Garmin Vector 3S - Low Power - Solved!
Discussion
I've splashed out on some Garmin Vector 3S pedals and, eventually, they work great. I'm leaving this here for anyone to find who experiences the same issues, because it wasn't obvious to me from any Google searches.
My Quarq power meter is still broken despite the best attempts of cyclepowermeters to run me through diagnostics/repair over the phone (absolutely brilliant guys though). So I'm on the hunt for a nice cheap Ultegra chainset whilst I ship it off for repair. In the meantime though I've been riding without power since about April and I'm really starting to miss it, especially the Wattsboard stress scoring.
I found Halfords had a great deal on some PowerTap P1S pedals - £389. With a discount here and a voucher trick there, I'd be looking at a little more than £300 for them, which seemed too good to turn down. So I placed an order for click and collect at Cycle Republic.
Long story short, they never arrived, and when I called to chase Cycle Republic, they did some digging and found there were none at all in the company and no plans to get any more. The eCommerce site should never have been listing them (which is a bit upsetting as it's run on my company's software - oops). They offered me the Garmins as an alternative, but these were listed as £499. I grumbled a bit, and they offered to price-match them with the PowerTaps! Can't argue with that. As I already had some money on one of my reloadable Halfords cards, which I consider to be sunk cost (shhh - man maths), my total outlay for these was under £300.
I was skeptical, of course, because of all the horror stories you hear about Garmin reliability. But I'm also accidentally becoming a bit of a Garmin fanboi - I have an Edge 810, an Index scale, and a Forerunner 235 watch. So I'm not un-used to the Garmin support chat.
I followed all the setup instructions (which, in comparison to previous Vectors, are basically "throw them on and swap the cleats" - no messing about with precise torque settings etc). And off I went for a ride. Returning home after just 30 miles, sweaty and exhausted, my weighted average power was 183W. Christ, I thought, I'm losing a lot of power because of the hot weather. Second ride, I went up and over a decent-ish climb, 381m of elevation. Weighted average was 174w, average for the climb was 217w. Now, I'm not a small chap, and 217w up a climb would either require a gear ratio of about 500%, or a tail-gale. So something was definitely amiss. In particular, I was able to compare a rolling 10 minute segment towards the top of the hill, with one I'd done a year before to the day. Heart rate, cadence, perceived exertion were almost identical, and my times were within 5 seconds of each other. The Quarq back in 2017 reported an average for the segment of 266W, the Garmins reported 174W.
The internet suggested that I turn off auto-zero and attempt a manual calibration. So, using the Edge (the computer, not the member of U2), I went to power meter --> calibrate. It then asked me to pedal at 80-90rpm whilst it set the install angles. After 5 minutes of this, I realised that if it hadn't figured out install angles by now it was never going to. Back to the internet I went, and found that install angles should be automatically calculated - I shouldn't be seeing this function at all. So now I went to power meter --> calibrate, then clicked "no" to the "install angles" message, and clicked calibrate.
This made no difference at all.
So I headed off to the Garmin Support chat, which was unexpectedly very helpful. I showed him the evidence of the two segments as described above. After doing some basic diagnostics he asked me if I'd set the crank length correctly. NOTHING in the setup instructions asked me to set a crank length; there is no option in the Garmin Connect firmware on the phone to set a crank length, and there's no option listed under the power meter on the Edge to set a crank length. Ah yes sir, that's because on the older Edges we've hidden it under the bike profile sir.
I went to menu --> bike profiles --> [appropriate bike] --> [randomly have to click appropriate bike again] --> scroll down and there it is, Crank Length, which had inexplicably defaulted to 110mm. Because I'm sure a lot of Islabike riders use Garmin devices.
I set this to the correct value, 172.5mm in my case, and went for another spin. Power figures are now much more plausible and make me look manly and strong once again.
The two rides for comparison:
https://www.strava.com/activities/1725356326#43232... With incorrect crank length
https://www.strava.com/activities/1729701128 With correct crank length.
One remaining question is how this will work with Zwift? If the pedals are relying on the head unit to provide a "multiplier" value for crank length, does Zwift provide a nominal value? Or does the pedal firmware remember the last known value until it's told to change it? In which case, why can't I have an option on the phone to view the value stored in the pedal?
Anyway, I hope this is vaguely useful to any users of older-generation Garmin Edge devices who buy some Vector 3s and wonder why they've suddenly become as powerful as Andorra.
My Quarq power meter is still broken despite the best attempts of cyclepowermeters to run me through diagnostics/repair over the phone (absolutely brilliant guys though). So I'm on the hunt for a nice cheap Ultegra chainset whilst I ship it off for repair. In the meantime though I've been riding without power since about April and I'm really starting to miss it, especially the Wattsboard stress scoring.
I found Halfords had a great deal on some PowerTap P1S pedals - £389. With a discount here and a voucher trick there, I'd be looking at a little more than £300 for them, which seemed too good to turn down. So I placed an order for click and collect at Cycle Republic.
Long story short, they never arrived, and when I called to chase Cycle Republic, they did some digging and found there were none at all in the company and no plans to get any more. The eCommerce site should never have been listing them (which is a bit upsetting as it's run on my company's software - oops). They offered me the Garmins as an alternative, but these were listed as £499. I grumbled a bit, and they offered to price-match them with the PowerTaps! Can't argue with that. As I already had some money on one of my reloadable Halfords cards, which I consider to be sunk cost (shhh - man maths), my total outlay for these was under £300.
I was skeptical, of course, because of all the horror stories you hear about Garmin reliability. But I'm also accidentally becoming a bit of a Garmin fanboi - I have an Edge 810, an Index scale, and a Forerunner 235 watch. So I'm not un-used to the Garmin support chat.
I followed all the setup instructions (which, in comparison to previous Vectors, are basically "throw them on and swap the cleats" - no messing about with precise torque settings etc). And off I went for a ride. Returning home after just 30 miles, sweaty and exhausted, my weighted average power was 183W. Christ, I thought, I'm losing a lot of power because of the hot weather. Second ride, I went up and over a decent-ish climb, 381m of elevation. Weighted average was 174w, average for the climb was 217w. Now, I'm not a small chap, and 217w up a climb would either require a gear ratio of about 500%, or a tail-gale. So something was definitely amiss. In particular, I was able to compare a rolling 10 minute segment towards the top of the hill, with one I'd done a year before to the day. Heart rate, cadence, perceived exertion were almost identical, and my times were within 5 seconds of each other. The Quarq back in 2017 reported an average for the segment of 266W, the Garmins reported 174W.
The internet suggested that I turn off auto-zero and attempt a manual calibration. So, using the Edge (the computer, not the member of U2), I went to power meter --> calibrate. It then asked me to pedal at 80-90rpm whilst it set the install angles. After 5 minutes of this, I realised that if it hadn't figured out install angles by now it was never going to. Back to the internet I went, and found that install angles should be automatically calculated - I shouldn't be seeing this function at all. So now I went to power meter --> calibrate, then clicked "no" to the "install angles" message, and clicked calibrate.
This made no difference at all.
So I headed off to the Garmin Support chat, which was unexpectedly very helpful. I showed him the evidence of the two segments as described above. After doing some basic diagnostics he asked me if I'd set the crank length correctly. NOTHING in the setup instructions asked me to set a crank length; there is no option in the Garmin Connect firmware on the phone to set a crank length, and there's no option listed under the power meter on the Edge to set a crank length. Ah yes sir, that's because on the older Edges we've hidden it under the bike profile sir.
I went to menu --> bike profiles --> [appropriate bike] --> [randomly have to click appropriate bike again] --> scroll down and there it is, Crank Length, which had inexplicably defaulted to 110mm. Because I'm sure a lot of Islabike riders use Garmin devices.
I set this to the correct value, 172.5mm in my case, and went for another spin. Power figures are now much more plausible and make me look manly and strong once again.
The two rides for comparison:
https://www.strava.com/activities/1725356326#43232... With incorrect crank length
https://www.strava.com/activities/1729701128 With correct crank length.
One remaining question is how this will work with Zwift? If the pedals are relying on the head unit to provide a "multiplier" value for crank length, does Zwift provide a nominal value? Or does the pedal firmware remember the last known value until it's told to change it? In which case, why can't I have an option on the phone to view the value stored in the pedal?
Anyway, I hope this is vaguely useful to any users of older-generation Garmin Edge devices who buy some Vector 3s and wonder why they've suddenly become as powerful as Andorra.
It will work perfectly with zwift. Somewhat like doing a zero offset at the start of the ride (hope you do that..), the head unit is triggering the process and the values are stored on the PM, in order for it to calculate watts. Crank length is somewhat fundamental to all pedal systems:
Strain gauges measure deflection, which is a proxy for force and with a known material / calibration for that gives you an accurate force.
Torque = force x distance (in this case, your crank arm length - for the quarq the distance is known, being the spider..)
Power is then a function of torque and RPM.
.. so if it doesn't know the crank length, or it is wrong, you're screwed. Surprised that isn't well documented, it certainly was when I had vector1's back in the day.
As above, once it's all set up, the PM is broadcasting watts to all who care to listen.
Strain gauges measure deflection, which is a proxy for force and with a known material / calibration for that gives you an accurate force.
Torque = force x distance (in this case, your crank arm length - for the quarq the distance is known, being the spider..)
Power is then a function of torque and RPM.
.. so if it doesn't know the crank length, or it is wrong, you're screwed. Surprised that isn't well documented, it certainly was when I had vector1's back in the day.
As above, once it's all set up, the PM is broadcasting watts to all who care to listen.
I guess what was concerning me was that in the head unit it’s stored as a property of the bike object, rather than giving any indication that it is pushing an updated value to the firmware on the pedals. Which would mean that the head unit receives raw data from the pedals, calculates against the stored value for crank length and then displays the output value.
Assuming this isn’t the case, as you say, it’ll work fine with a turbo. It’s just unintuitive design for updating the value in the firmware.
Assuming this isn’t the case, as you say, it’ll work fine with a turbo. It’s just unintuitive design for updating the value in the firmware.
Yes. If your head unit is capable of setting crank length, it acts as master - when it connects to the pedals it sends a message to say "hey vectors.. your crank length is xyz mkay?". So, if you had a TT with 165mm cranks, a roadbike with 172.5, moved the pedals around and set / used the right profiles, it would all work nicely. I'd hazard a guess that is exactly why it's on the bike profile..
If the head unit is incapable, the pedals go with what they know / was last set. Before many head units supported, you had to set from the laptop..
If the head unit is capable and the setting is wrong, you're screwed.
ANT is a broadcast protocol - not paired. It only makes sense to set on the sender. For one without it the vectors can't send watts which the power meter profile requires, and for two it would be a royal pain in the *** if you had to set the crank and zero offset on every receiver (zwift, watch, edge)..
If the head unit is incapable, the pedals go with what they know / was last set. Before many head units supported, you had to set from the laptop..
If the head unit is capable and the setting is wrong, you're screwed.
ANT is a broadcast protocol - not paired. It only makes sense to set on the sender. For one without it the vectors can't send watts which the power meter profile requires, and for two it would be a royal pain in the *** if you had to set the crank and zero offset on every receiver (zwift, watch, edge)..
Edited by upsidedownmark on Friday 27th July 16:46
Gassing Station | Pedal Powered | Top of Page | What's New | My Stuff