Elevation going bananas

[bucko]

Hi! I've had my PowerPod a few months, though I only started using it properly recently (I broke my arm). I bought it mostly for wind recording and to try to get power numbers from London cycle hire bikes; I already have a DFPM (a PowerTap Pro upgraded to a PowerTap G3). Comparing the power numbers shows the PowerPod often puts out pretty ludicrous numbers (100W while coasting at 10kph, or 0W while pedalling) but is often within 15%. My belief is that it's been doing this since I started using it, though I don't have records of the earlier rides (I transferred them with erase turned on, then corrupted the hard disk!).

I used Isaac's merge feature to merge in the TCX from Garmin Connect (it didn't want to align the FIT for some reason), and the problem was immediately obvious: the elevation is almost completely unrelated to anything I'm actually doing (I thought it looked pretty odd before I merged!). The measured speed seems fine, and even the slope seems to mirror something closer to what was happening than the elevation plot. The Garmin elevation in red is pretty much spot on, and you can see near the middle (before the final climbs) me making 2.5 laps of a park in that plot.

I'm guessing the device gets elevation from a barometer, so I don't think any offsetting is going to help here. Is it possible it's somehow clogged? Anything you can suggest trying other than sending it in for replacement? (If replacement, how do I do this?)

[Velocomp]

Please post the .ibr ride file. We will be able to tell if something is not right.

You should be within a few % of your PT, so something appears to be amiss...

[bucko]

I noticed some stuff about 'tilt cal', so I've reset all of the profiles on the device. I will do a ride tomorrow and recalibrate. I've moved it around the bike a fair amount, so it's possible it's calibrated to an incorrect tilt and remembered as such. I also noticed it has the wrong weight (my bike is a super-heavy 20kg due to panniers and heavy tyres) and wheel circumference (Garmin says 2144) and so-on for me; should I enter these myself, or are they auto-detected?

The attached ride is the one from earlier, unmodified.

[Velocomp]

You have a fair number of sensor dropouts on your ride file. This is likely the source of your elevation problems...

Please do the following:

1) press-hold your button for about 10 seconds, until the light stops flashing momentarily, and then starts to flash again. This resets your PowerPod

2) I'm guessing that you have a Garmin GSC-10 sensor. If so you might want to replace it; they do not broadcast a very strong signal. The new Garmin magnet less sensors are great, as are the Wahoo sensors.

3) Make sure your speed sensor magnet passes within 1-2 coin widths of the sensor probe

4) the date and time of your next ride will be "off" but when you connect to Isaac the date/time will be restored.

If you get a solid speed sensor signal I suspect your elevation will sort itself out. Please post a ride file.

Once we get a good speed reading we'll sort out your weight and circumference.

[bucko]

You have a fair number of sensor dropouts on your ride file. This is likely the source of your elevation problems...

How do I spot sensor dropouts? The messy speed is caused mostly by me cycling in London and having to stop every few minutes at traffic lights. The clean section in The Regents Park is clean because it only has 5 sets of lights in the 4.45km lap and I was quite lucky today. My Garmin doesn't have any trouble with speed when on a stationary trainer. The speed graph aligns perfectly with the GPS speed graph after alignment; I assumed the misalignment was simply caused by the PowerPod skipping over stationary periods.

Please do the following:

1) press-hold your button for about 10 seconds, until the light stops flashing momentarily, and then starts to flash again. This resets your PowerPod.

I will do this at the start of the next ride. Is that meant to put it back to calibrate mode? I've done this before, but I didn't get the 1-100 count a second time. Must I do it stationary with no sensors synced?

2) I'm guessing that you have a Garmin GSC-10 sensor. If so you might want to replace it; they do not broadcast a very strong signal. The new Garmin magnet less sensors are great, as are the Wahoo sensors.

You are correct, though I did deliberately try to sync to the new magnetless sensors which I also have (I keep losing pedal magnets, so I used the magnetless ones to get cadence for the PowePod). I did the initial sensor sync by spinning the front wheel and pedal only. I suppose it's possible that when I reset it, it picked up the rear sensor. I currently have no cadence magnet though, so if there's cadence at all it is from the magnetless sensor on the left crank.

I guess since I'm not actually using it for anything I can just take the old sensor off entirely, which I've been too lazy to do. I'll do that.

The Garmin has the GSC-10 synced but I think the PowerTap signal overrides it anyway.

3) Make sure your speed sensor magnet passes within 1-2 coin widths of the sensor probe

(It actually hits the probe sometimes, though I think that pushes it away again! At least will save me from that issue!)

4) the date and time of your next ride will be "off" but when you connect to Isaac the date/time will be restored.

If you get a solid speed sensor signal I suspect your elevation will sort itself out. Please post a ride file.

Once we get a good speed reading we'll sort out your weight and circumference.

Righto. I will do the above before I go out for a ride today and see how I fair.

[bucko]

Right, I removed the combined sensor so I'm surely synced to the magnetless ones now (unless the PowerPod picked up my PowerTap's power profile). I still got a crazy elevation plot. Attached is a .ZIP of the initial calibration, the rest of the out-and-back, the Isaac combined file of both of those bits, and the .FIT files from my Garmin 820 (recording the Garmin sensors and the PowerPod - the -PowerPod file) and my Garmin 510 (recording only from the PowerTap - the -PowerTap file).

The PowerPod is still convinced the ride was almost entirely uphill despite it being an out-and-back ride up and down a hill twice. I had to slow a few times in the cal section, and the first hill descent has a lot of braking as it hits a junction with little visibility. The second long part of the return descent was a pure coast, though.

[bucko]

Messing around the settings, the software seems to want me to use a "riding tilt" of 3.2% but when I press send, it tells me this is out of range. Looking closer at the mounting, I see the "PowerPod" text was pointed a little upwards -- I suspect I aligned it while the bike was on a slight lean with the handlebars turned (it looks flat when I do that!). I've wedged a bit of plastic into the top of the supplied mount to step the pod from being able to rock back and forth then re-aligned it with the bike vertical, and I'll try a second reset and recalibrate.

[Velocomp]

Looks like your elevation sensor has problems. Please email technicalsupport@velocomp.com for next steps.

Good news is that your sensor dropouts are gone...

[bucko]

OK, I've mailed them. Thanks for your prompt response.

[bucko]

Just to be close off this discussion, I sent my PowerPod off and it's been replaced with a PowerPod with working barometer (I assume a new one unless it also had the scratches on the bracket repaired!). Now my elevation traces look like those from my actual GPS device!

Power seems a bit low; I just messed with the weight settings and such to setup a profile, and as the profile was lower I guess that explains some discrepancy. I'll have a play and maybe do a calibration ride on the weekend.