due to its weight distribution and its softly sprung. Stiff dampers and a big bar in most cases are a backward step in a softly sprung car. Once you go down the road of fitting stiffer springs everything changes and stiff  shocks, are needed to control the springs, the Koni set is ok but in reality its only a compromises as the dampers at full stiff overpower the springs, which is why the need to be run 1/2 hard. tire pressure needs depending on temperatures 30F - 32R, soften the rear if the rotation is to aggressive.  The alignment is one of the most important mods on the Miata.

Base line: Front camber 2.3, rear 2. rear tow, zero, front tow, just a tad out to sharpen up the turn in. Mild rotation is always better as you can keep your foot in it, and it usually points you in the right direction. I played around with my car for over a year, its now just right for my driving style.

Good luck Murray, I'm sure once you got it sorted you'll be kicking our arse.

This advice was given to me by Chris Harvey. US National winner.

[Terry Johns]

Somewhat off-topic, but I wonder if what's happening there is that you're hitting your front bump stops, which is in turn "shocking" your tires and causing them to break loose.

So that would suggest that a stiffer sway bar setting would work?  I am soooo confused

Sorry Stephen with all respect that's nonsense LOL