Hello Vital MTB Visitor,
We’re conducting a survey and would appreciate your input. Your answers will help Vital and the MTB industry better understand what riders like you want. Survey results will be used to recognize top brands. Make your voice heard!
Five lucky people will be selected at random to win a Vital MTB t-shirt.
Thanks in advance,
The Vital MTB Crew
I don’t have the full run in front of me, but based on what’s shown here, I can echo the suggestions from @TheSuspensionLabNZ and @carlinojoevideo.
1) SAG looks fine, but it could probably be improved (playing with air and compression).
2) I’d try better to balance rebound and compression between the front and rear. You can go either way: speed up the front or speed up the rear, depending on which end you liked most.
I’d also recommend trying lower air pressure and compensating for the reduced spring support with more damping (which may include revalving). Running the rebound completely open isn’t ideal; it would be good to revalve the rebound shim stack as well so you can move the adjuster back toward the middle while achieving the same on-track speeds you’re seeing now.
Suggestions:
- Identify one meaningful section of the track and do more runs on it.
- Try to ride consistently and on the same lines.
- Check results after each section and keep track of changes in a tidy manner.
- Change one parameter at a time (but be aware that sometimes 1 click isn’t enough to see a real difference).
- Start by adjusting air pressure, then move to rebound, then compression and volume spacers.
Yes, even without the leverage ratio, you can use WHEEL. This will result in an automatic (more approximate) interpolation, but it should still be close to the real values
Based on all the info you posted, I looked at the other data also, you are close! I have Levo 4 with Cascade link also. This was from Motion Instruments data, the numbers are different than BYB but you can still see the balance in them.
I think the 3 minute trail recording is more the issue. Like Enrico said, try a short section and run that over. If you can find 60 seconds that has a good mix, even if it’s in the middle of that 3 minute run. Just make a start and stop location. Or find a shorter trail so you can do laps faster.
For the data about “total stroke movement” that is a calculation of ever mm your fork or shock is compressed and rebounded. For me, on mellower grade trails, I find I can get them very close and that is very balanced feel. Steep trails and tech sections where you are locked up going down something steep, these will be much different.
Any reason why the position distribution isn't showing full wheel travel on the desktop software?
I think that might be a bug where the axes is labelled in mm but the chart is plotting % values - ive seen something similar, it might be a setting you can change, or related to displaying the rear wheel vs position instead of shock position
I think I am a bit in a corner with my current shock tunes. The only way I can increase my rebound speeds are by increasing my pressure.
I don't know if I can decrease air pressure with the rebound tune I currently have but if I did then I am guessing my compression speed would increase then I could run more compression to slow it down? The fork is already using more travel than the rear and not being able to run faster rebound then I need to
A. Run more compression or
B. higher spring rate, upping the spring rate is going to lower the compression speed, correct?
But my compression speed on my fork is already slower than my rear.
So how do I bring my rear compression speeds down? Adding compression?
I know I need to do one change at a time and currently there might be 2 ways of doing it. I need the fork to ride higher in the travel to get a better overlap. I can either go up in spring rates or go up in compression. I am just a little confused how close I should be shooting for F/R compression speeds. I know there might be some circular logic going on here. Ideally I would get at least a rebound reshim so I could try to increase speeds.
Also I noticed that when I reuploaded my run the speeds on the compression in the rear changed. Whats up with that?
I'm guessing one of them had a leverage file applied and the other was using an average leverage ratio? if you compare the data at the shock they should be the same?
A few more things -
Don't get too hung up on targeting specific numbers, always keep in mind how the bike feels for YOU and what your goals are. Actual peak numbers depend on how consistent each run is, plus different frame kinematics, coil and air springs, tyre pressures and riding style will all affect the peak speeds you see. Plus noise in the data from a loose sensor or interference will cause spikes that make the peak value much higher than you actually achieved, which is why I use the histogram or avg/95% values more than peak speeds
2.5m/s on the rear of a bike like the Levo is very fast by the way - not unusual for a DH bike but when its happening on something with 3/4 of the travel it is a lot faster than I would expect
Thanks for pointing this out.
Looks like the last columns are not properly shown. i checked the other combinations (SUSP/WHEEL - %/mm) and they're fine.
We'll fix it in the next release.
In your position, I would certainly try one approach, then try the second approach as well, to understand how the changes are reacting. Note everything down so you can return to the original point before trying the second option.
If you want to reduce the compression speed on the rear (to match the front ones), I would try by increasing the compression first.
If the spring is correct and delivering the right SAG, I'd go with the compression clicker (HSC and LSC).
@thegromit
About the short trail section, remember that you can always ride the full run (which can be good for the pace) and then cut the section accordingly.
You can do it in 3 ways:
1) Apply a selection from the stack or suspension view > Click the "Sync" button on the desired charts/summary/autoTuning to see the applied selection only
2) Save the new section as a new run, manually (apply a selection > File > Save selection as new.byb file)
3) Save the new section as a new run, using a set of saved GPS waypoints. Create or load your GPS waypoints and click "Apply". The stack and suspension view are automatically resized according to your waypoints (from start to finish line).
Here you can see the stack view, which reflects the selection from the start to finish:
Now you can decide either to do 1) or create a new trimmed run with 2).
The 3) option is the recommended way to compare the exact section between when dealing with different runs.
New member here. I am happy to see that there is an active topic on telemetry with interested people, unlike on other forums or reddit.
For the last couple of months I have been actively working on the open source telemetry system https://github.com/sghctoma/sst
I am very new to this and unfortunately there is little introduction or guidance - a few light on details reviews of commercial systems (MI & Byb) on youtube and couple of MTB sites, a few scattered forum threads and finally I managed to dig out the MotionIQ user guide, which is surprisingly heavy on the science and thinking behind the measurements and graphs. I tried to find the BYB user guide, to see their thinking behind the data processing, but looks like that is customers only.
The original author, sghctoma, has done amazing work and I know that he has been working on improvements, mainly a new desktop app and a way to get travel/suspension data from a bike photo/schematic, similarly to Leverage X3.
Meanwhile I started experimenting with extending the firmware and the existing dashboard to use more sensors. I have added initial support for inertial (accelerometer & gyro) data, in addition to the suspension data. I have been running experiments with an IMU on the fork lowers and another one on the frame. I have coded support for a sensor on the rear axle too, but don't have spare hardware right now to run three sensors. I have added calibration that orients the sensor axis to the bike frame of reference - X is forward, Z is upwards and Y is to the left.
I tried to run some correlations between inertial data and travel or GPS, but didn't get anything much conclusive. Best I got so far was to compare fork vs frame IMU and see how much of the impact is absorbed by the fork.
If anybody wants help to build a system, let me know. The hardware is about $30.
If anybody has some insight about what additional data analysis or processing would be helpful - happy to hear!
This is all very cool!
Are you asking which information on the whole bike is helpful or just IMU?
I’m using 6 axis IMU data with SynBike system and I would love to see the average movements of IMU compared. I always do this manually and it’s time consuming. But information like fork or frame move laterally “this much” would give you a good insight for each setup. Does it have more “grip” or is it moving vertically more because maybe spring rate is too stiff. Endless questions!
Welcome! I have actually followed the project for a year or 2 - its super impressive. I noticed more work happening on it lately too. The dashboard is very slick and hopefulyl will have some more time soon to contribute something to these projects.
As for IMU data - you can use frequency domain analysis to calculate the transmissibility and see how what frequencies are being passed through to the rider, also in motorsport a common use is the "traction circle", G-G plot or Gsum channel which all different ways of combing the inertial values to build a picture of how much grip the vehicle has under braking and cornering. eg combining lateral and longitudinal g to see if you are maximising the amount of grip when turning. A little more complicated on a 2-wheel vehicle that can lean but there could be insights there
I appreciate the inspiration to get creative with mounts. The front uses a brake pad separator with one of the BYB plastic shock mounts for stability & the rear is a tire levers that i cut the spoon off & drilled a hole in. Then on the backside I have a nut to hold it. I'll be using the 3D printers at local library to improve upon this setup.
Question is, measuring shock stroke was 40mm of sensor shaft movement & I set the total sensor shaft at about 45mm from its bottom out, is it necessary to change in the setup the shaft sensor length? I thought I read that the system will self calibrate when you start a run & pick the bike up.
I'll be doing some bench tests later after work but wanted to know from those in the know @enricorodella22 etc, what they do / best practices. Thanks.
Yeah the system will self-zero itself as long as you top out once during the recording, and if you set the shock stroke as 40mm and wheel travel to whatever your bike is, the software will interpolate a linear relationship between the sensor position and wheel position. So the figures won't be 100% accurate, which means you might want to manually measure the curve between sensor movement and wheel movement, or at least factor that in when interpreting the results.
ie if you don't measure the actual curve the average position numbers might show wildly different to an expected number but is actually OK. I would just use relative measurements eg if the average position/dynamic sag reads 25% and feels soft so you add pressure and now reads 22% you can be confident it is riding higher, and got the desired change, but your dynamic sag isn't actually 22%, its jsut stiffer than it was before. make sense?
I have the byb in use for front and rear suspension. Really fun to do laps and compare the different suspension settings. Takes a bit of learning curve to learn how to read all the data and what is the recommended settings etc.
Then again, if you ride with bad technique and your weight is not centered then this will not be fixed with any telemetry device out there.. just saying.
Also when doing laps. You need to record your settings / clicks somewhere.. Unused this https://mtb-tracker.com for storing my suspension settings and tunes and commenting them on how they felt etc. Very easy to use and read
Hello together,
I was a silent reader for the last few months, but I've also been working on my DAQ system mainly improving the filtering of the data (which was more substantial than I thought!) and the display of the data. Unfortunately it's only been raining and snowing for a pretty long time now so I wasn't able to get many "proper" data runs in. I was wondering if I could get some feedback of you guys.
Do you miss anything on these screenshots you consider essential for suspension tuning?
Also since the weather is not really rideable (at least not in a fun way) I was also thinking If I could get some feedback or setup improvement ideas on my latest data run. This was the best bike feel I was able to get till now.
Bike: Santa Cruz Heckler SL
Fork: DVO Diamond D1 OTT
Shock: Rockshox Vivid Air Ultimate
Weight: 85kg
"Final" Settings Fork: 125 PSI, 0 Tokens (Not available with the OTT system I think, LSR -14 from full slow (Of 22 available), LSC and HSC fully open
Static Sag: 20%
DVO reccomendations: 125 PSI, LSR between -10 and -14, LSC fully open, HSC 2-3 (Of 5 available)
"Final" Settings Shock: 230 PSI, 2 tokens, LSR -6, middle +1, HSC middle, HBO middle
Static Sag: 30% (SC recommends 29-30%)
RS reccomendations: 235 PSI, (no reccomendation for tokens), LSR -10, LSC middle, HSC middle, HBO middle
I started with the recommended settings and ended up with the ones above. I mainly slowed down the rebound of the shock a lot, because the data showed that it was always faster than the fork. The data of the screenshots are from a run with the "Final" settings in slippery/wet conditions so about 70% commitment. I'm still learning a lot about the topic so please tell me if I interpreted something wrong 🙂
I also got this info by Santa Cruz:
You’re right that the Heckler’s high leverage ratio can make the rear very active, and seeing a dynamic sag around 39 % when you’re targeting 29–31 % static sag is not unusual. We’ve seen similar numbers in testing – the bike does sit deep in its travel, which actually contributes to that playful, plush feel the Heckler is known for.
In terms of dynamic sag for a balanced setup, most riders find that 35–38 % tends to be the sweet spot. It keeps the bike active over small bumps while still giving you support for bigger hits. Of course, the “perfect” setup depends on your riding style, terrain, and even the suspension tokens or volume spacers you might be using.
EDIT: In case the screenshots are not really readable in the forum, here is a link to the raw screenshots :-)
https://drive.google.com/drive/folders/1PFgAd1Lg4W3NmrtJIczGKMBRAwMmi1_…
Cheers!
that data system looks great and I like the charts. I know there is not a right or wrong but 35-37% dynamic sag at the rear axle to me seems very deep to me. I don’t know the steepness of the trail you are on but 27 up front and 37 rear axle seems like you would be really slacked out.
Why is there so many more comp/reb strokes on the front than the rear? The total movement was close but that caught my eye.
Thank you, glad to hear :-)
I was thinking the same thing about the dynamic sag, but when I did setup the bike so the rear dynamic sag matches the front, I was a whopping 40psi over what Rockshox or Santa Cruz recommended and the stiffer rear did not feel good. That's why I asked them and got this answer.
You’re right that the Heckler’s high leverage ratio can make the rear very active, and seeing a dynamic sag around 39 % when you’re targeting 29–31 % static sag is not unusual. We’ve seen similar numbers in testing – the bike does sit deep in its travel, which actually contributes to that playful, plush feel the Heckler is known for.
In terms of dynamic sag for a balanced setup, most riders find that 35–38 % tends to be the sweet spot. It keeps the bike active over small bumps while still giving you support for bigger hits. Of course, the “perfect” setup depends on your riding style, terrain, and even the suspension tokens or volume spacers you might be using.
Which is where I am now. But that also means that the position distribution balance in the data looks pretty off since the rear basically lives ~10% deeper than the front😅 The test trail isn't very steep.
Good point about the amount of strokes. I think thats because the rear gets more super short strokes that get filtered out, but I'll check this.
Hi everyone,
As I am looking at new projects for this year and will have some free time in the next month I stumbled across this thread. Suspension setup has been an exiting topic for some time now, but also kind of a big mystery when you don’t have any data to base your feelings/findings off (maybe it’s also my engineering background needing data to verify
). I can’t justify buying e.g. a BYB at my level of riding as the cost to benefit ratio would be absurd and would also take away a great learning experience.
Do you have any recommendations on how to diy suspension telemetry? And I mean for everything, hardware, software, setup, instructions, etc. The cost doesn’t have to be the main focus, it’s more about building a system without being a suspension/data expert from the get go.
What year heckler are you riding? What travel is that front and rear? Not saying you’re wrong, just wild setup, maybe it’s a bandaid fix for bad leverage ratio.
My 2 cents, you will spend so much more time and money trying to make a system vs the least expensive BYB system. You can spend all that time testing and setting up your bike. The software seems like the most difficult part to make. Look for a data system if you are trying to save money.
I'm riding the Heckler SL Model year 2024.
It has 160mm in the front and 150mm in the back.
The beginning leverage ratio is very high on this bike (Around 3.25). I haven't seen any other leverage ratio this high yet.
Hi and welcome to this thread :-)
I totally know what you mean, I also have an engineering background and love to tinker with electronics and software. Also my riding level is for sure not as good as many others here in this thread, so I like having data to check if what my "Bum-O-Meter" felt could be correct ;-)
As carlino said, you'll spend an awful lot of time engineering this, but you'll also learn a lot through it. I know it because I did exactly the same thing ;-) It will be cheaper if you don't think of working hours as money.
I'm still working on mine, further improving it and adding more stuff to it. Maybe it will become a comercial product, maybe an open source project, I don't know yet.
I'll send you a DM :-)
The sufni system linked above by @moridinbg is one of the best DIY systems I've come across that includes really good documentation on how to build it yourself - the parts used are very common and relatively cheap (most can be bought from places like ali epxress). It includes the software side which is definitely the hardest thing, so they have instructions on how to build it yourself - that part might be pretty daunting so maybe have a go at that first and see how it works out for you.
Alternatively, learning Python is super fun and extremely useful - One of the easier languages to learn and super versatile. You can also get Claude AI to do the software side for you (both the logger & analysis parts) but I would highly recommend getting a grasp of it on your own as there will inevitably be debugging to do yourself.
Lastly, building the logger is probably the easy part - you can get resonably plug-and play parts from adafruit, sparkfun or seedstudio that even use minimal soldering (with qwiic/stemma QT conectors). Reading sensors and logging to a csv is pretty simple - again Claude is really really good at writing the code there as you can just tell it you have sensors A, B and C connected to pins X, Y, and Z and you want it logged to a text file with timestamp, sensors A, B, C at whatever sample rate and it will make that in no time! Then just analyse the data in a spreadsheet to make that side a bit simpler
That leverage ratio isn’t very wild to me. About 27% progression. What shock are you running?
Not very wild, but the starting ratio seems pretty high compared to other bikes.
I tried different shocks (Originally the Rockshox SuperDeluxe, then the Rockshox SuperDeluxe Coil) but currently it has a Rockshox Vivid Air Ultimate. The data is with that shock.
Ahhh the vivid air with that leverage ratio can be the issue. That shock has a big negative chamber. So you might be adding more and more air to get to a smaller sag number and then that make the rest of the travel super stiff and ramp even more. That shock would be better on a linear bike. Coil is probably your best bet on that leverage. If you are looking to keep progressing your set up. Not that you have, just 2 cents! Stoked you are grinding out there with data!
Well I did try the Super Deluxe Coil Ultimate...First with a 500lbs spring, then a 550 and ended up with a 600 lbs in the end to achieve about 32% static SAG. This made me feel fat, but believe me I checked my weight 😂 The problem with that setup was actually too little ramp up. The rear felt super plush and good over rocky/rooty stuff but blew through the travel on small jump landings. Haven't tried a progressive spring or springdex yet. But especially the progressive spring kinda kills one of the main benefits of a coil shock IMO 😅
I think I'll just have to wait until the ground here gets rideable again without the super slippery mix of mud,snow,ice and leaves 😅 And then get some more data and experiment some more with different setup approaches 🫡 Good thing we all love tinkering here!
Data system on Factory Honda Supercross bike this weekend in San Diego. I love that rear shock pot mount. They would be using a leverage ratio equation from pot to figure out rear axle movement.
Post a reply to: Suspension Data Acquisition