Mountain Bike Reviews Forum banner

1 - 20 of 20 Posts

·
Registered
Joined
·
512 Posts
Discussion Starter #1
Anyone ever done this? I am wondering how valid this is as a way of suspension tuning.

an accelerometer would show how many oscillation cycles it takes to stabilize over a curb and the intensity of the bike movement. But I have no idea how that would translate to the trail. But it should easily show the difference between one or two clicks on the damper...

It’s not talked about, so I guess it doesn’t work? Why?
 

·
Ride More - Suffer Less
Joined
·
2,576 Posts
How about a linear potentiometer ?

 

·
Registered
Joined
·
13 Posts
an accelerometer would show how many oscillation cycles it takes to stabilize over a curb and the intensity of the bike movement. But I have no idea how that would translate to the trail. But it should easily show the difference between one or two clicks on the damper...
Yes I know of at least 1 major brand that has used accelerometers to develop bikes with great success in terms of sales and racing.

As with anything it's just one tool inside your whole toolbox of analysis methods. Used properly with other things it can make a great bike.

Looking at individual acceleration vs. time traces from trail data is NOT useful whatsoever in my experience. It is not the pretty smooth line you are thinking of, even when doing isolated tests. In my experience, some kind of summary statistic or performance parameter that can be computed from accelerometer data is much more useful.

The auto industry calls this NVH (noise, vibration, and harshness). Read up! there are many, many books/iso standards about it.

In terms of suspension tuning for an individual, average rider, I don't think data acquisition is a useful tool, as the real magic of data aq is what data you chose to record and the analysis of that data.
 

·
Registered
Joined
·
952 Posts
Ive borrowed and liked friends Shockwiz. Uses shock air pressure changes and accelerometer to give you computed suggestions. Doesnt give you raw data aside from air/jump time and deep event and some other statistics.
 

·
Registered
Joined
·
1,682 Posts
Anyone ever done this? I am wondering how valid this is as a way of suspension tuning.

an accelerometer would show how many oscillation cycles it takes to stabilize over a curb and the intensity of the bike movement. But I have no idea how that would translate to the trail. But it should easily show the difference between one or two clicks on the damper...

It’s not talked about, so I guess it doesn’t work? Why?
Accelerometers are certainly useful for suspension, I can only assume they aren't used more due to a lack of understanding (like Data Acquisition in general) but the Stendec system is built around them, so its not completely unheard of.

It can be tricky to get a clean signal at a high enough sample rate, plus it takes a little bit more processing to get something meaningful out of the data. eg if you are just looking at acceleration values at the bar they are proportional to the speed of the trail so won't tell you a lot without something to compare to. A sensor on the hub and a sensor on the bike at both ends is what you really need and then you can start to get useful information

They do have advantages like low weight, small size and no moving parts
 

·
Registered
Joined
·
134 Posts
One data stream by itself can't tell you much. One of the best ways to capture lots of data is video.
For the average Joe, can't beat the visual that a video gives. You can obviously see if your riding at 50% dynamic sag, bottoming, etc...

Sure data acq systems are awesome for the initiated, but if you don't understand the data and more importantly what it is suggesting for you to change, it's useless.

I think an important part of potentiometers is that they can isolate suspension movement from bike movement. I think that would be hard to do with an accelerometor.

I think two potentiometers on the bike and an accelerometor on the rider would give interesting data

Sent from my moto g power using Tapatalk
 

·
Registered
Joined
·
9,730 Posts
For the average Joe, can't beat the visual that a video gives. You can obviously see if your riding at 50% dynamic sag, bottoming, etc...

Sure data acq systems are awesome for the initiated, but if you don't understand the data and more importantly what it is suggesting for you to change, it's useless.

I think an important part of potentiometers is that they can isolate suspension movement from bike movement. I think that would be hard to do with an accelerometor.

I think two potentiometers on the bike and an accelerometor on the rider would give interesting data

Sent from my moto g power using Tapatalk
The pros use video paired with data acquisition. It's a vital part of analysis. There are plenty of enthusiasts out there generating lots of useless data and drawing the wrong conclusions.
 

·
Registered
Joined
·
4 Posts
Interesting topic. Since December I have been playing around with professional grade accelerometers and data logger . So far I have been running hub and body accelerometer on the front fork to try and quantify the system behavior through frequency analysis.
One of the main limitations is the trail bike I'm running it on while waiting for a proper enduro.
The main thing that is clearly visible is the differences in average acceleration levels with different settings (rebound damping, air pressure, lock-out).
Some difficulties in describing the behavior is the non-linearity or progressiveness of the air spring.
I therefore think that a combination of linear potentiometer with the accelerometers would help in tuning of the system.

This is also the next step when building up a system for the Enduro.
 

·
Registered
Joined
·
4 Posts
I've not done any of this, but it seems you'd want a pair of accelerometers to compare the data on the fork crown and lower tubes (or the rear wheel analogy). Then tune for the smallest difference.
What do you mean by smallest difference?
You want the hub/lower tubes and fork crown to move in phase and with the same amplitude? Then the difference is zero...

That's what my racing cycle does when you don't count in the compliance 😁😁😁
 

·
Registered
Joined
·
239 Posts
What do you mean by smallest difference?
You want the hub/lower tubes and fork crown to move in phase and with the same amplitude? Then the difference is zero...

That's what my racing cycle does when you don't count in the compliance 😁😁😁
I edited out the "smallest difference" part in my earlier post - that made little sense. My idea for using two accelerometers is that you'd be able to compare the input and the transmitted accelerations.
 

·
Registered
Joined
·
4 Posts
I edited out the "smallest difference" part in my earlier post - that made little sense. My idea for using two accelerometers is that you'd be able to compare the input and the transmitted accelerations.
Definitely, the transmisability was about 90% on my fork up to roughly 30 hz or so with zero phase, after that there was a small phase delay up to 20 degrees or so.
But this analysis is only valid up to a certain level as the system is non linear and trying to get the transmisability and phase would be treating it as a linear system first of all.
1919369
1919370
 

·
Registered
Joined
·
1,682 Posts
Definitely, the transmisability was about 90% on my fork up to roughly 30 hz or so with zero phase, after that there was a small phase delay up to 20 degrees or so.
But this analysis is only valid up to a certain level as the system is non linear and trying to get the transmisability and phase would be treating it as a linear system first of all.
View attachment 1919369 View attachment 1919370
nice work! What are the specs of the accelerometers you're using?
 

·
Registered
Joined
·
4 Posts
They are from a manufacturer called Entran, the EGCS series with a range of - 30/+30G.
These were used by one of the F1 Team, after which I acquired them for some race car projects.
Since they were still laying around I put them on the bike together with a 2D data logger. But like said, since it was only a test on the trail bike I did not want to invest too much money into a decent wiring loom. I will build that one once I can get my hands on a proper Enduro.

The interesting part is, I can't remember if somebody mentioned that here or elsewhere, that you can't use a similar approach as cars as you cant neglect the geometry changes. Due to the changing fork inclination angle you are also measuring in a different plane, perhaps I will change this when I get front and rear damper Potis to back calculate the geometry of the bike.
Still there are a lot of similarities and a lot of useful techniques that apply to the bicycle (as well as some stuff from motorcycle racing).

Don't get me wrong, I'm doing this for fun besides my job at a motorsport manufacturer. I know the scope of the research at the moment completely outweighs the results I will achieve (which in general are talent limited, not material).
But like said, it's just a hobby...

I would appreciate any input or any information to what could be important, or important parameters for for example road holding/following and rider feedback from the suspension.
 

·
Registered
Joined
·
1,682 Posts
For the most part I think using a similar approach to cars, especially if you are working in the frequency domain. It is a good point about fork angle, but lean angle could be another factor so if you have a method for correcting this it wouldn't hurt.

I've been testing mostly with damper pots and a handlebar accelerometer and its definitely useful to have the comparison with suspension movement. I'm currently building the mounts and wiring to add sensors to the hub too

There really isn't much data out there yet but there has been a few papers published on the topic so I can DM a link to those for you which has some useful information
 

·
Registered
Joined
·
77 Posts
I love geeking out over this stuff. Recently had the opportunity to play with a linear pot and map that data to some video:
. Not sure what's up with the sound - but you can see the big compressions. I think I have red as highspeed and blue as low-speed green as in-between - the grayish/dark grey area split represents 'sag.' Utilizing the video really helps identify what's going on. Would love to have an accel on the bars or headtube area to compare with this data along with a setup for the rear. fun stuff!
 

·
Registered
Joined
·
1,682 Posts
I love geeking out over this stuff. Recently had the opportunity to play with a linear pot and map that data to some video:
. Not sure what's up with the sound - but you can see the big compressions. I think I have red as highspeed and blue as low-speed green as in-between - the grayish/dark grey area split represents 'sag.' Utilizing the video really helps identify what's going on. Would love to have an accel on the bars or headtube area to compare with this data along with a setup for the rear. fun stuff!
Nice one, what are you using to do the overlays? I’ve played with a couple of programs, they do the job but were a bit clunky
 

·
Registered
Joined
·
77 Posts
Nice one, what are you using to do the overlays? I’ve played with a couple of programs, they do the job but were a bit clunky
I'm a web developer by trade – so I used what I knew and wrote some custom graphing utilizing PixiJS which allows the browser to render with the GPU. Then did a screen record. Hoping to free up some more time to make this a bit easier to get the video output.
 
1 - 20 of 20 Posts
Top