2000+W power peaks with Stages Power meter

Do you have the latest software update from Stages? What version are you on?

My 520 randomly records 2000+ Watts efforts every 10-15 minutes or so lasting 2-3 seconds. It happens when I don't pedal at all, recorded cadence is 0.
I made sure it's zeroed before every ride. It never happened when I was using it with Cateye Stealth Evo +.
Happened to any of you guys?

I have the Stages and haven't had any problems and haven't seen any posts with that kind of problem. My guess is that the problem is with the power meter. With the surges happening at zero cadence I wonder if the Cateye software might have blocked them. Can you get your hands on another Garmin to do a comparison?

That would be my guess as well, that something is wrong with that power meter. The Edge and Garmin Connect do not filter out power spikes, but rather just report what they get from the power meter.

I would suggest doing a search across the forums you will see that other users have had this issue and it has turned out to be the power meter. Either a bad battery door or the unit needed replacing.

This is one area where it is unlikely to be an Edge issue. The Edge receives a message from the Stages power meter that contains torque and cadence data that is then used to compute power. The issue is typically due to an error in the torque or cadence value. Do these spikes occur upon restarting after coasting for example?

It is possible to capture the ANT+ messages sent from a device. If you were able to capture one of the events that one confirm if the problem is on the sending or receiving side, but I'm quite sure that the problem is on the sending side.

I'll investigate it. Just did a factory reset, maybe that will help. If not, I'll borrow another Garmin, unless I'll just send this one back, my patience is running out.
I find it hard to believe that there is something wrong with a power meter.
Especially after loosing half of my first ride with Garmin 520, power off mid ride the next day, lost power meter connection another, freeze while I tried to get it to follow gpx course, bugs just keep coming..

Especially if your Stages is still under warranty, you should try to sort it out. You might want to see what Stages has to say. They might send you another crank arm to try.

The 520 seems to be solid but that doesn't mean you don't have a doggy unit with losing data and the power shutdown. Maybe that is a hardware issue of some sort. But my guess is the power surges are coming from the meter.

Was the power meter connection completely lost or where there brief dropouts?

Your rides on Connect must be private, I can't access them.

I'm running a stages and the 520. No power spikes at all since i got the Garmin (several weeks now)

Thanks for the feedback. Although I had suggested that maybe the Cateye was filtering the spikes I was having a hard time buying that.

There have been some issues with power dropouts with the Stages. Maybe it's not that common but something to be aware of. Here is a thread on that. But that was dropouts only and not spikes.

I don't know if it would cause the unit to freeze when following a gpx course but Turn Guidance doesn't work. There is separated and different way to get turn prompts using Course Points that can be added to tcx file. Check out my post and others here

It does not sound like an issue with the power meter calculating bad data, considering the cadence reading is 0.

The power meter sends torque and cadence reading (in the form of event counts/time stamps) to the garmin and the garmin calculates those into power values. So if it sent 0 for cadence, power should be 0. There may be a mishandling of data, but not the same as a bad calculation (or bad reading from water ingress, etc). Power is torque x speed (Rpm in rad/s), so if one is 0, it is all 0, unless there's some sort of mishandling of data exchanged between the two.

I have seen this happen on some 510 files. Not sure what causes it exactly, but it doesn't ever happen on the older garmins or most other head units. You can email Stages about it though. The bad values are easy to fix in Golden Cheetah or another software, then you can save it and upload it to your site of choice.

280

I wasn't able to download goobeq"s files earlier because of the privacy setting. I just downloaded one of them and there are missing cadence and power values where the surges occur. There have been other posts about signal interruption from the Stages causing power drops but not surges. Maybe the 520 software is a little different and ends up calculating surges due to the timing of data. But it looks like the basic problem is the interruption of the Stages signal.

Stages are aware that interruptions can occur. see "Know issues" under Recommended Settings here. There have been a couple of posts observing that drops occur when the Edge is mounted out front but not when mounted on the stem. So making a slight change in where the Edge is mounted may be the difference. If your Edge is mounted out front maybe changing the angle of the mount might make the difference.

I wasn't able to download goobeq"s files earlier because of the privacy setting. I just downloaded one of them and there are missing cadence and power values where the surges occur. There have been other posts about signal interruption from the Stages causing power drops but not surges. Maybe the 520 software is a little different and ends up calculating surges due to the timing of data. But it looks like the basic problem is the interruption of the Stages signal.

Stages are aware that interruptions can occur. see "Know issues" under Recommended Settings here. There have been a couple of posts observing that drops occur when the Edge is mounted out front but not when mounted on the stem. So making a slight change in where the Edge is mounted may be the difference. If your Edge is mounted out front maybe changing the angle of the mount might make the difference.

Of course interruptions can occur, with any sensor, but as other members have seen they don't yield power spikes. The torque values all correspond to event counts, so the power and cadence can line up. If you power drops, it comes bakc and has an updated cadence count to go with it. The cadence values are null because the rider is coasting (and most people don't include 0s in cadence averaging, so no 0 is placed in the file, just a null value)

Having seen a friends files who had this happen, it always occurred after coasting. Even while coasting, there is still accumulated torque on the meter, but no event, so there should still be accumulated torque ant+ messages - but no new event count (so no cadence).


A similar, but different situation can happen in the event of signal interruption:

Although having ridden the 520 and 1000, i did notice occasionally I would get high values after a drop out occurred (this was not due to a power meter problem - i found out they dropped out if i put them side by side. seems they interfere with each other). Those numbers only showed up on the screen and not in the file. I assumed the processor was able to parse the data correctly to fix these. They were when I was actually riding though, so there was a cadence value to put in to make it a reasonable power number.

Also when using trainer road, a similar thing happens. At my office, the microwave causes drop outs.. whenever the ant+ dropped out on my sensors (it would kill all my ant+ sensors), I continued riding/pedalling. When the connection returned after the microwave stopped disrupting it, it'd pop back up on trainer road with a huge power spike - but then when I saved my ride, trainer road would process it and fix it. It actually shortened the gaps in the drops as far as I can tell - since I believe the accumulated torque includes several seconds, so it was able to reprocess it and fix some of the drop out. It took out all of the spikes that were shown on the screen.