Often in science a mistake or accident can prove to show you more than your original experiment, as I just confirmed with my informal testing of my Quarq CinQo powermeter. So here’s the backstory:
After a season of really enjoying my CinQo, I had it shipped back to Quarq to have it recalibrated and get some firmware updates,.
I have to pause here to say that Quarq are a model of customer service and I am happy to have invested in their product: it was trust well-placed and I’d do it again. As a matter of fact, I can’t think of any cycling company, and few companies at large, that have been as good to me as a customer than Quarq has.
Also, after badgering the busy guys over at Saris / Cycleops (thanks Jesse and Geoff!) I received the beta Joule 2.0 firmware, which adds a manual zeroing feature for Quarq users.
So I decided with all of this new tech that I would grab a weight, some cord, and hang weights off of my cranks to see how the small and big rings and left and right crankarm affect power measurements.Since I could now hit a button and get a report on the torque measured by the Quarq when a certain weight was applied, I could see what it reported for torque compared to what the theoretical value should be given I knew the weight and the crankarm (leverarm) length.
I had a 20-lb weight right there, and after scrounging around a bit, decided I’d use an old power cord to make a loop on it to hang over my pedals.
I used the instructions on the Quarq FAQ page and built a spreadsheet and was merrily on my way to geek nirvana when I noted a strange thing: every time the cord that I had used to tie the weight touched the ground, I saw about 2-3 fewer 32nds of a Nm than I had when the cord wasn’t touching the ground.
My attention shifted: was the Quarq so sensitive that it could tell a 30 gram weight difference on top of the suspended 20.5 lbs when the end of the cord was resting on the ground or not?
Oh yeah – the CinQo is that sensitive. I tried this at least 15 times, then also tried it with some smaller weights in the 10-15g range added or taken away from the 20.5lb suspended weight. And every time, my quarq reported the difference.For good measure, I did weigh the cord end too!
At this point I had seen a few things:
1. The Quarq is *remarkably* sensitive
2. The Quarq showed a remarkable consistency, but even 2 degrees of slope either way in my crankarm away from horizontal would affect my results in a huge way, so it’s very hard to test your CinQo with weights and improve anything.
3. I don’t own a scale that’s good enough to weigh an object well enough to test my Quarq!!
This is a crazy-good piece of kit.
I just spoke with Jim from quarq (www.quarq.com) about how their Cinqo bike powermeter picks up and sends power signals to head units. This isn’t covered in their FAQ’s in depth, so I thought I’d post what I understood here:
- The Cinqo reads torque at 60HZ (60 samples a second)
- At each pedal revolution completion the torque values for the rotation are computed to come up with a power figure for the full rotation, using an averaging method rather than a “latest data” method.
- 4 times a second the data from the latest rotation is checked and if there’s a new value (i.e. you just finished a pedal stroke), transmitted via ANT+ to the head unit
You can use the above to go through some scenarios, but the take-home is that when you’re spinning at 90RPM, which is 90/60s = 1.5 rotations a second or a rotation every .67s, you’re generating a new value a little better than every second, and within a 1/4 second of that value being generated, it’s transmitted to the head unit. That means that at worst, the lag between the new value (which is generated every .67 seconds) and the transmission (every .25 seconds) would be limited mostly by the .25 seconds between sends.
The head unit, (mine a Garmin edge 500) on the other hand, is a different matter. Some say that the head unit takes the power measures and just shows, and records, the latest value, which is a bad idea – it should average multiple values in a second, if it receives them, at least for storage purposes, which is widely reported to be 1 measurement stored per second.
On a test ride last night, I saw changes in the head unit that correlate roughly with the above. Tonight I will try to spin up to 200RPM, but am guessing that I’m going to have a hard time reading the display at that RPM to determine if I’m getting faster updates.
There are rumors that head units like an upcoming one from cinqo will show multiple values per second, and that they will also allow for much better data resolution, which may be important for sprinters, or those interested in peak power in general.