First few days ride experience

[Mark_H]

Hi guys, first post. Bunch o' questions!

Been riding with the iBike Aero for a few days and am enjoying the power experience immensely, however one issue that is conspiring against me is a certain type of road surface... that being the very poor condition surface where much of the tarmac has worn away and the stone chips are exposed. When riding over such surfaces the bike starts rattling away and power display often drops to zero even though the same, or indeed, higher effort is being made to get over the poor surface. This was a real problem during an FTP test a couple of days back because after 15 minutes of "busting a gut" I encountered such a surface and power simply vanished.

Does iBike compensate for these conditions during Route Analysis?

Also, while riding, if moving from an older asphalt surface onto a newly tarmacked one, there is an immediate jump in power, even though power from my legs is the same. Again, does Route Analysis recognise these sorts of transitions or does a change in speed (due in this case to lower road friction) always indicate user power input? Am I to assume from what I've seen that power is always going to be a little off depending on the road surface?

In the iBike software I have general settings set to automatically analyse downloaded rides. But after downloading, if I do a manual analyse route the results always change. Are these two separate analysis, or should automatically analysed routes give same results?

Regarding wind offset. Was out on a five hour ride today and checked wind offset regularly during the ride and each time it had drifted. I reset and an hour or so later when checking, it had drifted again. Temperature was very variable today with a 14 degree F shift during the ride. Do I need to be concerned about this? Do I *actually* need to recal during a ride or will the iBike post analysis take care of this drift?

Regarding the initial wind offset, is it ok to do indoors? I leave my unit outside to acclimatise to temperature and then stick it on the bike indoors to do wind cal before nipping straight out again on the bike. I figure there's no wind indoors so it's the best possible wind cal???

Could the software be updated to allow user selection of how temperatures are displayed? I display distances in metric, but want to show temperature in Celsius as I don't do Fahrenheit...

I guess that's enough for now

Thanks,

Mark

[Velocomp]

Please post a ride file.

Software should switch to Celsius when metric is chosen. Also, you can choose metric units in the Fast Start sequence.

Do your Cal Wind outdoors; just cup your hand over the wind sensor.

Don't worry about drift during the ride; it really does not contribute to meaningful errors.

The iBike assumes you are always on the same road surface type. This is generally a good assumption but in your case it sounds like you've got some significant variations.

When you are on a rough road frictional drag increases. Assuming that you've calibrated your iBike for smooth roads, the iBike doesn't know you've moved to a rough spot so could under-report power. When you return to the smooth road, where the iBike was calibrated, your bike speed jumps and the iBike reports higher watts.

[Mark_H]

Here's yesterday's ride. A 100k+ tour. Lots of road variation over that distance.

Re: Celsius - made a mistake - I ride with MPH but want to see Celcius...

Thanks for the answers.

Tyres are Continental Four Seasons (I hate punctures!) running around 100PSI front/110back.

Did my cal ride and coastdowns and let software calculate Crr - The Crr is higher than I see in discussions here but I put that down to the tyres tread and lower PSI than many ride. Unfortunately there was wind on the cal ride, but it's rarely calm around my location.

Cheers,

Mark

[Mark_H]

Another ride today that took in some poor surfaces (maybe this is a UK thing?!) and as soon as bike starts rattling over the surface power vanishes, even though I was clearly outputting watts, stomping on the pedals to get over the rough stuff. Also, noticed on cobbles the other day that the power vanishes too. As soon as the bike was on the cobbles and rattling, power vanished. At all times though cadence and speed output was present on the Garmin 705 so it wasn't like the sensor was mis-firing due to vibration...

Something clearly going on in ibke software... Ironic thing is that it's on these rough and tough surfaces where monitoring effort is more important because it's more difficult to judge from feel what sort of effort you are putting out because you are also contending with bike handling issues...

Is this something we can expected to be improved as right now it's undermining my joy of iBike ownership...?

[Velocomp]

I looked at the ride file you posted for August 18.

It would be helpful if you could point out some spots where you think watts weren't correct.

All of your calibrations look fine and there is no evidence of vibration problems.

I do notice one suspicious thing: your speed readings spike from time to time. This is suggestive of improper speed sensor and speed magnet spacing.

[Mark_H]

My cycling buddy, also an iBike user, sees the same issue over the same roads, so it's unlikely to be my bike, although he also uses a Garmin 705, so perhaps there's an issue there...

Attached is today's full ride and a snippet (borked.csv) from it containing sections where power is lost. I was riding with a power target of 200w and as you can see in this data there are sections where my cadence remains constant at ~100rpm and yet power drops away though I was in places adding more power to combat the poor surface.

Next time I go out over roads I know cause a problem I take some explicit samples.

Hope there's something useful in these files!

Thanks, Mark

[Velocomp]

On the flat and uphill portions of the ride your power output is roughly 200W. However, on the downhill sections your power declines significantly.

The downhill sections have slopes of -5% to -10%; quite a bit! Yet, your cadence is virtually unchanged...

While it certainly is possible on downhills that, as speed picks up, you are going into higher gears and maintaining both constant cadence and constant power, on steep downhills it is much more difficult to maintain a constant wattage output, simply because gravity is pulling down with a strong effect. For example, on the downhill highlighted below, the average downward watts caused by gravity is -623W; that is, gravitational forces are pulling you down with a force equivalent to 623W. For the iBike to measure any additional effort by your pedaling, your combination of gearing, cadence, (which remained constant at 100 rpm) and pedal effort would have to add to the 623W already pulling you downhill. This is not an easy thing to do; the iBike is reporting no additional wattage from pedaling.

Since your reported perceived effort is constant, apparently including downhills as well as uphills, we will look at your files further, as something else may be going on. We do not think this is an issue of rough roads.

[Mark_H]

Just to add to this, I lose power in exactly the same way if riding those sections of road the opposite direction, ie generally uphill... although infact I have to increase my power to maintain pace in that case...

Next time I'm out over there or over the cobbled areas I have to ride I'll take some specific snapshots for you to look at.

Cheers,

Mark

[rruff]

Something clearly going on in ibke software... Ironic thing is that it's on these rough and tough surfaces where monitoring effort is more important because it's more difficult to judge from feel what sort of effort you are putting out because you are also contending with bike handling issues...

Power declines on my unit when I hit rough surfaces. Because the Gen3s recalculate the tilt using the last 5 minutes of data, after 5 minutes of this it will sort of correct, but it won't be accurate because the degree of roughness and it's effect on the unit is highly variable. It will average out over longer time spans though. If you ride on rough pavement and then move to smooth again it will take it 5 minutes to sort it out and give accurate readings again.

Since good quality tires and tubes provide the most suspension on a bike, this might be something to look at. Also, lowering pressure might help.

[Mark_H]

Something clearly going on in ibke software... Ironic thing is that it's on these rough and tough surfaces where monitoring effort is more important because it's more difficult to judge from feel what sort of effort you are putting out because you are also contending with bike handling issues...

Power declines on my unit when I hit rough surfaces. Because the Gen3s recalculate the tilt using the last 5 minutes of data, after 5 minutes of this it will sort of correct, but it won't be accurate because the degree of roughness and it's effect on the unit is highly variable. It will average out over longer time spans though. If you ride on rough pavement and then move to smooth again it will take it 5 minutes to sort it out and give accurate readings again.

Interesting. And good to know!

Since good quality tires and tubes provide the most suspension on a bike, this might be something to look at. Also, lowering pressure might help.

As noted above I do run my tyres softer than many riders. This is mainly a compromise between performance and wet weather riding. Tyres are decent enough (Continental Grand Prix Four Seasons); I don't like punctures so give up a bit of performance for their superb puncture resistance and they have great wet weather handling.

Cheers,

Mark

[crazymadcyclist]

W

[Mark_H]

Ok, today's ride took in some short cobbled sections and, as promised, I took the time to capture some data. I reset trip before and after to capture the effort most accurately and attach it below for inspection.

Things to note: the road is *FLAT*. I accelerated off normal road surface onto the cobbles. You can see the power applied in the file. Then I had to stop peddling for a pedestrian before once again picking up the cadence. Next you can see in the file the section with maintained cadence and pace and ZERO POWER, even though more power was being applied to keep pace up over the cobbles. Cadence stops again at the end because I popped a bottle and had to recover it!

The slope data presented by the file is inaccurate. As noted above, the road is flat, and you can see it here of Street View:

http://maps.google.co.uk/maps?f=q&sourc ... 02706&z=19

It would seem that once the bike starts rattling, slope goes a bit "mental" and iBike assumes the bike is descending and thus gravity wipes out the power?

OK, this is an extreme example of rough road, but as noted above I've seen the same behaviour on normal poor road surfaces. Hopefully something useful in the data, and hopefully something "smart" can be done in software to spot such instances and compensate...

Oh, and make sure you view the original file prior to any analysing!

Cheers,

Mark

[rruff]

The newer iBikes are a big improvement over previous models, and I can't think of a way they could improve it further unless... maybe... they incorporated even more sophisticated accelerometers that were able to filter the vibration. They used to sell a damping device that fit inside the stem, so maybe something like that would help.

[Mark_H]

The newer iBikes are a big improvement over previous models, and I can't think of a way they could improve it further unless... maybe... they incorporated even more sophisticated accelerometers that were able to filter the vibration. They used to sell a damping device that fit inside the stem, so maybe something like that would help.

I guess it all depends on what's in the data. If the slope data shows abnormal swings (due to the vibrations) perhaps that can be taken as a sign of rough surface and coupled with the presence of steady cadence and steady pace it might be possible to interpolate the numbers a bit better (and certainly in the post processing of the data I supplied the iBike software does find power...) If, on the other hand, the hardware is simply presenting the slope data as fact then I guess there's little that can be done because doing so would affect those sections that are truly downwards and where power should be zero.

[Velocomp]

Please post a more complete ride file.

[Mark_H]

Full ride attached. I have reverted it to original data, so there is no post-processing in this file as far as I know.

Mark

[Velocomp]

I'm looking at the full ride now. Please tell me where in the full ride file your cobble-tastic portion is...

[Mark_H]

John, it should be in the joined section:

2:06:15 -> 2:07:25

Again, make sure that you revert to original file to reveal collected data.

Cheers,

Mark

[Mark_H]

John, did the data reveal anything useful??

Cheers,

Mark

[Velocomp]

Sorry, I'm just catching up on this thread now. We kind-of were interrupted by Interbike and iBike Dash activities.

The environment in which the accelerometer operates is extremely noisy. The noise is caused by road roughness and, the rougher the road, the more noise there is. The accelerometer and the electronics need to sift out the slope and bike acceleration data from all the noise generated by the road.

In the "old days" of the iBike rough roads could cause crazy-high watts. Unfortunately, at higher speeds the roads did not have to be all that rough for the problem to occur (a fast pace on a good chip seal road might set things off...). Fortunately, that rough roads problem was eliminated years ago; the iBike works very well on the rough roads situations encountered by most cyclists.

The photo link you provided was very useful. Your situation suggests that the severe roughness of cobblestone roads may still present a challenge. Interestingly, watts going down on severely rough roads is quite different from our past experience; this is new to us and we don't have an immediate explanation for it. We will investigate.

In the original days of the iBike we said specifically that the iBike might provide inaccurate readings on rough roads such as cobblestone surfaces. After making improvements in the Gen III we thought that this language no longer applied so we removed it.

Your example shows we might have to add back the cobblestone notice.