Excessive speeds causing too high powers?

[doib]

Hello John,
some time ago you advised me to use a sensor with two magnets on the wheel to have precise power measurements on very steep climbs at low speeds (I have a PPV3 and a 29er MTB). Well, I did as you suggested, inverting the sensors between the mtb and the road bike and, of course, halving the wheel size for the mtb profile. I also redid the calibration runs, without problems; on the mtb I mounted the garmin gsc-10 (I know it is not recommended, but on the road bike it always worked well). The solution worked very well, except that after each race, isaac asked me to save the two profiles again, so i have a long series of equal profiles saved ... I come to today's problem. I started with the MTB and the Garmin 820 computer was showing me excessively low speeds. I immediately noticed that one of the two magnets was moved. As soon as I put it back in place, the speeds on the garmin come back correct but the power data become very high. In order not to save on strava races with exaggerated powers data that could affect the future training, I turned off the pp. I then tried to turn it back on several times, but the problem persisted. I also made an attempt at recalibration, but from the files that I downloaded I would say that it was not successful. May be I messed around with my smartphone, trying to intervene on the profiles with the power house app. To close the (long, sorry) speech, once downloaded the files I noticed that the speeds reported by the powerpod are very high, I would say about double the real ones. I attach some files (there are five). I wonder if the error was caused by the initial problem with the magnet. Should I reset the powerpod? If so, should I redo the profiles or can I recover them from the saved rides?
Thanks and sorry for the long speech.

Davide

iBike_05_23_2020_1126_2_km.ibr

(21.17 KiB) Downloaded 197 times

iBike_05_23_2020_1041_2_km.ibr

(24.24 KiB) Downloaded 184 times

iBike_05_23_2020_1049_3_km.ibr

(32.92 KiB) Downloaded 184 times

[Velocomp]

Please post a ride file using the two magnet solution, from a time where you did not experience this problem.

[doib]

Here it is.

iBike_05_19_2020_1843_23_km.ibr

(245.51 KiB) Downloaded 212 times

[doib]

Hello, waiting for a reply to the previous post, an update, or rather three.
1. The problem with exaggerated power readings has been temporarily solved by adjusting the position of the magnets (I think it also created problems with the cadence). Until yesterday. Today, although sensors and magnets seemed to work well, I had recorded powers even over 2000 watts (not visible because I turned off the pp and probably the rides were too short to be recorded. The peaks are however visible on strava). I attach the two files of today's ride (partial, I turned off). The speeds seem almost double than the real ones, despite having halved the diameter of the wheel (29er mtb).
2. every time I connect the pp to isaac it asks me if I want to overwrite a calibration run (which I attach), and then starts the procedure for the formation of a new profile (n. 3, which I use for the road bike ). As a result, I have a long series of unused profiles in isaac.
3. On the road bike, on which I moved the Garmin hub and crank sensors, the powers are now very low. In this case, comparing the rides with the corresponding strava or garmin connect files, the speeds recorded by pp are much lower. The Garmin 820 says that the sensors are working properly and the batteries are ok.
What can I do?

iBike_06_02_2020_1038_2_km.ibr

(28.27 KiB) Downloaded 175 times

iBike_06_02_2020_1048_1_km.ibr

(30.31 KiB) Downloaded 192 times

iBike_05_29_2020_1730_1_km_CalRide.ibr

(14.88 KiB) Downloaded 168 times

[Velocomp]

I looked at your previous ride file but forgot to respond.

Your profile is set for 2 magnets (wheel circumference around 1100mm) but your speed is too high. This suggests that your speed sensor was sending pulses 2x the required amount. If your magnets are improperly positioned this could result in spurious speed sensor signals, as your most recent post suggests.

How exactly are you downloading files in Isaac? There is no reason it should be asking you to overwrite calibration unless there is some sequence of commands that is prompting this.

Is it possible that you have your MTB and Road bike profiles reversed (or at least, the sensor connections)?

[doib]

I think the magnets are placed well, because the garmin measurements are correct.
To download the files in isaac, I wait for the automatic download window to close by itself and start the transfer. As for the inverted profiles, I thought about it too, but also checking the profiles on the race files they seem right.
yesterday's rides are associated to the epic GSC10 DATA 09_05_2020, where the wheel measure is halved (1150 mm) and the sensor has the same code for speed and cadence, so it should be the gsc-10. For the road bike I used this: Road Bike 01/06/2020 Prof # 3, which has two distinct sensors. I attach the file.
Thanks so much

[doib]

Hello John,
more updates from my endless story. Following the latest events, I decided to reset all the profiles and start again, correctly associating the sensors to the respective bikes. In the beginning, it seemed to be working. Last Sunday I took a road bike ride of about 55 km, and the distance and power values are absolutely consistent (even compared to what is recorded by the garmin 820). Today I took the same bike, same powerpod and I did a 74 km race. The pp recorded a run of about 40 km, with very low powers and speeds. The recorded speeds and distances do not match the strava file. The profile, in the two rides, is the same and the circumference measurement corresponds to the real one (I checked with the analyze device setting command). The sensors are the hub and crank garmin and on the garmin 820 they return the correct values. At this point, I think that the device has failed. Or could this be an isaac problem? I am a little disappointed, for me to send the PP for assistance would be a big problem, because I bought it in the United States, from the clever training shop and already the first shipment to Italy cost me € 65 of customs duties .. What can I do ? I enclose the two races (the correct one and the wrong one, complete with screen of strava with the right numbers), so you can please take a look and suggest a solution.
Thanks, see you soon.

[Velocomp]

I am really, really confused about your setup.

Please provide the following information:

1) Your bike tire size
2) what kind of speed sensor(s) you are using (photo please)
3) Do you have more than 1 speed sensor on your bike? If so, what sensor is paired to Garmin? What sensor to PP?
4) The settings on your Garmin for speed sensor (is it manually set by you, or automatically by the Garmin)
5) The settings in Strava. Where is Strava getting its speed values?
6) Please provide all 4 profiles stored in your PP

Your 74km race shows ½ the correct speed, which implies that either its profile setting is incorrect, or if you are using a magnet speed sensor, that one of the magnets is not pulsing the sensor.

There is no need to return your PP.

[doib]

Thanks for the quick reply, I will try to be clearer. Perhaps the confusion arises from the fact that I no longer specified that the bikes are two, because I had already done it in the previous discussion and I took it for granted. I apologize for this.

Point 1
Bike associated with profile no. 2: mtb, tire size 29x2.3 "(measured 1150 mm) with Garmin Gsc-10 sensor and two magnets
Bike associated with profile no. 3: road, tire size 700x25c (isaac list) with garmin hub speed sensor and gramin crank arm cadence sensor

Point 2
See previous point. I add photos of the bikes where you can see the positioning of the sensors; later I will try to put better ones (now I'm at work and I only have these available).

Point 3
No, each bike has a single speed sensor paired to the Garmin 820 and to the PP.

Point 4
Profile # 2 (mtb 29x2,3) manual insertion with the same size inserted on PP (1150 mm)
Profile # 3 (road 700x25c) automatic insertion on Garmin 820 and chosen in the list on PP.

Point 5
Strava shares the tcx file loaded on garmin connect by the garmin 820 (I think in part you process it, but the speed is certainly the one read by the garmin, as well as the power values).

Step 6
The pp includes only 2 profiles relating to the two bikes described (which I attach), plus two more that are the same but outdated (I changed my weight). I haven't set 4 different profiles, because I only have three bikes and I still don't know if I will equip the one I use for travel with sensors.

The strange thing is that, when I set up the profiles and perform the calibration, everything works well for a ride or two, then the problems recur. With the road bike I always read lower power values, with the MTB very high values (often over 1500 watts).
Thank you.

Davide

[doib]

Bike photos. On the road bike, the sensor is on the front wheel hub; the mtb has the combo sensor with the two magnets.

[Velocomp]

Please use Edit/Edit Profiles/Extract from Device to confirm that there is nothing stored in profile 1 and 4.

If something is stored in these profiles, please post these profiles too.

[doib]

First 2 profiles. Profile #2 is the mtb one.

[doib]

Profiles number 3 and 4. The #3 profile is the road bike one. This evening, when i download the weekend files, isaac ascks me if i want to compose the profile number 3 (but it should have already been composed...). When I download the profiles, Isaac sends the message in the picture attached.
If you need more information, I'm here. Thank you, bye

[Velocomp]

Well, this post may demonstrate that the saying "You can't teach an old dog new tricks" might not be completely true (at least, for me)

In the past 15 years (which is pretty old in the power meter business) I have never seen anything like this. I have stared and stared at your ride files and puzzled about what you are seeing:

1) On your MTB files, sometimes the bike speed is 2x the expected amount
2) On your road bike, sometimes the bike speed is ½ x the expected amount

Very strange!

To prove this to myself, I took your MTB ride file 05232020_1049_3km, and used the command Analyze/Analyze Bike settings, to change the wheel diameter to 575mm (½ the stated amount of 1150mm). The modified file looks perfect

I took your "wrong" road race file, 74 km, and used the Analyze/Analyze Bike Settings command to change the wheel diameter to 4210, twice the indicated amount. Again, this file looked perfect.

So, it appears that inside of PP, the wheel circumference values are getting scrambled. How could this be?

I think you have provided three crucial clues: "1) Every time I connect the pp to Isaac it asks me if I want to overwrite a calibration run (which I attach), and 2) then starts the procedure for the formation of a new profile (n. 3, which I use for the road bike ). 3) As a result, I have a long series of unused profiles in Isaac."

This sequence of events should not happen.

I saw, exactly one time earlier this year, a peculiar bug in Isaac that was causing wireless IDs to be copied over from one profile to another profile. This of course is not supposed to happen...

I am somewhat speculating here, but I believe what is happening to you is your wheel circumferences are being copied over from one profile to the other. If your MTB 1150 circumference is copied into your road bike, then your road bike speeds will be 1/2 the correct value; if your 2105 road bike circumference is copied into your MTB profile, then bike speeds will be 2 x.

I know that your profiles indicate the correct circumference values for each bike. But mysterious things appear to be happening in Isaac data transfers...

So, here is what I would like you to try:

1) Use Edit/Edit Profiles/Delete to delete all the extraneous profiles you have. Make sure to click Accept after deleting them

2) You should still have in your Edit Profiles list your road and MTB profiles in your profile list

3) Connect PP To Isaac, and use the command Device/Setup/Reset Profiles-Device to reset profiles 2 and 3 to factory values

4) Use Edit Profiles to set your active profile to profile 2, select profile 2 in the Profile window, and then send to Device. You should now have your MTB profile in profile 2.

5) Do a short ride with your MTB. Hopefully watts will look normal

6) Download your MTB ride from 5). If you get a messages requesting overwriting profile, please take screen shots (so I can see what is going on) BUT DO NOT ACCEPT THEM.

7) Hopefully your bike speed will be correct. Post the ride file

8) Just to be sure there are no devils in your device (!) do another short MTB ride, download and post it. You should still have correct bike speed

8) Next, use Edit/Edit Profiles/Set Active Profile/Profile 3. Select your road bike profile in the window, then send it to your device. You will now have your road bike profile stored in your device profile 3 and your MTB profile in profile 2.

9) Do a short ride with your road bike, then download. If you see any messages about new profiles, etc., take screen shots and post them, but do NOT accept any changes to the profile!

10) You should see correct ride speed on your road bike. Post the ride file.

11) Do another road bike ride, download, post. Speed should be correct.

12) Click PP button 5 times to turn it off. Attach PP to MTB, click to turn it on, start riding. After a few seconds PP light should go solid green, meaning that it has shifted back to your MTB profile. Do a short ride, post. Speed should be correct.

13) Click PP button 5 times to turn it off. Attach PP to road bike, click to turn it on, start riding. After a few seconds PP light should go solid green, meaning that it has shifted back to your road profile. Do a short ride, post. Speed should be correct.

I am sorry for the very long length of this post. Your problem borders on baffling! I think the problem is with Isaac sending bad profile data back to PP; if so this is a bug. We will see what your tests show!

Corrected ride files for your MTB and road bike rides attached.

[doib]

Thank you very much John, I really appreciate your support. In the w.e. I will do the tests, I also think that the problem is related to Isaac. I'll update you as soon as possible!
Bye

Davide

[doib]

First mtb ride (point 5), after performing all the operations referred to in points 1-2-3-4. Aside from an initial peak, it looks good. Now I go out for the anti-demon race

[doib]

Second mtb ride (point 7). Looks like there is no devil in pp

[doib]

Uploaded profile no. 3 and done the first road bike ride. On the road all right, the watts seem correct. I arrive at home and connect pp to isaac. First problem: there are 4 rides still to be downloaded, instead of just one (I attach the image). I suspect that they are duplicates of those already made before, with the profiles inverted (I attach the files). Isaac doesn't ask me to edit any profiles. The file that comes closest to the strava values for the ride in question is the iBike_06_20_2020_1053_9_km.ibr; the profile is the right one. Now I go out for the second ride on a road bike, but I think the values will be correct and that the error will reappear when I ride MTB again.

[doib]

The other two rides

[doib]

Second ride with road bike. Correct power and speed, even distance. Isaac wanted to load an already loaded ride again. I did not overwrite the ride (is the first of today). I attach screenshot and ride file. I don't load the race with distance 0, I activated the garmin by mistake and immediately canceled it.

[doib]

Mountain bike ride point 12 ... doubled watts again. Isaac shows 4 files still to be downloaded, I attach photos (in fact all new). Still, some are duplicate ones this morning..I attach everything and go out on the road bike, but I already know what awaits me ..
ps no requests to modify profiles.

[doib]

Last file of the point 12 mtb ride.

[doib]

The zero km rides are indicated but not downloaded by isaac..I think they are born from the fact that I turned on pp after a few hundred meters

[doib]

Made road race point 13 .. contrary to what I thought, correct watts (file iBike_06_20_2020_1818_5_km.ibr). Again Isaac asks me to upload three rides and not one. But now I understand that the pp saves some old ones with a new number, perhaps repeating the saving of all those with the same profile? However, I am attaching new images and files.
I apologize for the amount of posts and files, but I tried to give as much information as possible. I look forward to your answers, thanks again.

[doib]

Image point 13.
Bye, thanks again

[Velocomp]

Mountain bike ride point 12 ... doubled watts again. Isaac shows 4 files still to be downloaded, I attach photos (in fact all new). Still, some are duplicate ones this morning..I attach everything and go out on the road bike, but I already know what awaits me ..
ps no requests to modify profiles.

When you switched back to the MTB, how did you do this? Did you use Isaac "Set Active Profile in Device" command, or did you let PP find the correct profile with the Automatic Profile Selection feature?

[Velocomp]

Made road race point 13 .. contrary to what I thought, correct watts (file iBike_06_20_2020_1818_5_km.ibr). Again Isaac asks me to upload three rides and not one. But now I understand that the pp saves some old ones with a new number, perhaps repeating the saving of all those with the same profile? However, I am attaching new images and files.
I apologize for the amount of posts and files, but I tried to give as much information as possible. I look forward to your answers, thanks again.

When you switched back to the road bike, how did you do this? Did you use Isaac "Set Active Profile in Device" command, or did you let PP find the correct profile with the Automatic Profile Selection feature?

[doib]

Both times with the automatic profile selection function. But today the situation has changed again. To avoid surprises, I reset profiles 2 and 3 again, loading only profile no. 2 (mtb). I made a 70 km ride on the mountain bike, the numbers of power, speed and distance are correct. However, when I downloaded the file with isaac, I got back a ride from yesterday, duplicated with a reduced wheel diameter. In fact, isaac immediately asked me to save a new profile for a road bike. I attach, in order, the files of the ride of today (70 km),of yesterday duplicated (3 km, but in reality there were about 6) and the sequence of images with the requests from isaac. It seems clear to me that Isaac tries to duplicate the profile but with the same sensor and different diameter, I don't know why. And most importantly, I don't know how to avoid it Could it be a problem related to the gsc-10

[doib]

?

[doib]

I forgot, isaac also wanted to save a new wireless id..here are the pictures (numbered in sequence).