Geoid PM500 power meter

[GLTRV]

I bought and installed a second PM500 on another bike one or two months ago.
As I've read a few reports of unsatisfactory results, I wanted to test my second unit to verify if it was as accurate as my first one.

For the first test, I had not calibrated the unit for several weeks as I was curious to see how it would falsify the data or not in light of my recent experience testing the Nexga S1.
The lack of calibration did introduce an error of ~5%.
Calibrating the unit is really simple, just spin the crank backwards for 5 turns, no need to go through menus. While I probably won't calibrate mine every time I use it, I'll certainly do it more often, at least once a week.

I'm not inserting the images of this test directly here as it's not the relevant test, but if you want to see how the results look like, here is the complete test result https://i.imgur.com/SX5deIi.jpeg and here are the details after the last sprint https://i.imgur.com/p6R4soH.jpeg.

After calibration, my second PM500 is spot on when compared to my Tacx Neo 2T.

Complete test (after calibration)


Details of the second sprint (only 850W but the Magene/Geoid power meters stay accurate after harder efforts, so I'm not concerned)



Very happy with the results with both of my PM500.
Again, I wish for a lighter crank option with 24mm spindle, or better yet, a non proprietary interface.

Note that I may not have calibrated my first PM500 before testing it earlier this year. I'll test it again with a fresh calibration at some point.

[zealot128]

I also made a video on my channel (German, but subs should be there); I only rode outside with Favero Assioma Duo as benchmark, so the cadence powers fluctuate much much more than indoors, as there ..you know, are people, cars, road furniture, curves etc.

Gist: Crankset is solid, especially for the price of 140 EUR currently on Ali etc. ; Mean/averages (various mean methods, NP etc.) are very close to Favero Assioma, especially the longer the ride, the less eventful, the more the lines match. Numbers also seem to be very consistent, even when not calibrating regularly.
Compared to Assioma Duo there are stark differences when switching cadences, short bursts etc. those always take a couple of frames/seconds to catch up.
Also, there are more occasional "zero" (null) values, compared with assioma, i tested Wahoo, IGPSports and Coros and switched pms around those to eliminate the source of the receiving end, there are definitively more null values compared to Coros, where even the cadence is null. this happens more often when reducing the cadence (going from 100-80 or so) - In the longer runs those make not much of a difference, but if there were stark differences between the PM's it was usually because of those drops. As far as I understand it ANT+ protocol allows for missing packages, and all packages also contain the accumulator instead of some moment value, so not sure how the "nil" values can happen, can only think of a not sent package; or my head units don't use the accumulator packages, that's some investigation for my next projects.
There seem to be no way to have any kind of Firmware update as of now (pls if one find any, share it), the Magene app can connect with it, but there is no option there. As the powermeter, at least to the FCC photos/docs, is identical to Magene's, it leaves the question if newer firmware updates would improve the accuracy.

As mentioned above, I did have a complete failure of my first unit, which still connects, but it appears the cadence sensor is shot. The only explanation i have, is that in 40ºC summer in my bike garage shed the temperatures spiked above the 50ºC from the manual. Besides the one fellow on the Magene reddit thread, that seem to be not a wide-spread issue, yet.