Announcement

Collapse
No announcement yet.

Power Readings Issue

Collapse
X
  • Time
  • Show
Clear All
new posts

  • Power Readings Issue

    Have an issue with the Edge 810 where I will be riding along at higher wattages and for 2 seconds, the wattage reported will cut in half for 2 seconds. So for example, wattage will be 280, drop to 147 for 2 seconds, and then return to between 270 and 290. The cadence will remain absolutely steady, within 1 revolution for the entire period.

    The above phenomenon has occurred with 2 different power meters, one of which just returned from Quarq.

    Also have an Edge 500 and this issue is not happening with the 500 so I assume this is a Garmin issue with the 810. Cannot imagine the Ant+ receiver is bad in the Edge unit because heart rate and cadence are being reported accurately.

    Anyone else notice this problem?

  • #2
    Yes - I have. Thought it might just be power drop outs though and not the Edge 810.
    Now you've got me thinking ......................... probably more than I should be.
    -FR

    Comment


    • #3
      I have the same problem. In fact, if you look at the raw data in golden cheetah, you can see that every once in a while there are 2 seconds where the power is drastically different. Seems to happen on the first 2 second of a minute.

      Distance Power HR Speed
      9059 55.9999 307 75 165 15.0408 215.8 37.2179 -121.943
      9060 56.0042 361 77 166 15.5304 216.2 37.218 -121.943
      9061 56.0085 147 78 165 15.5304 216.4 37.218 -121.943
      9062 56.0128 147 75 165 15.6564 217 37.218 -121.943
      9063 56.017 363 73 166 15.0552 217.4 37.2181 -121.943
      9064 56.0211 324 72 166 14.778 217.6 37.2181 -121.943
      9065 56.0251 402 71 166 14.5008 218 37.2181 -121.943


      I went and rode with both the 810 and my old 500 computer and it confirmed that the problem is with the 810 and not the power meter. I have saved the ride files if someone from garmin wants examples.

      Comment


      • #4
        Awesome that you saved the data. I would definitely call or open a CASE with Garmin as while I know they watch the forum, I think it will get more direct attention through the proper venue. Doubt it will make the first firmware update (although there's always hope).
        -FR

        Comment


        • #5
          I too have seen lower power readings at times and I multiple rides. I am using a Power Tap G3 hub that is less than 4 months old. These problems do not occur when I am using my 800.
          AKA luv2climb.

          Comment


          • #6
            Everyone with this issue .......
            Submit a support ticket to Garmin.
            I'll add mine now.
            Thanks guys!
            -FR

            Comment


            • #7
              Yep! Same problem here. I submitted a ticket two weeks ago and they told me they have never hear of the problem and I should just reset the unit. I did so and spent the time to reconfigure all my settings. Did not fix the problem. So I just set my power display to the 10s average so it's not so distracting. Though it is frustrating still... This also happens on two different Quarqs. My 500 and 705 do not have this problem.

              I guess this is the ticket/case number: KMM27472591I15977L0KM
              ---

              MTB
              http://www.thecyclenut.me

              Comment


              • #8
                What you are seeing are 2 independent things.

                There is currently a known bug in the smoothing process for average power (3, 10 and 30 second). This impacts the UI only.

                Garmin have also changed the data averaging process on the 510/810 for the 1 second recorded value. This could result in different instantaneous power readings than that seen of earlier units, but the values returned by the 810 are a more accurate reflection of what is being reported by the power meter.

                If you use the ANT+ simulator you can capture the ANT+ messages from the power meter and you can determine see what is being done.
                The change to the recorded 1 second value is most noticeable for those using an event based power meter like an SRM/Quarq/Power2Max. It will not be that noticeable to those using a time based power meter like a PowerTap.
                - Alan

                Comment


                • #9
                  Originally posted by AWEATHERALL View Post
                  What you are seeing are 2 independent things.

                  There is currently a known bug in the smoothing process for average power (3, 10 and 30 second). This impacts the UI only.

                  Garmin have also changed the data averaging process on the 510/810 for the 1 second recorded value. This could result in different instantaneous power readings than that seen of earlier units, but the values returned by the 810 are a more accurate reflection of what is being reported by the power meter.
                  Okay, this concurs with what I think I'm observing. I set a screen that shows 3s and instantaneous power reading and even though the 3s shows a massive fluctuation the instantaneous does not.

                  Originally posted by AWEATHERALL View Post
                  If you use the ANT+ simulator you can capture the ANT+ messages from the power meter and you can determine see what is being done.
                  The change to the recorded 1 second value is most noticeable for those using an event based power meter like an SRM/Quarq/Power2Max. It will not be that noticeable to those using a time based power meter like a PowerTap.
                  I tried to look at this using Qalvin, the Quarq iOS tool. It was quite difficult to come up with anything. Looking at the detailed level of my workouts on Training Peaks though, it does appear that occasionally there is an unexplained power drop (this while riding on a trainer). However it is strange because it does not coincide with a change in cadence. Unfortunately I only notice these issues when doing efforts around or above threshold, which compounds the frustration when I get the beep that I'm under the zone I'm supposed to be in!

                  Thanks.
                  ---

                  MTB
                  http://www.thecyclenut.me

                  Comment


                  • #10
                    All have responded to the very issue I brought up.

                    So there appear to be 2 bugs, a known bug in the UI for 3 second average power and also a bug in the way the unit is recording power with the Quarq.

                    I see the sudden fluctuation on the UI when riding and after uploading into Trainingpeaks and looking at the raw data, see the 2 second drop in power.

                    Comment


                    • #11
                      Originally posted by HEROTHEDOG View Post
                      Okay, this concurs with what I think I'm observing. I set a screen that shows 3s and instantaneous power reading and even though the 3s shows a massive fluctuation the instantaneous does not.



                      I tried to look at this using Qalvin, the Quarq iOS tool. It was quite difficult to come up with anything. Looking at the detailed level of my workouts on Training Peaks though, it does appear that occasionally there is an unexplained power drop (this while riding on a trainer). However it is strange because it does not coincide with a change in cadence. Unfortunately I only notice these issues when doing efforts around or above threshold, which compounds the frustration when I get the beep that I'm under the zone I'm supposed to be in!

                      Thanks.
                      Yes, i have noticed the same issue with random half power drops with my edge 810 and quarq cinqo. I will be riding steady on the trainer at 350W then it will show 180W. I checked the raw data on trainingpeaks and confirmed. This happens about every 2 min or so on avg. I will also submit a support ticket.
                      As for the power zone training, you can modify your training program in the garmin to do a rolling 3 sec or 10 sec avg power in that zone. I started doing that recently for structured intervals on my trainer and that made it so much easier to stay in that power zone.

                      Comment


                      • #12
                        I am having the same issue with my Quarq PM (SRAM Red Exogram model) too. Lots of random half values as described in previous posts.

                        My experience:

                        On Sunday, I used my Edge 810 for the 1st time proper (while on the turbo trainer), rather than messing about with settings etc, it was a steady tempo effort with a nice constant cadence. As a sanity check, I ran my old Edge 500 in parallel so I could spot if the data was different and check all the settings were correct. I noticed that the Edge 810 was giving me lower average power or normalised power for each lap than the Edge 500 even if the %FTP was reading the same on both units, so I checked all settings, everything is the same.

                        Data recording, 1 sec
                        Power averaging, include zeros
                        Cadence averaging, do not include zeros

                        Both devices were used to zero the Quarq, both recieved the same zero offset value back at each point, before, during and after the session. When reviewing the data I noticed these odd readings but I let it go this time thinking I would know more the next time I used it as one file is not enough to come to any conclusion, I couldn't be bothered to compare with the Edge 500 file of the same workout at that time.

                        Well now my Edge 500 is away getting the mount repaired, so today I rode with just my Edge 810. The same thing, riding at say 100% FTP (minor fluctuations ofc) but the average and normalised power for the lap is below my FTP figure by some way, it made no sense to the point I couldn't use the raw power figures in any form. I finished my session and uploaded the data, only to find these half values again. My 1st response was to check if there was a corresponding drop in the cadence data, there was not. So I checked everything possible with the Quarq, magnet is in the right place, unit passes its own diagnosis mode, battery good etc.

                        Tried a few short mini-rides trying to eliminate the issue, no luck! So I decided to review the file I recorded in parallel with the Edge 500 (I backed it up before sending off to Garmin) on Sunday as I was becoming very annoyed and lo-and-behold it is free off these half values. So 2 Garmin devices, both set up exactly the same, recording the same training session, simultaneously, and one is free of these glitches, the other plagued by them.

                        There is something not right with the 810, for sure. I have been using my Edge 500 with the Quarq flawlessly.
                        Last edited by Rob3rt_H; 02-05-2013, 05:56 PM.

                        Comment


                        • #13
                          Thank you for sharing. I hope you will forward your experiences along to Garmin support for all our benefit. I have sent over mine as well as several other "glitches".
                          Hoping they get these irone out ASAP.
                          As the days go by, I am less and less optimistic.
                          -FR

                          Comment


                          • #14
                            I will drop them a phone call tomorrow.

                            Comment


                            • #15
                              Originally posted by ROB3RT_H View Post
                              I will drop them a phone call tomorrow.
                              i might also call garmin too. this is a major issue honestly. I was riding today and noticed this issue and that it occurred about every 3-4 min. It was a 2 sec drop of about half the power. I think this is not a major effect in the normalized power scheme since that is a rolling 30 sec smoothing, but seems to affect the avg values much more. I really hope they can fix this soon.

                              Comment

                              Working...
                              X