Low Power Readings on rough surfaces-please help

[smhagger707]

I've been using my iSport for about a month now, and have noticed that the power readings seem to drop 30-70W when I hit a bump in the road, and drop off even more when I ride continuously on a poor road surface. I have the green sticky pads installed the way the directions say, and the unit tightened down snugly (to the point it feels very firm and doesn't feel like it will move much, but not overtightened). I'm starting to get frustrated as it really makes doing workouts or holding power during a power test very difficult due to under-reported power. On smooth roads, it works great.

One example I can recall was when I was on a group ride taking turns to pull. The first pull I did was on relatively flat ground (11' of climbing for the interval, no descending), but directly into a strong headwind. I averaged 220W, with a max of 283W, and an average speed of 15.1 mph (did I say it was windy?), average HR for the pull was 171 bpm (max of 173 bpm). A little while later, while riding on a pretty rough road, still heading directly into the wind (East), I did another pull (still relatively flat with only 7' of climbing, no descending): average speed was 17.2 mph, average HR was 174 (max of 176 bpm). Power for this second interval: average power = 128W, max power was 267W. The max looks about right, the average reflects the typical power number that was seen while riding (the 128W number was straight from the Garmin data for the average power for that lap). Thus, in this example, the average power was 90W lower for a similar pull, but on really rough pavement (I estimate I was probably holding more like 240W-250W for the interval based on how badly my legs burned while doing it.

Since I have an iSport, I don't have the iBike data to send, but I can provide a .tcx file if it helps at all (probably not).

Any ideas on what's going on? I really want to like this unit, but it's driving me nuts when my power drops off for no apparent reason (thus skewing my averages-and probably my maxes as well)-just because I hit a bump in the road.

More details:

Combined bike + rider + gear weight: 158 lbs
iSport (iBike Gen III) GT, purchased April 12, 2010
Cst (Aero) = 0.222
Cst (Fric) = 4.541
Data recorded on a Garmin Edge 500

Tires = New Conti GP4000s @ 112 psi
I don't know how to get the Crr or other coefficients from the iSport


Help please-any ideas?


Sean

[smhagger707]

Update: I decided to redo my 'other' profile by redoing the 'fast start' data, tilt cal, wind cal, and cal ride. The wind was nearly calm this time when I was doing my cal ride (of about 2.5 miles total). I made sure I hit some of the regular bumpy surfaces rather than trying to stay on smooth surfaces. My cal numbers ended up being: Cst (Aero) = 0.217, Cst (Fric) = 3.867, Tire press = 106 psi. My power readings did seem to be somewhat impervious to bumps after this cal ride (I don't know what would have changed to 'fix' it-any ideas?). Not wanting to be too optimistic here, so I'll go for a ride tomorrow morning to see how the readings look, over the course of a full ride.

Question: what would happen to my power readings if my coefficients went to Cst (Aero) = 0.217, Cst (Fric) = 3.867, from the 'other cal' that I have that I've been trusting at Cst (Aero) = 0.222, Cst (Fric) = 4.541? I'll just be glad when I can get this thing dialed in where it isn't so 'jumpy' and I can believe the power numbers based on the burn in my legs (and not see the data dive when I hit a bump-how discouraging).

I'll report back on what I learn tomorrow...

[Velocomp]

Rough roads present more frictional drag than smooth roads.

It looks like you are using the Fast Start sequence to set aero and fric.

The default fric coefficient used in the iSport is for average smooth roads (Crr = 0.005), so on rough roads your iSport will read low.

To account for rough-surface roads you can do one of two things:

1) In the Racer sequence, go to Est Fric/Yes and set Crr to a higher number, more representative of rough roads, say 80. Store this rough-road setting in Profile 2. You will find that your fric number will increase significantly, causing watts to be higher.

During your ride, when you are on smooth roads use profile 1 and when you get to the rough road section switch to profile 2.

2) Alternatively, you can measure aero and fric for rough roads using coast downs. This MIGHT give you better results but it's a tedious process with the iSport. If you want to try it here's how:

a) keep your smooth road profile in profile 1

b) set your iSport to profile 2

c) go to the Racr sequence and set Est Fric to No (when set to No the iSport will measure both aero AND fric)

d) Do 4-5 coast downs on your rough-surface roads. Read the coast down instructions attached. Since you have an iSport you'll need to write down your aero and fric numbers after each coast down.

e) average the results of your five coast downs. If the average aero and fric numbers you calculate are close to the aero and fric numbers displayed in your iSport (the numbers from your final coast down), then you're done. If the average aero and fric are substantially different from the displayed aero and fric (meaning that your final coast down was "off" for some reason), you can "trick" your iSport into having the average values, even though you can't use the iBike software to load them automatically.

Let's suppose, for example, your averaged coast down values are aero = 0.24 and fric = 6.72. The values stored in your iSport are aero = .28 and fric = 3.8.

So, for this example, you'll want your iSport to show aero = 0.24, and fric = 6.72

f) To change the value of aero, go to the Est CdA step in the Fast Start sequence and enter your height, weight, and riding position. At the end of the Fast Start sequence exit setup. Go to the totals screen and check aero. You'll see that aero has been changed to something else (the Est CdA step overwrites the value of aero from your last coast down). If aero isn't 0.24, then play with the height and weight settings on Est CdA to get your aero number close to 0.24.

g) To change the value of fric, go to the Est Fric step in the Racr sequence. Set Est Fric to "yes" and click the center button. You will see a fric number (a Crr number). Increase this number, then exit setup. You'll see that fric number has increased to something else (the Est Fric step overwrites the value of aero from your last coast down). If fric isn't 6.72, go back to Est Fric and continue to change the number until your fric is close to 6.72.

Once you've done all this your profile 2 will have aero and fric numbers determined from your averaged coast down results. Use profile 1 for smooth roads, profile 2 for rough roads.

This is a complicated alternative. If you had an iPro you could do everything automatically with a few computer clicks.

[smhagger707]

Thanks for the info John. I updated the Fric coefficient as you suggested in step one (set to 'yes' and '80'). This seemed to help the power from dropping off so much on the bumps. Now my coefficients are as follows: Cst (Aero) = 0.217, Cst (Fric) = 5.659. Should I try to tweak my 'Aero' to get it higher as well (does .217 seem too low-I'm 5'4", 131 lbs, riding a 52 cm bike with 40 cm c-t-c bars)? Also, the power numbers seem more reasonable now, even on 'relatively smooth' roads-is there a downside to keeping my 'Fric' at 5.659 (is that too high or too low for a good, all-around coefficient)?

When I get the money, I'll upgrade to the iPro-I just wanted to make sure that the settings can be changed so that the power doesn't drop off each time I hit a bump (should I tighten or loosen the clamp mount, I wonder?).

Thanks for your assistance-I'm moving in the right direction it seems...


Sean

[Velocomp]

Your aero is fine; you are a small and very light person!

You bring up a good point about your mount. Make sure that your mount is tight enough so that it cannot rotate AT ALL on the handlebars. The double-sticky tape helps in this regard. Tighten the screws enough so that the mount doesn't move, but don't over-tighten or the clamp might break.

[smhagger707]

Your aero is fine; you are a small and very light person!

You bring up a good point about your mount. Make sure that your mount is tight enough so that it cannot rotate AT ALL on the handlebars. The double-sticky tape helps in this regard. Tighten the screws enough so that the mount doesn't move, but don't over-tighten or the clamp might break.

Yes, I'm pretty small for sure.

I just checked my mount and it feels secure, but I can get it to rotate slightly if I apply some torque to it. I'll try tightening it more to see if that helps (that's a good way to describe it BTW-not being able to rotate the mount...).

The power readings seemed pretty good from today's ride (after I recal'd the tilt meter), with the new coefficients (0.217 / 5.659.


Sean

[lorduintah]

The sticky tape from 3M works OK (you may have run across this stuff with many other possible applications and uses besides this particular one), but if you do use it, you really have to be extra carefu,l as noted, about tightening the screws too much. Plastic with that torque or deformation pressure can snap easily.

I ended up with a strip of an old inner tube cut out and put that under the top of the mount and let the bottom clamp directly on the handlebar. This is probably working as well as the tape and I know I do not get any movement with the iBike mounted.

Tom

[smhagger707]

The sticky tape from 3M works OK (you may have run across this stuff with many other possible applications and uses besides this particular one), but if you do use it, you really have to be extra carefu,l as noted, about tightening the screws too much. Plastic with that torque or deformation pressure can snap easily.

I ended up with a strip of an old tube cut out and put that under the top of the mound and let the bottom clamp directly on the handlebar. This is probably working as well as the tape and I know I do not get any movement with the iBike mounted.

Tom

Tom,

Thanks for the reply. You bring up an interesting solution-I'll definitely try that.

This discussion does bring up an interesting question though: are the 3M squares supposed to provide damping for the iBike or are the there to lock the mount down tight so it can't slip or move? Initially I thought it was to provide damping to the iBike, but based on the comments by John and Tom-regarding tightening down the clamp so there is no possible movement, it doesn't appear that one would want any damping (or perhaps the 3M strips provide high-frequency damping only?). If no movement is desired, and damping isn't one of the reasons for using the 3M squares, then it seems like mounting directly to the bars would be the way to go (I use carbon bars, BTW).

So I guess my question it: what is the real purpose of the 3M tape strips?

[lorduintah]

It is to keep the iBike and mount from rotating on the stem or handlebar - whichever type you have. I think the Velocomp dudes found the 3M tabs to be an easy solution to securing the clamp - and leave little doubt to anyone about rotation, if the owner actually used them. Mounting directly on the bars will NOT prevent rotation - metal or carbon bars inclusive. Tie-wraps would not offer the same grip either - so you need a little cushion and just enough tension on the clap - but too much tightening and you buy a new bottom clamp from Velocomp.

The vibrations are going to be there just from riding - unstable tilting, if the clamp is loose, will wreck any profile and any data you take with a mount that is not secure.

Tom

[smhagger707]

Update: I just completed my first significant distance ride after tightening down the clamp further and I have to say that, overall, the problem of power dropping off seems mostly resolved. The only reason I say 'mostly' and not 'totally' is that I did notice it would still drop off when I would hit 'major' bumps in the road. Those type of bumps, however, are understandable as they sent considerable shock through my hands and interrupted my pedal stroke to the point that I may have momentarily dropped power output anyway-so I'll call it too close to call.

The main thing that I noticed was that my power wouldn't drop off 50W every time I'd hit a crack in the road (a relatively small surface imperfection of any type). The result was my average power coming up, for a solo ride (averaging 17.5 mph, for 29.7 miles, 802' Ascent, avg power 151W) 21W over a comparable solo ride (averaging 17.8 mph, for 39.4 miles, 1,201' ascent, avg power 130W)-actually I think it would have been more like 25-28W higher if I had averaged the same average speed and did the same amount of climbing and distance for both rides.

The one thing I did notice was that my power seemed a little low when I went to sprint tonight. I don't know if this from the Cst (Aero) being down to 0.217 vs. 0.222 or if it was from my legs being shot from a weekend of nasty climbing. I've now gone back to my other profile and have updated my Cst (Fric) to nearly the same setting (as close as I could get) as the profile I used tonight so the only real difference that I can see is from the Cst (Aero). Interestingly: I had to set the fric on profile 2 = 0070 to get nearly the same result (Cst Fric = 5.623) as profile 1 on 0080 (Cst Fric = 5.659). Anyone know why?

I'll keep watching it as I ride more over the next couple weeks to see if I'm as satisfied then as I am now--but so far I'm a happy camper.

Thanks for all the help.

[rruff]

You need to look at your CdA and Crr values... the other numbers depend on individual unit calibration and so are not comparable.

Also, the vibration/tilt phenomena will *not* effect the average power for a ride. The auto-tilt adjustment sees to that. So if you are seeing higher averages, something else is going on.

[smhagger707]

You need to look at your CdA and Crr values... the other numbers depend on individual unit calibration and so are not comparable.

Also, the vibration/tilt phenomena will *not* effect the average power for a ride. The auto-tilt adjustment sees to that. So if you are seeing higher averages, something else is going on.

How do I view those numbers on an iSport (or can I)?

Be patient with me: I would think that with the power numbers 'dropping down' as frequently as they were before from hitting even slight road 'imperfections' that the average power figures actually *would* get affected (it happened a LOT-hence the reason I posted this thread in the first place)? Honestly the slight increase in the Cst (Fric) values (from 4.541 up to 5.651) doesn't really seem to have bumped up the instantaneous power numbers much that I've noticed--but I suppose they could be having an 'upward-shifting' effect on the power averages as noted above. Honestly though-just not having the numbers 'dropping down' all the time really seems like the reason that the average power is higher now (just 'gut feel' from watching the real-time ride data).

Also, looking at the data, it appears that the data is less 'jumpy' now as well-perhaps I'll post a screen shot of the power data from a segment of the same road for comparison to illustrate. *edit: I just looked at data for the same section of road (1 mile's worth of data) from the 'old' setup compared to last night's ride and the different is noticeable in the 'jumpy-ness' of the data for sure. It also shows up in the numbers: the 17.8 mph ride from last month had an average speed for the 1 mile (warmup) section of 13.9 mph and 79W average power with a LOT of data jumping up and down. The ride from last night over the same 1 mile warmup section had an average speed of 14.5 mph and an average power of 125W-that's a 45W jump and the data looks a whole lot smoother (like I would expect) from last night's ride. Yes, a small portion of that 45W difference (maybe 10-15W is from the higher average speed)-but the rest is clearly from the 'jump-ness' of the data from the power numbers constantly dropping down from hitting bumps. I did also notice my cadence data seemed to drop-down a lot from the earlier ride so perhaps that was also driving combining to cause the power numbers to 'drop' in some cases.

Ultimately I wish I had access to all the coefficients via software to make this whole 'setup' process easier--I'll have to upgrade to the iPro, iGuess?

[rruff]

If vibration is/was effecting your unit, then it would record low during bumpy sections because it thinks the slope is less than reality, but it will compensate by auto adjusting the tilt after 5min of riding. For this reason the vibration will not effect your average during a ride.

I noticed you are running GP4000 tires at 112 psi and only weigh 158 lbs with bike and everything. These tires do not ride particularly smoothly, and that pressure is much higher than ideal. I run the same tires with 90f and 95r psi, and my total weight is 196lb. I'd recommend dropping the pressure to at least as low as I'm using, or lower until you feel the ride smooth out. It will be faster as well as more comfortable, and your possible vibration issues will likely disappear.

[smhagger707]

OK, i dropped the tire pressure to 95 psi on both tires and used the 'other profile' (profile 2)-the one I *didn't* use last night to see if adjusting the estimated Fric up to about the same Cst (Fric) value as profile 1 would give me what I was looking for: the resistance to wattage dropoffs when hitting bumps, but the peak power responsiveness that profile 2 was giving me (profile 1 just didn't go up as high as I think it should have for the effort I was putting out, and compared to wattage readings from other riders, adjusted for weight, when sprinting with them up hills). The result? It still had poor performance over the bumps. I guess that's good in a way though as it means the mount tightness wasn't a factor after all (or at least not 'the' factor), but the difference really lies in the coefficients. Now, if I could just simply *see* what those coefficients actually are in both profiles, I could get more insight into what is affecting what: for instance which coefficient is responsible for my peak power during short burst sprints (10 sec or less) being too low--is it CdA where the wind resistance at those speeds is not properly adding to the forces overcome? Which coefficient is responsible for the 'resistance' to wattage 'dropouts' when hitting bumps (is it Crr)? How can I get to those parameters in an iSport so I can actually get believable, consistent readings and use this thing like I originally wanted to-or am I going to be forced to do the upgrade just to simply get access to manually tweaking the coefficients?

I guess in the end, I'm finally seeing enough 'potential' in this device to make me keep trying to get the calibration dialed in, rather than just ditching it altogether and building a PT wheel...

I suspect the calibration I need lies in the 'best' combination of the two profiles--I just need to get access to them. Any suggestions?

[lorduintah]

A significant change to tire pressure requires re-calibration. You cannot just put a data set from another tire pressure into the mix.

Tom

[rruff]

Actually, changing the pressure will have a trivial effect on the Crr in this range, and since the unit automatically adjusts tilt, that won't matter at all.

I'd suggest using the iBike's set coefficients for Crr and CdA... not sure how you do that but it is one of the cal routines. Then you just do a cal ride to get your tilt and wind scaling. You can repeat this as often as you like to see if the Fric and Aero values are consistent. Then put your unit on the "coast" screen and do some coast-downs at various speeds and conditions. If it reads close to zero watts +- then you are good. If it always reads high or low, then you know something is off.

[smhagger707]

Ok, so after a while of riding now, it seems that I have my calibration dialed in to be pretty consistent with minimal drops in wattage readings when hitting bumps (like the watts will still drop off, but for only 1 second or so, instead of 'hanging' down there for 3-4 seconds like it used to do). This week I've decided to take the plunge and upgrade to the iPro + trainer option (which should arrive today or tomorrow).

My question: am I correct that, when I actually perform the upgrade on the unit, that data recording will just 'unlock' and my current calibration numbers won't be changed? This is VERY important to me as I really need to know where ALL my calibration values are set to currently as I now have months worth of data logged (via my Garmin) and I would like some way to know if those are all 'off' and be able to estimate by how much they are off. I realize I can't 're-apply' any new cal numbers to my current legacy data since it was just 'Watts' recorded by the Garmin, but I can record the current parameters, then perform whatever calibration I need to do on 'profile 2', and ride with those parameters. After collecting data with the converted iSport (to iPro) using the new parameters (I'm assuming they may still be a little off-it would be great if they weren't-but I'll post a ride file as soon as I have one so John or Coach Boyd can confirm my parameters for me), I can 'revert' some of the new data to the 'old' coefficients (that I'm using today) and see what happens to my watts: average, peak, etc. and get a better understanding if my values are currently being under-reported, over-reported or are generally 'right on'.

I'm looking forward to upgrading but am a little hesitant about loosing my current coefficients / parameters in the process. Please, anyone who's upgraded: alleviate my concerns for me...

Thanks,


Sean

[racerfern]

None of your calibration values or settings will change when upgrading. It's as simple as that. The biggest difference is you will be able to download your data. However, the power numbers will not change because of a firmware upgrade.

It is important to note that now that you're going to be able to download files, you'll be able to analyze them and you might come up with slightly different wattage after analysis. However it won't be enough to throw off any pattern.

Enjoy your upgrade.

[smhagger707]

Ok, so I've now upgraded to the iPro and have a ride file attached. A few points about my current configuration: I've now been riding with my tire pressure set to 92 psi (front) and 97 psi (rear) and I mainly ride on the hoods. Up to the upgrade, I've used a crr = 0.0080 as suggested in this thread, and that has helped. Now, after the upgrade (and firmware update?), my power seems WAY more stable in terms of not dropping off. However, I think it's too easy to hold higher wattages as they seem to 'float' high for a time.

In the attached file, I've used a Cst (aero) = 0.222 (resulting in a CdA = 0.318), and a crr = 0.0080 (which is what I was using before the upgrade to the iPro). I now question whether the higher crr is needed as the power drops (the reason I raised the crr in the first place) seem largely eliminated with the newer firmware (I'm assuming). I still question whether my CdA is too low. Playing around with the 'Tweak CdA, crr, Cm' feature (what is 'Cm'?) I tweaked my CdA up to 0.355 and my crr down to 0.0055 and the values seem 'correct'-but all sorts of combinations seem 'correct'. Thus, now knowing where to go with the coefficients (and thinking my crr should be somewhere between 0.0055 and 0.0080), I've decided to just post a file and let you take a look.

I should also add that I ride mostly on chip sealed surfaces. I would say the distribution of the roads I ride on would be something like: 6 month or older chip seal = 70% (some cracks / bumps in road, mostly smoothed out), new chip sealed roads (VERY rough, but not quite gravel) = 10%, New asphalt = 15%, concrete 5% or less.

Please HELP!

How do I know if a CdA of 0.318 or 0.355 is what I should use? Should I now do another calibration ride? Should I do some coast downs on various road surfaces (and how do I average them)? If doing coast downs, do I turn est crr 'off' or 'on'? Do I try to borrow my friends PowerTap and correlate the iPro to the PT output?

Confused and needing help.

Thanks,


Sean

[smhagger707]

I should also mention that, over the course of the last two months, I've had patellar tendon knee pain / popping (due to riding the horribly hilly hundreds ride and kneeling too much, the later being what I'm really tracing the problem back to). As a result, I've had to raise my saddle by about 5 - 6 mm and raise my bars accordingly to compensate.

After reading several posts, I'm starting to believe that I just need to 'start over' and do the fast start, cal ride and coast downs to re-baseline.

Also, I've switched to Conti GP4000s ('S' version) so that may also have an impact on my calibration.

I'm kind of bummed because I feel like I'm back to the place I was months ago: not really knowing what my power output really is-steady state or for sprinting. This also means my critical power curve in Golden Cheetah is also likely junk. In any case, I did seem to get a somewhat repeatable number for power output over the last two months based on a given effort...

So it looks like I'll need to start over completely, scrap my CP curve in Golden Cheetah, and move forward with a new power / fitness test, etc. to see where I'm really at (I hope it's not lower than believed...).

[lorduintah]

If your power calibration was correct before you had to make all of these changes, then after a new calibration you should really expect the power to get back to the same. It is this period of in between that is probably the only portion of your riding where there may be some differences.

So your CP may be back in line, if it was during the good values.

Tom