Hand dynos are more or less intended for testing the bleed on a shock after reassembly. When talking about the machines, there are different kinds of dynos...
Hand dynos are more or less intended for testing the bleed on a shock after reassembly.
When talking about the machines, there are different kinds of dynos for sure. Cheaper ones (~10k USD) tend to go to shaft speeds of ~2.5m/s. If you want really high shaft speeds of about 7m/s, those things get expensive fast. Like ~80k USD expensive and upwards.
Thanks for the info!
That raises another question, which is how fast shaft velocities can be while mountain biking? Greater than 2.5 meters/second? Greater than 7.5?
That raises another question, which is how fast shaft velocities can be while mountain biking? Greater than 2.5 meters/second? Greater than 7.5?
Forks frequently see shaft speeds of 4m/s and in general up to about 8m/s (see this video by Vorsprung). Rear shocks see quite a lot slower shaft speeds in general due to two reasons: Leverage ratio and often not being the first in line for the hit.
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the...
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the typical dyno test look like?
Is a dyno test always necessarily an automated machine process, or can it be reliably performed manually in a hand dyno?
Are there different types of dyno machines, or is every suspension brand and independent tuner pretty much using the same machine? How much do dyno test protocols vary between suspension brands and tuning shops? And if there are different machines, is data from one machine or test protocol comparable to data from another?
What kinds of information does a dyno test actually meaningfully provide, and what kinds of info can that test data never tell you? It seems like there isn't a 1:1 correlation between what the machine says about a fork or shock vs. what riders experience on the trail, meaning that real on-trail testing is still essential and you can't truly product test damper tunes from start to finish in a lab.
I'm betting this has already been answered somewhere in our vast internet, so if you links to share, that's great too.
Hand dyno is basically just jerking off a shock. Need a load cell and velocity sensor to be able to generate graphs.
The big flaw of the shock/damper dyno is it only tells you what it tells you, it's isolating things. So like in a real suspension system you have the chassis and the spring and those things interact in complex ways. If you watch video tours of fox R&D they seem to maybe dyno whole forks, but looks to be purely vertically which eliminates a lot of the side loading that happens in real life (actually loading the bushings etc).
All damping is, from a physics perspective, is an energy bleed. It's bleeding off energy from the system (which could be bike+rider on single bump event). Friction does the same thing, so quantifying what your damper is doing is great, but it's only one part of the picture of damping happening in the system, and depending on the scenario, the main source of damping isn't even the damper. Which is not a good thing as the damper provides the most controllable form of damping.
Additionally in something like a fork, you have an imperfect lubricaton system, where the fork is trying to pump oil up to the foam rings and bushings, and then the action of the fork is also pushing oil out of the bushings, where it's needed most. This is something that I think fox and rockshox have made a lot of progress with over the last 6 years or so. It used to be that turning your fork upside down before a ride made a massive difference. That seemed to kind of go away for me when fox introduced the cast oil channels into the lower legs. Remember when WC mechanics would talk about doing a fork rebuild for practically every run? Well if your fork lubrication system is awful, it makes a lot of sense...
But all the complex chassis stuff isn't really tunable, and it's real easy to turn knobs or swap shims. So we end up talking about knobs and shims and PSIs.
More sophisticated test setups/machines can probably go a long way to evaluating something like fork performance more holistically, but tiny bicycle industry is tiny and this stuff is actually expensive. This is why, in my opinion, we get things like fox releasing a new fork and being able to honestly say that it fixed the old fork "not using its travel" aka not being a fork (first gen 34 sc anyone?). They've also had to boil the frog (us) into accepting weight penalties for performance. In the past you wanted the light fork cause.. well it didn't work anyway. Then you get into the whole question of: looking at all these charts, what's better? Now some of the chat at my LBS is: the forks are so good now, what does this mean for riding if these trails aren't gnarly anymore?
Hand dynos are more or less intended for testing the bleed on a shock after reassembly. When talking about the machines, there are different kinds of dynos...
Hand dynos are more or less intended for testing the bleed on a shock after reassembly.
When talking about the machines, there are different kinds of dynos for sure. Cheaper ones (~10k USD) tend to go to shaft speeds of ~2.5m/s. If you want really high shaft speeds of about 7m/s, those things get expensive fast. Like ~80k USD expensive and upwards.
That raises another question, which is how fast shaft velocities can be while mountain biking? Greater than 2.5 meters/second? Greater than 7.5?
A 5 foot huck to flat could probably get you to an inital shaft speed of around 5m/s or 5000mm/s. Simple equation: time to drop = (2*height/g)^(1/2) so sqrt of 2*5/32 = .56s Speed at impact -> speed = time * acceleration or .56 g aka .56*32=17.92ft/s. You can do some more math to figure out how landing slope angle and head angle affect shaft speed, but it's not that important.
If a huck to flat gets you to 5m/s i bet you can find bumps that do more.
Another limitation of dyno charts: spiking occurs where exactly? Oh, off the chart? Cool. 😂
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the...
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the typical dyno test look like?
Is a dyno test always necessarily an automated machine process, or can it be reliably performed manually in a hand dyno?
Are there different types of dyno machines, or is every suspension brand and independent tuner pretty much using the same machine? How much do dyno test protocols vary between suspension brands and tuning shops? And if there are different machines, is data from one machine or test protocol comparable to data from another?
What kinds of information does a dyno test actually meaningfully provide, and what kinds of info can that test data never tell you? It seems like there isn't a 1:1 correlation between what the machine says about a fork or shock vs. what riders experience on the trail, meaning that real on-trail testing is still essential and you can't truly product test damper tunes from start to finish in a lab.
I'm betting this has already been answered somewhere in our vast internet, so if you links to share, that's great too.
The machine is basically a motor that turns a shaft, which is turned in to linear up and down motion. The software runs the machine faster and faster, and measures how much force the damper is producing at every point and speed in the stroke. My machine is a CTW "rotary" type dyno with 2 HP and one of the more basic, lower power machines on the market (I think about $15,000 USD today). CTW was born out of a company called Roehrig who pioneered workshop damper testing in the early 90's - many of the older racecar and motorcycle tuners still use Roehrig Dynos today, along with mtb companies like Push and Vorsprung. They are the most experienced and knowledgeable when it comes to building damper dynos but there are other brands out there, Laba 7 is a recent one pushing hard, mechatronics labs is smaller but seem to make very nice machines. There are 2 main types-
Rotary Dynos like mine use a motor to drive either a crankshaft or a scotch yoke. Crankshafts are not very common anymore because like a piston in a car, the damper is accelerating faster at the bottom half of the stroke compared to the top. A scotch yoke has constant acceleration which smoothly turns the rotation into linear motion in a sinusoidal wave. The limits of these are they can only input up to about 7Hz with a fixed stroke, whereas trail inputs could be up to 20-30Hz. Trail inputs are often much more sudden and faster accelerating than that though, so it can't measure more extreme events. It's also tough to create high speeds and forces with these, so while my 2HP machine goes just over 1m/s no problem, a 20HP "only" reaches 3m/s in most cases. It depends on the stroke and force involved, an mtb damper with very light damping can get over 2m/s with much lower power but most machines are built with cars or motorbikes in mind. But since 90% of impacts are under 1/ms and the more noticeable damping is more like 0.25m/s, most tuners use a dyno around 3-5 HP for normal tuning and revalving work.
The other types of dyno are linear actuators which have an electromagnetic or hydraulic actuator - these are way more powerful (and 100's of thousands of dollars) but can not only go 5-10 m/s but have square or triangle wave inputs which are basically instant acceleration and direction changes if needed. They can create inputs up to 100Hz, drive the damper at different compression and rebound speeds and have infinitely variable stroke which means they can also replay track data collected on the bike. These are crucial for developing dampers as they truly test the limits of the valving for things like cavitation so manufacturers like Push, Vorsprung, Fox etc use them but they are extremely rare in a tuners workshop. fluid Focus use one for developing their pistons though.
A typical test is run at a fixed stroke on a rotary dyno (this can be changed though, my dyno does 0.5, 1 and 2") and looks like -
-Running a warm up to operating temperature (about 25-30 degrees)
-Then the dyno does a Rod force test which compresses the damper and stops to measure the internal force from the IFP and removes it from the test. This can be in the middle for an average force or taken at several points for a curve. How its done is important - if the adjusters are set very stiff and the the machine moves quickly, there can be residual pressure inside which throws out the reading so you need to tune you rod force test for allowing a "settle" time. By inputting the shaft diameter you can fairly accurately measure how much pressure is in the shock as well.
-Then you might do a seal drag test - just a very slow cycle that doesn't generate much damping but tells you how much friction is in the system which is useful in certain cases.
-Then the actual test begins - starting slowly (usually about 1in/s) the software is logging the whole time, but it will wait until the machine is settled at the speed and record a full "lap" then accelerate to the next speed. Again it will wait to reach the correct speed and take each recording, something like 1, 3, 5, 10 and 20 in/s.
-These are recorded as individiual constant velocity plots (CVP) (these are the plots which have a loop where the force is slightly different speeding up to slowing down), but the peak speed from each is turned in to a peak velocity plot (PVP) which are the nice straight line versions which normally get posted as they are a little clearer to read the basic function.
Those results can then be used to compare to calculated "target" damping rates, looks for problems or inconsistencies, or test for the effectiveness of damper adjusters.
And each person will have their own protocols - not many companies publish their own tests and each damper behave differently. If you have a perfectly linear damper you could do 2 speeds and draw a line between them - it would represent the damper fine, but there is normally some kind of change in rate a long the way so you will pick the speeds which work for you case. If time is important you will do as few speeds as possible, but if you need more detail across a range of speeds then you will do much more. 2 PVP plots of the same damper could look wildly different if you pick different speeds to test at. You do have to be careful comparing tests from different people, let alone other brands of machine - its tough to put all the context in to one post (which is why I'm careful about what I post) so different calibrations and test protocols can create different results. Tiny things like how the rod force test is performed can make the damper look wildly different and not everyone understands those effects.
As far as what it provides - the key thing is it gives very consistent and known results with controlled variables, even if it is a little different to how it responds on the trail. Basically you can measure 2 dampers - one that the rider likes and one they don't and look at how each is different. Do that a few thousand times and you can build up a model of what works well and what doesn't, then use that to predict what might work in the future. Just using trail data only goes so far because you can't always account for different conditions, rider ability, fatigue etc between each run. Its actually quite an indirect measurement because the input (ie ground) has to go through the tyre, wheel and flexing the frame and overcoming friction before it reaches the damper. How much of the impact went in to the suspension and how much went in to the parts or in to the rider - if you don't know that, you can't conclusively know what needs to change. Or once you attempted to make a change in the damper if it actually did anything. Trail data is super important but not the only tools to use. If I had to pick one or the other I would have a damper dyno. At the end ofd the day, what the riders feel is the most important - if I measure something that I think should work and they don't like it, I adjust model based on that feedback and try again.
Typical peak speeds are about 6-8m/s on downhill bike forks, and shocks will be up to around 2.5-3m/s. Most dampers "should" be fairly linear up to those max speeds - some aren't by design, but its rare for the ports to "choke" and start generating excessively high rates at those speeds (despite what some people think). It's more likely that high acceleration causes issues (so very high frequency but relatively low speed) which is why the linear actuator dynos exist.
Oh and the data card at the back of the machine - Typical peak speeds are about 6-8m/s on downhill bike forks, and shocks will be up...
Oh and the data card at the back of the machine -
Typical peak speeds are about 6-8m/s on downhill bike forks, and shocks will be up to around 2.5-3m/s. Most dampers "should" be fairly linear up to those max speeds - some aren't by design, but its rare for the ports to "choke" and start generating excessively high rates at those speeds (despite what some people think). It's more likely that high acceleration causes issues (so very high frequency but relatively low speed) which is why the linear actuator dynos exist.
This is all great info, and thanks to everyone who answered my question. We're really spoiled here in the Vital forums.
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the...
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the typical dyno test look like?
Is a dyno test always necessarily an automated machine process, or can it be reliably performed manually in a hand dyno?
Are there different types of dyno machines, or is every suspension brand and independent tuner pretty much using the same machine? How much do dyno test protocols vary between suspension brands and tuning shops? And if there are different machines, is data from one machine or test protocol comparable to data from another?
What kinds of information does a dyno test actually meaningfully provide, and what kinds of info can that test data never tell you? It seems like there isn't a 1:1 correlation between what the machine says about a fork or shock vs. what riders experience on the trail, meaning that real on-trail testing is still essential and you can't truly product test damper tunes from start to finish in a lab.
I'm betting this has already been answered somewhere in our vast internet, so if you links to share, that's great too.
A dyno is a machine to run standardised tests to quantify damper performance. It is not meant to replicate trail loads and doesn't as they load suspension in a very different way. Yes I know electromag dyno's can play back trail data but that's a party trick and not their day job.
A simple hand tester is not a dyno because it doesn't measure anything. You can instrument them though and people have.
Mechanically they are very simple. It's literally a gearmotor turning a cam in a greasy slot. This gives you a true sinusoid (sine wave). You can adjust the stroke and rpm to run different velocities and forces. Electrically they are a bit more complex. Digital speed controllers, position sensors on the yoke and a load cell to measure the forces. The real complication is all in the software.
Dyno plots are created in a CVP (continuous velocity profile) or football style. You can see this being created in real time in the video above. Then multiple CVP/Football plots can be stitched together to create PVP (peak velocity profile) graphs by literally joining the dots across the peaks like this:
Dyno Plot. Creating a PVP from multiple CVP plots
The interesting thing about PVP plots is they can hide all sorts of horrible behaviour. You can have cavitation holes like this that are completely hidden. This is a FVV (force vs velocity) plot which is CVP data straightened out by velocity instead of position:
Dyno plot. FVV showing cavitation hole.
Here's where the traps lie. Because dyno's load suspension differently to rough ground they can hide or cover a lot of unexpected behaviour. You can build a damper with preloaded shim stacks, hide them by bleeds and it can graph out nicely but ride like absolute garbage. Many examples of that.
You can build two dampers that graph out almost identical just with port size and shim stacks. But they will ride very differently. Inertia of parts and fluids is one factor that is hidden completely by a sinusoid dyno.
These two curves were the same fork damper at 1m/s with completely different port sizes and shim stacks. They graph out pretty much identical (within margin of error) at 1m/s and had about 10% deviation by 2m/s. But rode very differently:
Another analogy - trail data or rider feedback is like going to the doctor - they will ask you questions, measure heart rate and blood pressure - thats like data acquisition and finding the symptoms, but getting more answers needs something like blood tests or an x-ray. They might not tell you the actual cause but they give much more information about whats going on
Hand dynos are more or less intended for testing the bleed on a shock after reassembly. When talking about the machines, there are different kinds of dynos...
Hand dynos are more or less intended for testing the bleed on a shock after reassembly.
When talking about the machines, there are different kinds of dynos for sure. Cheaper ones (~10k USD) tend to go to shaft speeds of ~2.5m/s. If you want really high shaft speeds of about 7m/s, those things get expensive fast. Like ~80k USD expensive and upwards.
A 5 foot huck to flat could probably get you to an inital shaft speed of around 5m/s or 5000mm/s. Simple equation: time to drop =...
A 5 foot huck to flat could probably get you to an inital shaft speed of around 5m/s or 5000mm/s. Simple equation: time to drop = (2*height/g)^(1/2) so sqrt of 2*5/32 = .56s Speed at impact -> speed = time * acceleration or .56 g aka .56*32=17.92ft/s. You can do some more math to figure out how landing slope angle and head angle affect shaft speed, but it's not that important.
If a huck to flat gets you to 5m/s i bet you can find bumps that do more.
Another limitation of dyno charts: spiking occurs where exactly? Oh, off the chart? Cool. 😂
A 1 metre drop gives you a contact speed of:
v = sqqrt(2*g*h) v = sqqrt(2*9.81*1) v = 4.4 m/s. This speed is obviously blunted by the tyre and everything else flexing (including the rider), but that's the limit case for a flat 1m drop. If the landing is up the impact speed is higher. If the landing is down it's lower.
Dampers spiking has to be pretty bad to be directly read off a dyno chart. You have to read between the lines or just pull the damper apart and you can see what's going to do it. Usually it's evidenced on a dyno chart by preloaded or very stiff shim stacks which are hidden by small bleeds and gap shims. Sometimes you have to get in and close off these hidden bleeds to see what the damper is trying to do on a sharp impact.
I was talking about preload in general and how they tend to realise it. Charger 3 base valve is dished. If I recall correctly, both Super...
I was talking about preload in general and how they tend to realise it. Charger 3 base valve is dished. If I recall correctly, both Super Deluxe Coil and Vivid have stepped base valves, too. Also the R85 rebound tune is preloaded via ring shim. What was wildly wrong here?
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright scary on jumps. (Would apply for the R85 rebound tune on the D1.)"
The R85 is the most progressive rebound tune you will see in a mainstream damper. Yes it is preloaded with a ring shim but that doesn't automatically mean its digressive. Most of the preloaded rebound stack Rockshox have made are so stiff they don't open, as you mentioned - which makes them progressive.
They have always used dished pistons or ring shims in the shock base valve- it's not inherently bad and its normally the safest way to quickly build pressure in the damper. Whether or not it "suddenly" releases can be controlled by the rest of the valve geometry.
to get back on track here, what are the thoughts on the new Super deluxe ultimate?
to get back on track here, what are the thoughts on the new Super deluxe ultimate?
The old one was better. They both have weird stock tunes but the 2018-22 Super Deluxe's (air and coil) would tune up very well with just shims as you can easily remove the ring-shims in the stacks. Use 2 stage shim stacks on compression and rebound with some basevalve reshimming on the 2018-22 to get you an excellent and reliable shock. Plus the shims were mostly the same sizes.
The 23+ need more changes to perform the same. Rebound ports on the coils are too small and the dished piston built into the base-valves causes a dead feeling. The coils use a whole heap of different shim sizes too so tuning is a bigger PITA. The lower end 23+ shocks are easier to tune as they have a simpler base-valve without the dished piston. The 23+ air shocks use a very similar main piston setup to the 2018-22 and tune up easier than the coils.
My recommendation: Get a base-mid spec shock with climb switch only if you have a choice with the 23+. Not the ultimate with HSC/LSC. For the 2018-22 get the Ultimate with LSC adjust.
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the...
Here's a dumb question, but when we talk about dyno charts for suspension components, what does the actual dyno machine look like, and what does the typical dyno test look like?
Is a dyno test always necessarily an automated machine process, or can it be reliably performed manually in a hand dyno?
Are there different types of dyno machines, or is every suspension brand and independent tuner pretty much using the same machine? How much do dyno test protocols vary between suspension brands and tuning shops? And if there are different machines, is data from one machine or test protocol comparable to data from another?
What kinds of information does a dyno test actually meaningfully provide, and what kinds of info can that test data never tell you? It seems like there isn't a 1:1 correlation between what the machine says about a fork or shock vs. what riders experience on the trail, meaning that real on-trail testing is still essential and you can't truly product test damper tunes from start to finish in a lab.
I'm betting this has already been answered somewhere in our vast internet, so if you links to share, that's great too.
The machine is basically a motor that turns a shaft, which is turned in to linear up and down motion. The software runs the machine faster...
The machine is basically a motor that turns a shaft, which is turned in to linear up and down motion. The software runs the machine faster and faster, and measures how much force the damper is producing at every point and speed in the stroke. My machine is a CTW "rotary" type dyno with 2 HP and one of the more basic, lower power machines on the market (I think about $15,000 USD today). CTW was born out of a company called Roehrig who pioneered workshop damper testing in the early 90's - many of the older racecar and motorcycle tuners still use Roehrig Dynos today, along with mtb companies like Push and Vorsprung. They are the most experienced and knowledgeable when it comes to building damper dynos but there are other brands out there, Laba 7 is a recent one pushing hard, mechatronics labs is smaller but seem to make very nice machines. There are 2 main types-
Rotary Dynos like mine use a motor to drive either a crankshaft or a scotch yoke. Crankshafts are not very common anymore because like a piston in a car, the damper is accelerating faster at the bottom half of the stroke compared to the top. A scotch yoke has constant acceleration which smoothly turns the rotation into linear motion in a sinusoidal wave. The limits of these are they can only input up to about 7Hz with a fixed stroke, whereas trail inputs could be up to 20-30Hz. Trail inputs are often much more sudden and faster accelerating than that though, so it can't measure more extreme events. It's also tough to create high speeds and forces with these, so while my 2HP machine goes just over 1m/s no problem, a 20HP "only" reaches 3m/s in most cases. It depends on the stroke and force involved, an mtb damper with very light damping can get over 2m/s with much lower power but most machines are built with cars or motorbikes in mind. But since 90% of impacts are under 1/ms and the more noticeable damping is more like 0.25m/s, most tuners use a dyno around 3-5 HP for normal tuning and revalving work.
The other types of dyno are linear actuators which have an electromagnetic or hydraulic actuator - these are way more powerful (and 100's of thousands of dollars) but can not only go 5-10 m/s but have square or triangle wave inputs which are basically instant acceleration and direction changes if needed. They can create inputs up to 100Hz, drive the damper at different compression and rebound speeds and have infinitely variable stroke which means they can also replay track data collected on the bike. These are crucial for developing dampers as they truly test the limits of the valving for things like cavitation so manufacturers like Push, Vorsprung, Fox etc use them but they are extremely rare in a tuners workshop. fluid Focus use one for developing their pistons though.
A typical test is run at a fixed stroke on a rotary dyno (this can be changed though, my dyno does 0.5, 1 and 2") and looks like -
-Running a warm up to operating temperature (about 25-30 degrees)
-Then the dyno does a Rod force test which compresses the damper and stops to measure the internal force from the IFP and removes it from the test. This can be in the middle for an average force or taken at several points for a curve. How its done is important - if the adjusters are set very stiff and the the machine moves quickly, there can be residual pressure inside which throws out the reading so you need to tune you rod force test for allowing a "settle" time. By inputting the shaft diameter you can fairly accurately measure how much pressure is in the shock as well.
-Then you might do a seal drag test - just a very slow cycle that doesn't generate much damping but tells you how much friction is in the system which is useful in certain cases.
-Then the actual test begins - starting slowly (usually about 1in/s) the software is logging the whole time, but it will wait until the machine is settled at the speed and record a full "lap" then accelerate to the next speed. Again it will wait to reach the correct speed and take each recording, something like 1, 3, 5, 10 and 20 in/s.
-These are recorded as individiual constant velocity plots (CVP) (these are the plots which have a loop where the force is slightly different speeding up to slowing down), but the peak speed from each is turned in to a peak velocity plot (PVP) which are the nice straight line versions which normally get posted as they are a little clearer to read the basic function.
Those results can then be used to compare to calculated "target" damping rates, looks for problems or inconsistencies, or test for the effectiveness of damper adjusters.
And each person will have their own protocols - not many companies publish their own tests and each damper behave differently. If you have a perfectly linear damper you could do 2 speeds and draw a line between them - it would represent the damper fine, but there is normally some kind of change in rate a long the way so you will pick the speeds which work for you case. If time is important you will do as few speeds as possible, but if you need more detail across a range of speeds then you will do much more. 2 PVP plots of the same damper could look wildly different if you pick different speeds to test at. You do have to be careful comparing tests from different people, let alone other brands of machine - its tough to put all the context in to one post (which is why I'm careful about what I post) so different calibrations and test protocols can create different results. Tiny things like how the rod force test is performed can make the damper look wildly different and not everyone understands those effects.
As far as what it provides - the key thing is it gives very consistent and known results with controlled variables, even if it is a little different to how it responds on the trail. Basically you can measure 2 dampers - one that the rider likes and one they don't and look at how each is different. Do that a few thousand times and you can build up a model of what works well and what doesn't, then use that to predict what might work in the future. Just using trail data only goes so far because you can't always account for different conditions, rider ability, fatigue etc between each run. Its actually quite an indirect measurement because the input (ie ground) has to go through the tyre, wheel and flexing the frame and overcoming friction before it reaches the damper. How much of the impact went in to the suspension and how much went in to the parts or in to the rider - if you don't know that, you can't conclusively know what needs to change. Or once you attempted to make a change in the damper if it actually did anything. Trail data is super important but not the only tools to use. If I had to pick one or the other I would have a damper dyno. At the end ofd the day, what the riders feel is the most important - if I measure something that I think should work and they don't like it, I adjust model based on that feedback and try again.
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright...
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright scary on jumps. (Would apply for the R85 rebound tune on the D1.)"
The R85 is the most progressive rebound tune you will see in a mainstream damper. Yes it is preloaded with a ring shim but that doesn't automatically mean its digressive. Most of the preloaded rebound stack Rockshox have made are so stiff they don't open, as you mentioned - which makes them progressive.
They have always used dished pistons or ring shims in the shock base valve- it's not inherently bad and its normally the safest way to quickly build pressure in the damper. Whether or not it "suddenly" releases can be controlled by the rest of the valve geometry.
Alright, I will concede that. But if the preloaded shim never opens that means no oilflow is going through the stack, right?
About the base valve preload: I tried stressing that removing the preload is more difficult if it is applied through piston geometry rather than a ring shim. Maybe it isn't bad in all situations, but if I don't have the means of testing that at home, I'd rather be able to remove the preload myself without hassle to see if I like that better.
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright...
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright scary on jumps. (Would apply for the R85 rebound tune on the D1.)"
The R85 is the most progressive rebound tune you will see in a mainstream damper. Yes it is preloaded with a ring shim but that doesn't automatically mean its digressive. Most of the preloaded rebound stack Rockshox have made are so stiff they don't open, as you mentioned - which makes them progressive.
They have always used dished pistons or ring shims in the shock base valve- it's not inherently bad and its normally the safest way to quickly build pressure in the damper. Whether or not it "suddenly" releases can be controlled by the rest of the valve geometry.
Alright, I will concede that. But if the preloaded shim never opens that means no oilflow is going through the stack, right? About the base valve preload...
Alright, I will concede that. But if the preloaded shim never opens that means no oilflow is going through the stack, right?
About the base valve preload: I tried stressing that removing the preload is more difficult if it is applied through piston geometry rather than a ring shim. Maybe it isn't bad in all situations, but if I don't have the means of testing that at home, I'd rather be able to remove the preload myself without hassle to see if I like that better.
The rebound shims on the preloaded Monarch and Deluxe/Super Deluxe shocks do open. The dyno plot shapes make that clear and the witness/wear marks on the shims do too. I have no idea why Steve said they didn't.
I'll post up some dyno plots later of a shock with no rebound shims (everything through the bleed) and a preloaded Rockshox. Something is buggy when I try to do it right now. The preloaded Rockshox go digressive near fully closed, the shocks without shims go quadratic and threaten to break themselves and the dyno if you push the speed too high.
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright...
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright scary on jumps. (Would apply for the R85 rebound tune on the D1.)"
The R85 is the most progressive rebound tune you will see in a mainstream damper. Yes it is preloaded with a ring shim but that doesn't automatically mean its digressive. Most of the preloaded rebound stack Rockshox have made are so stiff they don't open, as you mentioned - which makes them progressive.
They have always used dished pistons or ring shims in the shock base valve- it's not inherently bad and its normally the safest way to quickly build pressure in the damper. Whether or not it "suddenly" releases can be controlled by the rest of the valve geometry.
Alright, I will concede that. But if the preloaded shim never opens that means no oilflow is going through the stack, right? About the base valve preload...
Alright, I will concede that. But if the preloaded shim never opens that means no oilflow is going through the stack, right?
About the base valve preload: I tried stressing that removing the preload is more difficult if it is applied through piston geometry rather than a ring shim. Maybe it isn't bad in all situations, but if I don't have the means of testing that at home, I'd rather be able to remove the preload myself without hassle to see if I like that better.
Here's a dyno plot from a Fox Vanilla X. This has fixed rebound with no shims. The plot goes quadratic in a hurry and it would break the shock or the dyno if you kept increasing speed. That big loop in the plot is hysteresis where the stress is high enough for the oil to compress/expand and the shock body to expand/contract too:
Dyno Plot. Fox Vanilla X with fixed rebound
This is a Rockshox Monarch with preloaded rebound circuit. It goes digressive near closed like all preloaded circuits do and can't really get a good balance between chassis control and following bumps at speed. The shims clearly open and close:
The old one was better. They both have weird stock tunes but the 2018-22 Super Deluxe's (air and coil) would tune up very well with just...
The old one was better. They both have weird stock tunes but the 2018-22 Super Deluxe's (air and coil) would tune up very well with just shims as you can easily remove the ring-shims in the stacks. Use 2 stage shim stacks on compression and rebound with some basevalve reshimming on the 2018-22 to get you an excellent and reliable shock. Plus the shims were mostly the same sizes.
The 23+ need more changes to perform the same. Rebound ports on the coils are too small and the dished piston built into the base-valves causes a dead feeling. The coils use a whole heap of different shim sizes too so tuning is a bigger PITA. The lower end 23+ shocks are easier to tune as they have a simpler base-valve without the dished piston. The 23+ air shocks use a very similar main piston setup to the 2018-22 and tune up easier than the coils.
My recommendation: Get a base-mid spec shock with climb switch only if you have a choice with the 23+. Not the ultimate with HSC/LSC. For the 2018-22 get the Ultimate with LSC adjust.
Was considering trying an air shock and the Rs super deluxe ultimate is tuned for my offering v4 by Evil and people seem to be mostly happy with it, but I guess I’m better off keeping my Push 11.6 gen 1
Thanks for the info!
That raises another question, which is how fast shaft velocities can be while mountain biking? Greater than 2.5 meters/second? Greater than 7.5?
Forks frequently see shaft speeds of 4m/s and in general up to about 8m/s (see this video by Vorsprung). Rear shocks see quite a lot slower shaft speeds in general due to two reasons: Leverage ratio and often not being the first in line for the hit.
Hand dyno is basically just jerking off a shock. Need a load cell and velocity sensor to be able to generate graphs.
The big flaw of the shock/damper dyno is it only tells you what it tells you, it's isolating things. So like in a real suspension system you have the chassis and the spring and those things interact in complex ways. If you watch video tours of fox R&D they seem to maybe dyno whole forks, but looks to be purely vertically which eliminates a lot of the side loading that happens in real life (actually loading the bushings etc).
All damping is, from a physics perspective, is an energy bleed. It's bleeding off energy from the system (which could be bike+rider on single bump event). Friction does the same thing, so quantifying what your damper is doing is great, but it's only one part of the picture of damping happening in the system, and depending on the scenario, the main source of damping isn't even the damper. Which is not a good thing as the damper provides the most controllable form of damping.
Additionally in something like a fork, you have an imperfect lubricaton system, where the fork is trying to pump oil up to the foam rings and bushings, and then the action of the fork is also pushing oil out of the bushings, where it's needed most. This is something that I think fox and rockshox have made a lot of progress with over the last 6 years or so. It used to be that turning your fork upside down before a ride made a massive difference. That seemed to kind of go away for me when fox introduced the cast oil channels into the lower legs. Remember when WC mechanics would talk about doing a fork rebuild for practically every run? Well if your fork lubrication system is awful, it makes a lot of sense...
But all the complex chassis stuff isn't really tunable, and it's real easy to turn knobs or swap shims. So we end up talking about knobs and shims and PSIs.
More sophisticated test setups/machines can probably go a long way to evaluating something like fork performance more holistically, but tiny bicycle industry is tiny and this stuff is actually expensive. This is why, in my opinion, we get things like fox releasing a new fork and being able to honestly say that it fixed the old fork "not using its travel" aka not being a fork (first gen 34 sc anyone?). They've also had to boil the frog (us) into accepting weight penalties for performance. In the past you wanted the light fork cause.. well it didn't work anyway. Then you get into the whole question of: looking at all these charts, what's better? Now some of the chat at my LBS is: the forks are so good now, what does this mean for riding if these trails aren't gnarly anymore?
A 5 foot huck to flat could probably get you to an inital shaft speed of around 5m/s or 5000mm/s. Simple equation: time to drop = (2*height/g)^(1/2) so sqrt of 2*5/32 = .56s Speed at impact -> speed = time * acceleration or .56 g aka .56*32=17.92ft/s. You can do some more math to figure out how landing slope angle and head angle affect shaft speed, but it's not that important.
If a huck to flat gets you to 5m/s i bet you can find bumps that do more.
Another limitation of dyno charts: spiking occurs where exactly? Oh, off the chart? Cool. 😂
The machine is basically a motor that turns a shaft, which is turned in to linear up and down motion. The software runs the machine faster and faster, and measures how much force the damper is producing at every point and speed in the stroke. My machine is a CTW "rotary" type dyno with 2 HP and one of the more basic, lower power machines on the market (I think about $15,000 USD today). CTW was born out of a company called Roehrig who pioneered workshop damper testing in the early 90's - many of the older racecar and motorcycle tuners still use Roehrig Dynos today, along with mtb companies like Push and Vorsprung. They are the most experienced and knowledgeable when it comes to building damper dynos but there are other brands out there, Laba 7 is a recent one pushing hard, mechatronics labs is smaller but seem to make very nice machines. There are 2 main types-
Rotary Dynos like mine use a motor to drive either a crankshaft or a scotch yoke. Crankshafts are not very common anymore because like a piston in a car, the damper is accelerating faster at the bottom half of the stroke compared to the top. A scotch yoke has constant acceleration which smoothly turns the rotation into linear motion in a sinusoidal wave. The limits of these are they can only input up to about 7Hz with a fixed stroke, whereas trail inputs could be up to 20-30Hz. Trail inputs are often much more sudden and faster accelerating than that though, so it can't measure more extreme events. It's also tough to create high speeds and forces with these, so while my 2HP machine goes just over 1m/s no problem, a 20HP "only" reaches 3m/s in most cases. It depends on the stroke and force involved, an mtb damper with very light damping can get over 2m/s with much lower power but most machines are built with cars or motorbikes in mind. But since 90% of impacts are under 1/ms and the more noticeable damping is more like 0.25m/s, most tuners use a dyno around 3-5 HP for normal tuning and revalving work.
The other types of dyno are linear actuators which have an electromagnetic or hydraulic actuator - these are way more powerful (and 100's of thousands of dollars) but can not only go 5-10 m/s but have square or triangle wave inputs which are basically instant acceleration and direction changes if needed. They can create inputs up to 100Hz, drive the damper at different compression and rebound speeds and have infinitely variable stroke which means they can also replay track data collected on the bike. These are crucial for developing dampers as they truly test the limits of the valving for things like cavitation so manufacturers like Push, Vorsprung, Fox etc use them but they are extremely rare in a tuners workshop. fluid Focus use one for developing their pistons though.
A typical test is run at a fixed stroke on a rotary dyno (this can be changed though, my dyno does 0.5, 1 and 2") and looks like -
-Running a warm up to operating temperature (about 25-30 degrees)
-Then the dyno does a Rod force test which compresses the damper and stops to measure the internal force from the IFP and removes it from the test. This can be in the middle for an average force or taken at several points for a curve. How its done is important - if the adjusters are set very stiff and the the machine moves quickly, there can be residual pressure inside which throws out the reading so you need to tune you rod force test for allowing a "settle" time. By inputting the shaft diameter you can fairly accurately measure how much pressure is in the shock as well.
-Then you might do a seal drag test - just a very slow cycle that doesn't generate much damping but tells you how much friction is in the system which is useful in certain cases.
-Then the actual test begins - starting slowly (usually about 1in/s) the software is logging the whole time, but it will wait until the machine is settled at the speed and record a full "lap" then accelerate to the next speed. Again it will wait to reach the correct speed and take each recording, something like 1, 3, 5, 10 and 20 in/s.
-These are recorded as individiual constant velocity plots (CVP) (these are the plots which have a loop where the force is slightly different speeding up to slowing down), but the peak speed from each is turned in to a peak velocity plot (PVP) which are the nice straight line versions which normally get posted as they are a little clearer to read the basic function.
Those results can then be used to compare to calculated "target" damping rates, looks for problems or inconsistencies, or test for the effectiveness of damper adjusters.
And each person will have their own protocols - not many companies publish their own tests and each damper behave differently. If you have a perfectly linear damper you could do 2 speeds and draw a line between them - it would represent the damper fine, but there is normally some kind of change in rate a long the way so you will pick the speeds which work for you case. If time is important you will do as few speeds as possible, but if you need more detail across a range of speeds then you will do much more. 2 PVP plots of the same damper could look wildly different if you pick different speeds to test at. You do have to be careful comparing tests from different people, let alone other brands of machine - its tough to put all the context in to one post (which is why I'm careful about what I post) so different calibrations and test protocols can create different results. Tiny things like how the rod force test is performed can make the damper look wildly different and not everyone understands those effects.
As far as what it provides - the key thing is it gives very consistent and known results with controlled variables, even if it is a little different to how it responds on the trail. Basically you can measure 2 dampers - one that the rider likes and one they don't and look at how each is different. Do that a few thousand times and you can build up a model of what works well and what doesn't, then use that to predict what might work in the future. Just using trail data only goes so far because you can't always account for different conditions, rider ability, fatigue etc between each run. Its actually quite an indirect measurement because the input (ie ground) has to go through the tyre, wheel and flexing the frame and overcoming friction before it reaches the damper. How much of the impact went in to the suspension and how much went in to the parts or in to the rider - if you don't know that, you can't conclusively know what needs to change. Or once you attempted to make a change in the damper if it actually did anything. Trail data is super important but not the only tools to use. If I had to pick one or the other I would have a damper dyno. At the end ofd the day, what the riders feel is the most important - if I measure something that I think should work and they don't like it, I adjust model based on that feedback and try again.
For more info, I highly recommend CTW's youtube channel https://www.youtube.com/@ctwautomation330/videos
Oh and the data card at the back of the machine -
Typical peak speeds are about 6-8m/s on downhill bike forks, and shocks will be up to around 2.5-3m/s. Most dampers "should" be fairly linear up to those max speeds - some aren't by design, but its rare for the ports to "choke" and start generating excessively high rates at those speeds (despite what some people think). It's more likely that high acceleration causes issues (so very high frequency but relatively low speed) which is why the linear actuator dynos exist.
This is all great info, and thanks to everyone who answered my question. We're really spoiled here in the Vital forums.
A dyno is a machine to run standardised tests to quantify damper performance. It is not meant to replicate trail loads and doesn't as they load suspension in a very different way. Yes I know electromag dyno's can play back trail data but that's a party trick and not their day job.
A simple hand tester is not a dyno because it doesn't measure anything. You can instrument them though and people have.
Here's a video showing the scotch yoke end of my Laba 7 dyno in operation: https://www.instagram.com/p/DRis7JOlJme/
Mechanically they are very simple. It's literally a gearmotor turning a cam in a greasy slot. This gives you a true sinusoid (sine wave). You can adjust the stroke and rpm to run different velocities and forces.
Electrically they are a bit more complex. Digital speed controllers, position sensors on the yoke and a load cell to measure the forces.
The real complication is all in the software.
Dyno plots are created in a CVP (continuous velocity profile) or football style. You can see this being created in real time in the video above. Then multiple CVP/Football plots can be stitched together to create PVP (peak velocity profile) graphs by literally joining the dots across the peaks like this:
The interesting thing about PVP plots is they can hide all sorts of horrible behaviour. You can have cavitation holes like this that are completely hidden. This is a FVV (force vs velocity) plot which is CVP data straightened out by velocity instead of position:
Here's where the traps lie. Because dyno's load suspension differently to rough ground they can hide or cover a lot of unexpected behaviour. You can build a damper with preloaded shim stacks, hide them by bleeds and it can graph out nicely but ride like absolute garbage. Many examples of that.
You can build two dampers that graph out almost identical just with port size and shim stacks. But they will ride very differently. Inertia of parts and fluids is one factor that is hidden completely by a sinusoid dyno.
These two curves were the same fork damper at 1m/s with completely different port sizes and shim stacks. They graph out pretty much identical (within margin of error) at 1m/s and had about 10% deviation by 2m/s. But rode very differently:
https://www.instagram.com/p/ClYLTiKSVK7/
Another analogy - trail data or rider feedback is like going to the doctor - they will ask you questions, measure heart rate and blood pressure - thats like data acquisition and finding the symptoms, but getting more answers needs something like blood tests or an x-ray. They might not tell you the actual cause but they give much more information about whats going on
A 1 metre drop gives you a contact speed of:
v = sqqrt(2*g*h)
v = sqqrt(2*9.81*1)
v = 4.4 m/s.
This speed is obviously blunted by the tyre and everything else flexing (including the rider), but that's the limit case for a flat 1m drop. If the landing is up the impact speed is higher. If the landing is down it's lower.
Here's a tracker video (not mine) showing 2.5m/s on an average trail with a small drop: https://www.youtube.com/watch?v=V5GtG4Fl7BI
Dampers spiking has to be pretty bad to be directly read off a dyno chart. You have to read between the lines or just pull the damper apart and you can see what's going to do it. Usually it's evidenced on a dyno chart by preloaded or very stiff shim stacks which are hidden by small bleeds and gap shims. Sometimes you have to get in and close off these hidden bleeds to see what the damper is trying to do on a sharp impact.
"You will always have pressure buildup before the preloaded shims open up, and on top of that, in my opinion, digressive rebound tunes are just downright scary on jumps. (Would apply for the R85 rebound tune on the D1.)"
The R85 is the most progressive rebound tune you will see in a mainstream damper. Yes it is preloaded with a ring shim but that doesn't automatically mean its digressive. Most of the preloaded rebound stack Rockshox have made are so stiff they don't open, as you mentioned - which makes them progressive.
They have always used dished pistons or ring shims in the shock base valve- it's not inherently bad and its normally the safest way to quickly build pressure in the damper. Whether or not it "suddenly" releases can be controlled by the rest of the valve geometry.
The old one was better. They both have weird stock tunes but the 2018-22 Super Deluxe's (air and coil) would tune up very well with just shims as you can easily remove the ring-shims in the stacks. Use 2 stage shim stacks on compression and rebound with some basevalve reshimming on the 2018-22 to get you an excellent and reliable shock. Plus the shims were mostly the same sizes.
The 23+ need more changes to perform the same. Rebound ports on the coils are too small and the dished piston built into the base-valves causes a dead feeling. The coils use a whole heap of different shim sizes too so tuning is a bigger PITA. The lower end 23+ shocks are easier to tune as they have a simpler base-valve without the dished piston. The 23+ air shocks use a very similar main piston setup to the 2018-22 and tune up easier than the coils.
My recommendation: Get a base-mid spec shock with climb switch only if you have a choice with the 23+. Not the ultimate with HSC/LSC. For the 2018-22 get the Ultimate with LSC adjust.
Not going to lie, you had me at googly eyes.
I need to get some bigger ones and maybe 3d print a mouth that goes on the bearing to make it look like its screaming
I do honestly do work, I pormise
Alright, I will concede that. But if the preloaded shim never opens that means no oilflow is going through the stack, right?
About the base valve preload: I tried stressing that removing the preload is more difficult if it is applied through piston geometry rather than a ring shim. Maybe it isn't bad in all situations, but if I don't have the means of testing that at home, I'd rather be able to remove the preload myself without hassle to see if I like that better.
The rebound shims on the preloaded Monarch and Deluxe/Super Deluxe shocks do open. The dyno plot shapes make that clear and the witness/wear marks on the shims do too. I have no idea why Steve said they didn't.
I'll post up some dyno plots later of a shock with no rebound shims (everything through the bleed) and a preloaded Rockshox. Something is buggy when I try to do it right now. The preloaded Rockshox go digressive near fully closed, the shocks without shims go quadratic and threaten to break themselves and the dyno if you push the speed too high.
Here's a dyno plot from a Fox Vanilla X. This has fixed rebound with no shims. The plot goes quadratic in a hurry and it would break the shock or the dyno if you kept increasing speed. That big loop in the plot is hysteresis where the stress is high enough for the oil to compress/expand and the shock body to expand/contract too:
This is a Rockshox Monarch with preloaded rebound circuit. It goes digressive near closed like all preloaded circuits do and can't really get a good balance between chassis control and following bumps at speed. The shims clearly open and close:
Was considering trying an air shock and the Rs super deluxe ultimate is tuned for my offering v4 by Evil and people seem to be mostly happy with it, but I guess I’m better off keeping my Push 11.6 gen 1
Post a reply to: Suspension Component Technology/Functionality Discussion