Matching power meters - how to interpret this test

Hey all,

I’m hoping someone here with more knowledge than myself can help. I have a Garmin Rally 100 PM (left only). Just purchased a 4iiii left only PM for the TT bike. Today I did a test to see how they compare. Per some reading here, I did a ramp test up to about 120% ftp, followed by about 30 minutes of endurance spinning. During that 30 minutes I did 2 hard sprints just to see how that showed up. I recorded the Garmin using the TR app, and the 4iiii on my Edge 530. Both PM’s were calibrated prior to testing. After completing the test I uploaded the files to DC Rainmakers analysis tool (very cool!).

This is where it got interesting. I noticed the 4iiii PM was reading low (5-8 watts) for most of the workout. Until the first sprint. Then the two lined up really well. For the rest of the workout, including the second sprint they were within 1-2 watts of each other.

So 4iiii low through the ramp test and first portion of endurance spinning. Then hard sprint, endurance spinning, another sprint, more endurance and they both matched up really well.

My intention was to adjust the 4iiii to match the Garmin since I’ve been using the garmin for some time and it’s tied to AI FTP. I will only use the 4iiii when riding inside or racing for Triathlon (not nearly as often as the garmin pedals).

Thoughts on what’s going on here? Do I make an adjustment and test again tomorrow, or just test again tomorrow without making an adjustment?

Here’s a link to the analyser with the data.

I see this happening often after hard sprints, across a few different types of power meters. A lot of the time it’s only a few watts drift which will go unnoticed without having another meter on the bike to use as a reference.

With only two meters it’s hard to tell which one has shifted. If you were to stop and zero just one of the meters after you see a change, you’ll get a better idea of which has changed.

When zero’ing (calibrating according to Garmin) the Rally it’ll show a real-time torque value on the Edge unit. If after a hard sprint when you zero it and it shows something like 1.12 or -2.34 (for example) before switching to 0.00 (or close to)… then you’ll know something is up. 4iiii will only return ‘10’ for a successful zero (1010 on dual).

Perform the same test again when you can. If both meters line up from the start and are ok after the sprints, it’s likely just a post-install settling in that has occurred. If the same issue is observed, it’ll be time for a support ticket to either Garmin or 4iiiii.


Have you noticed this type of behavior with recent Quarqs?
Wondering if the magic zero feature (or similar auto zeroing) software features can help avoid this issue.

Some sort of shift in zero after prolonged and/or particularly hard efforts seems like a difficult issue to solve. Perhaps it is the same or similar to the issue of “load cell creep” - What Is Load Cell Creep? - Arlyn Scales. From my limited understanding, it seems logical that the zero point of a load cell/strain gauge would be liable to shift a small amount under certain loading conditions (like if you left a heavy object on top of a kitchen scale for example overnight).

We ask a lot of our power meters :slight_smile:

That makes sense. I’ll do the test again tomorrow and report back. Do you think I need to do the whole test, or can I just get on the bike and spin for a few minutes to see? Thanks for your help, much appreciated!

Make sure your 4iiii pinch bolts are torqued to spec.

1 Like

Yup, with a torque wrench! :+1:

Update. Did another test today. Zero’d both PM’s at the beginning. Things looked much better! Looks like possibly the 4iiii was just off for the first portion of the last test. It was the first time I used it. I remember having a similar issue the first time I used the Rally pedals (had a set, they took a crap. Garmin sent me a new set, got them the day before a race so I put on and sent it. They showed huge PR #'s, like not even realistic. Ran them again a few days later and everything was good.)

I am seeing some drop outs from the 4iiii file, which is weird. Other than that I think I can move on and just leave them both alone.