Jump to content

kevinstillwell

Members
  • Content Count

    481
  • Joined

  • Last visited

About kevinstillwell

  • Rank
    TT Bronze Member

Contact Methods

Profile Information

  • Location
    Texas

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. ---------------------------------------------------------------------------------------------------------------- "If you run a test and it confirms your hypothesis, you have taken a measurement. If you run a test and it doesn't perform as expected, you have made a discovery". ---------------------------------------------------------------------------------------------------------------- We dyno and pressure tested rec_id 1724s. With this setup, we were testing an aftermarket mv piston with the stk cadj. Unfortunately, we made a discovery. We noticed the stk cadj forces had changed from a previous test (1662) with this same adjuster. They should have been the same. This table shows the difference in cadj forces between 1724_01 vs. 1662_05. - - Stack 1 is the test with the stiffer cadj forces (1724s_01). - - Stack 2 was a dyno test for the stock valving (1662s_05). - - - - The big difference is at the lower velocities from 1-20 ips. The force more than doubles at 2-10 ips. With most valving changes, you might shoot for a 10 % change. Doubling the forces is huge. - - - - We'll need to run a series of tests if we want to determine what's going on. - - - - - - - - - - - - - Both tests have the same shock and stock cadj. But when we went to assemble the adjuster for this test (1724), we couldn't find the shim stack from test 1662. So for this test, we grabbed all new shims from the bins. Obviously, the first thing to test (to find the difference) would have been the shims. Simply put back the original shim stack and see if that was the difference. But since we didn't have the original shims, we have to continue with what we have. We'll start here. - - We knew from previous experience that if the piston washer and housing washers are distorted or damaged, this can effect the damping. So the first thing we did was simply flip the piston washer. - - - - - - - - - - - - - Here are the cadj force numbers. - - Stack 1 is with the piston washer flipped (1724s_05). - - Stack 2 is from above (1724s_01). - - - - It looks like flipping the piston washer reduced the cadj force numbers. But it's still stiffer than the original 1662s_05. There must be something else going on. - - - - - - - - - - - - - Good testing practices dictate that you cannot rely on one test. So we decided to flip the piston washer back again. If the piston washer is causing the increase in force, the numbers should jump right back up. -----> NOTE: Stack 1 is always the current stack in question. Stack 2 is what we are comparing against. - - Stack 1 is this test where we flipped it back (1724s_08). - - Stack 2 is the previous test where we first flipped the piston washer (1724s_05). - - - - We can see the cadj forces didn't change. This tells us that the piston washer isn't the culprit. Must be something else that is causing the difference in cadj forces between 1724s_01 and 1662s_05. - - - - - - - - - - - - - Next we'll try the nut torque. We generally use 40 in/lb. We will set up 1724s_13 with 25 in/lb. - - Reducing the torque on the nut actually increases the cadj forces. Hmmm, these numbers look similar to 1662s_05, time to put the thinking cap on. . . . - - - - I do recall torqueing 1662s_05, and I remember forgetting to put a drop of oil on the nut. So I torqued the nut dry, probably resulting in an inaccurate torque. - - - - So for now, we will say this mystery is solved. The difference between 1724s_01 and 1662s_05 was the nut torque. - - - - But wait. The nut torque doesn't account fully for the differences. . . . But as I mentioned, we also swapped the shim stack. We have determined from previous tests that there are differences between shims. On a light stack with many large shims, the differences will be less noticed, but on a stiff stack with only a few small shims, the differences are more noticed. So we will 'assume' the difference between 1724s_01 and 1662s_05 to be a combination of the torque and shims. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Why stop here. Lets test more stuff. Next we will do something extremely simple to the cadj. This wouldn't really be considered a mod, as we aren't machining or changing out any parts. This is something that some tuners might do, and some might not do. We are going back to 40 in/lb torque, so we'll compare this change against 1724s_08. - - Stack 1 is this test with the mystery mod (1724s_16). - - Stack 2 will be the 'stable' stack with 40 in/lb (1724s_08). - - - - Hello! This made a huge difference. We'll have to say this difference even surprised us. - - - - - - - - - - - - The next obvious step is to put the cadj back, retest and see if the numbers go back to where they were. - - The mystery mod required us to 'do something' to a couple of the pieces in the cadj. Since we can't undo this, we will simply grab new stock pieces for this test. - - Stack 1 has the new/exchanged stock pieces (1724s_24). - - Stack 2 is still the 'stable' stack (1724s_08). - - - - Dang, I hate it when this happens. The numbers don't match. This test with the new/exchanged pieces is stiffer. - - - - Compared to some of the changes we've seen, this isn't a major difference. The forces have changed 15-30%, but we'd still consider that to be significant. Time to put the thinking cap back on. . . . Conclusion: We will come to a conclusion based not only on these few tests, but from the hundreds of previous tests where we have noticed these same discrepancies. The conclusion is that the stock KBY cadj is not very sophisticated, and by simply disassembling and reassembling the adjuster, and possibly performing some very common 'assembly procedures', your cadj force numbers will be all over the place. - - When you think about it, the adjuster is made up of very small pieces, but it handles a lot of pressure. It actually creates the pressure in the shock. So any minute difference in tolerances will be magnified. - - Since most don't have a way to measure the adjuster, there is really no way to know. One way you might notice is when you put the same valving in several different bikes (for the same rider). - - -- For example. A rider might have 4 bikes, all valved the same. He is almost always going to have one of the bikes that he likes the best. It will have the best power and the best suspension. Discrepancies in stock compression adjusters could be one of the reasons that all the bikes suspension doesn't turn out exactly the same. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - We can run one more test to see if our conclusion is fairly accurate. We will test a couple aftermarket cadj piston assemblies. These pistons (and related pieces) have a different design, and are machined to exacting tolerances. We'll use this same shock and run two more tests (with two different cadj pistons). - - This is a quick test, and the cadj forces from these aftermarket pistons aren't intended to mirror the stk forces. We will simply install two different assemblies in the same shock and see how much variation we get. - - - - Again, we are surprised. These cadj forces are very close. - - - - As I mentioned, we have tested many stock compression adjusters, and they definitely lack any consistency. Some of you may recall the post where I found a stock adjuster that produces over 300 lbs force, when it should have been around 130. That's an extreme, but you can almost be certain that the forces will vary. - - - - - - - - - - - - The End - - - - -
  2. I'd like to get a few opinions on how much the balance spring is scratching the inside of the ctg tube on the 13-14 CRF 450 and 13-14 KXF 450. You would have to remove the seal assy end to access the balance spring, then look inside the ctg to see any wear. So the question is, how much wear has anyone noticed, and if there are scratches, how much time is on the bike. Here is a picture of the ctg tube.
  3. kevinstillwell

    shim changes and damping curves, based on dyno testing

    Darren, First, I don't want anyone to get the impression that Scott recommended using PVP. Obviously, nobody knows who Scott is, but we know he is two notches above the average bloke when it comes to this sort of information. So I don't want to give the wrong impression. Basically, running a PVP test gives you exactly the same data as running a CVP. You just get the option of one additional bit of information (the PVP info). You get this info in the report data as well as in the graphs. It would be extremely difficult to try to explain this in a Thumper Talk post, and we would be the only two guys even slightly interested. kevin
  4. kevinstillwell

    Does anybody like their stock suspension?

    Hi geaux, I'm guessing you have ridden the bike with the same shock for a couple / few years. If that's the case, you simply need to have the shock rebuilt. On the flip side, if you buy a shock off eBay, it will be used and worn out. So it will need to be rebuilt anyway. And, it won't be designed for your bike anyway, so you are opening the proverbial 'can of worms. kevin
  5. kevinstillwell

    shim changes and damping curves, based on dyno testing

    - - - - - - - - - - - - - Darren, This run is at 14 ips (inch per second). What are the peak velocities you test shocks at?
  6. kevinstillwell

    shim changes and damping curves, based on dyno testing

    That depends on the clicker setting. And that specific type of info might fit into the 'top secret' category.
  7. kevinstillwell

    shim changes and damping curves, based on dyno testing

    For anyone interested, the two dyno tests we are talking about are: 1. PVP = Peak Velocity Pickoff. In all my discussions, I show test results based on the data collected from 11 individual test speeds; 1, 2, 3, 4, 5, 10, 20, 30, 40, 50, 60 ips. The dyno simply runs each test at the selected speed, and then pulls out the peak force for that velocity. For example, from the above graphs, 115 lbs peak force at 2ips. 2. CVP = Constant Velocity Pickoff. This type of test collects many more data points. For example, I have my machine set to get 2000 samples per second. This is why you could question looking at one data point, when you can look at thousands of data points. 3. I might as well answer that. Why PVP tests. - - We use the PVP because it gives a quick snapshot of the forces for the speeds tested. We have found this works well when trying to understand how the valving changes effect the forces at the different velocities. We always have the CVP to fall back on as needed. - - In essence, we are frequency testing, with the different velocities being the frequency. (frequency = the rate of occurrence of anything, in this case the velocities). - - - - - - - - - - - - - Just the other day I had a long conversation with Scott at Roehrig about this. The consensus was that there is so much data to look at, you have to find some way to make it manageable. This is the method we've adapted to make sense out of everything. And when you start pressure testing, the amount of data goes up 4x. When you run the PVP test, you still get the CVP data. So I can still look at the CVP graphs. I just get one extra graph, the PVP graph. - - This first graph is the PVP graph. - - This second graph is the CVP - - I have found the PVP shows a little more detail at the lower velocities. - - - - - - - - - - - - -
  8. kevinstillwell

    shim changes and damping curves, based on dyno testing

    We've only done a couple specif tests with the rebound separator valve. We can basically do the same thing by turning the reb screw in until the 'reb bleed flow area' is about equal to your separator valve hole diameter. Of course, this makes the compression forces much, much stiffer.
  9. kevinstillwell

    shim changes and damping curves, based on dyno testing

    Kyle, Good thing we have inquiring minds and have determined the difference is the gas force. Basically, we wouldn't want to compare graph 2 with graph 3. They each represent something different. Graph 2 has the gas force and seal drag removed, graph three does not. This will throw the percentages and proportion off if trying to compare them against each other. I believe my initial mistake was using the dyno data that included gas force for this discussion. I should have kept it simple and removed the gas force and drag right from the start.
  10. kevinstillwell

    shim changes and damping curves, based on dyno testing

    grayracer513, WrenchDaddy, Mog, MRW, Bot of these adjusters have the small 2.15mm piston holes. Since they are both the same hole sizes, that is not a factor. And going from 2.15 to 2.75 hole size does not make even close to the differences we are talking about. And for the spring to make this much difference, it would have to be about 10 times stiffer.
  11. kevinstillwell

    shim changes and damping curves, based on dyno testing

    Darren, I want to make sure we are talking about the same thing. The data I am showing is from a PVP (peak velocity pickoff) test, and these numbers are then plotted on Excel. I left out the zero velocity and forces. For arguments sake, we will just say that at zero velocity there is 68 lbs (from the gas force). Here is a another look at the numbers. - - In these numbers, the gas force is not removed. - - Gas force = 68 lbs.
  12. kevinstillwell

    shim changes and damping curves, based on dyno testing

    This is a likely place to look. But it doesn't account for enough difference in force. We run cadj click range tests. In the case of a normal cadj, the difference between lc-10 out and lc-0 out (closed) is about 50 lbs. So that could be part, but not all of the discrepancy.
  13. kevinstillwell

    shim changes and damping curves, based on dyno testing

    Looking at the damping coefficient looks like a good idea. Kyle and I have discussed these damping coefficients before. As I mentioned, I haven't used them much, and lack an understanding of what they might show. I'll just lay out the force numbers again, and maybe ya'll can expand on ways to interpret them using the damping coefficient. - - This first table shows the compression force numbers and their differences. - - Notice the % diff column (the % diff isn't graphed). - - - - - - - - - - - - - This second table shows the damping coefficient as shown buy Clicked. He used an offset of 90lbs. I asked him why, but forget. Maybe he can expound on that. - - This graph shows a bit more distinction, making it easier to see the differences. - - But, the percent differences have changed. - - - - - - - - - - - - - Here is another damping coefficient table and graph, but it doesn't use the 90lb offset. - - The percent difference without the offset is the same as the first graph. This we could use some clarification. - - - - - - - - - - - - - - - - - - - - - - - - - -
  14. kevinstillwell

    shim changes and damping curves, based on dyno testing

    We usually run a complete series of tests for each shock. The default test is at clickers - - lc10 - - hc1 - - r10 We also run a low compression click range: - - lc18 - - lc10. - - lc6 - - lc0 Then we run a high compression click range: - - hc2 - - hc1 - - hc0 Then reb run a rebound click range: - - reb14 - - reb10 - - reb6 - - reb0 This gives us a good idea of what is going on. The drawback is, it takes about 3 hours to run this complete series. But it definitely tells you exactly what's going on in the shock. -----
  15. kevinstillwell

    shim changes and damping curves, based on dyno testing

    These are all good places to look. As you can see, if you really wanted to know exactly what was up with this adjuster, you would have to test each and every possibility, one at a time. Eventually, you'd either find the problem or die of old age, depending on how persistent you were. But now that we've brought up this 'bad seed' of an adjuster, it makes us want to figure it out, once and for all. Perhaps in a few days, when we get caught up, we can run some more tests. - - - - - - - - - - - - -
×