I think I read it I DC rainmaker site, magene will only focus on Chinese market now.
unlikely as their UK distributor has just had a major push to all retailers
Magene's New Trainer & Power Meter, But Pulls Back From Europe & US Markets My English isn’t good maybe it’s only for newer models
That definitely reads as ‘not outside of China’ but seems to be mainly talking about the trainers. Checking the distributer’s site, there are no trainers listed in the UK but the power meters, head units and light/radar are. That’s good news for me!
It looks like shipping is happening soon according to the latest update on Kickstarter.
@Aram does this mean we will have review units sent out to the usual suspects?
I just signed up here, and read the thread. (Google mentioned it as one of the few spots where the Cycleclick power meter is discussed.) It seems nobody here did participate in the kickstarter, or maybe they didn’t receive their devices yet?
Anyways, I have one, but I have only been able to test it against a built-in power meter on my Yamaha Wabash RT e-bike. I also did record the meter in parallel with a SRAM ‘Quarg’ in my one of buddies’ bikes, but it seems very hard (or impossible) to define stretches of data where neither of us was drafting, and where we were maintaining the same steady speed. (We’re also on different bikes, different tires, different weight, different air resistance, so probably moot.) My main use for the meter will be on my non-motorized Canyon Roadlite 6.
A few questions:
Is the Yamaha power meter worth comparing to? Has anybody compared it (the Yamaha built-in meter) or calibrated it against other power meters here? The Yamaha seems to read about 10% more than the Cycleclick. ** I have not found a good way in intervals.icu to plot the two power readings against each other (not an intervals.icu expert; obviously I can overlay the traces), so I don’t know whether it’s a constant (linear) scale factor, or maybe more complicated. Extracting the average power readings for both meters at different power levels manually seems tedious. I do like numbers as much as the next guy, but I don’t like typing a lot of them by hand. There must be a better way. Hence the next question:
How does one download or otherwise get hold of @dcrainmaker’s comparison tool? I clicked on the link in one of the posts above, but all it seems to show are screenshots from the tool, with no link to the tool itself, unless I’m doing something wrong.
** Added: After exploring the intervals.icu compare feature for a while (and not making sense of its interval selection), I finally now used fitfiletools.com to extract 18 minutes worth of raw Yamaha and Cycleclick power data, recorded while riding the Yamaha without motor support, and put them into a spreadsheet (I had been hoping to avoid that 
). I get 1119 data points, one for each second in the ride. Looking at the data in detail, it seems the Yamaha might record things a fraction of a second later than the Cycleclick. Or maybe this is an offset introduced by the recording app. In intervals.icu, the overlay of data traces seems to have a slightly sharper appearance (minimal apparent vertical line width) for a -1 second offset of the Yamaha data, as compared to 0 seconds offset.
Interestingly, while it’s true that the Yamaha power vs time trace in intervals.icu appears 10% higher than the Cycleclick trace, if I actually average the ratio of Cycleclick power over Yamaha built-in meter power for those data points (1-second intervals) where both measurements are non-zero, I get an average power ratio of 1.013 (Cycleclick over Yamaha).
For the integrated work (in Watt seconds or Joule) done during those 1119 seconds, the Yamaha gives a value that is a factor 1.091 higher than the Cycleclick, with an integral of 140.4 kJ from the Yamaha sensor, and 128.7 kJ from the Cycleclick sensor. So that agrees with the visual appearance of the power curves, and the numbers I read from the intervals.icu plots.
For both of these things to be true at the same time (average ratio of individual power levels within 1.3%, and work over time = average power 9% different), there must be some transformation going on in one (or both) of the units. It could be built-in to the electro-mechanics of the meters, or some processor could apply some weighted averaging or smoothing; I have no idea. I assume there’s a bunch of people here who have looked at and compared power meter data for a lot longer than I have (which is 2 days, and counting…). I’m very interested in suggestions (as long as they don’t involve typing lots of numbers ).
And here’s the plot of both series against each other, x= Yamaha value, Y=Cycleclick value:
I included two trendlines, one of them forced through (0,0).
And here’s a screen grab from interval.icu; time units are minutes (e.g. 0:16 to 0:17 is one minute), the whole segment is 2’40" long; Yamaha power in red. Doesn’t it look like some smoothing or peak hold is going on for the red curve?
Sorry, I don’t have a lot of input about your analysis, but here is the information for how to reach the DCR Analyzer tool:
I would take a look at how gplama tests. Without extensive testing like that we don’t know if the power meter is accurate or not.
I know @Aram said they were working on getting some third party reviews but those never materialized. Or, seeing as it was delayed in release maybe they are still in progress?
There is also this:
Thanks. I looked at that same page again, and this time I clicked on the live example. That does lead to analyze.dcrainmaker.com, so I registered for an account there. Unfortunately, after registering and confirming my email, I get a ‘server error: error getting user’. Oh well. I now also @-tagged rainmaker in my post above, so maybe he will chime in.
Thanks. I’ll google gplama. Also thanks for the heads-up. I think all the kickstarter participants got an email from @Aram. I checked my two parts (the clip-in ball joint mounts), and they are both ok. I’ve also done test rides already with both of them, and as @Aram can confirm, I’m really good at breaking stuff
The first 100 devices are out, as far as I know, but it seems nobody here was interested enough to participate in the kickstarter, or at least not early enough
This would be a lot easier if you had a “trusted” power meter to compare to. I would start by doing the equivalent of a ramp test, i.e. holding steady-state power across a range of power intervals. For example, start at 20W for 30 seconds, then 40W for 30 seconds, and so on. You could start at a higher initial power if 20W is too low. I would try to find consistent conditions for this - a trainer would be best, otherwise a consistent road surface and direction where you can hold constant power through constant cadence. You should use one of the powermeters as your reference for your effort, and let the other one just measure. You can then plot this and see how well the staircases align and whether there are any trends across the power range. Similarly you can try sprints from a low power to a high power and back, and compare those.
Until you have confidence in the basics, it is hard to make conclusions from real-world data, especially with two powermeters of unknown accuracy. Your X/Y plot is not very encouraging …
Thanks, @ukbloke. What I have done in the past is relatively slowly (below 10 mph) biking up a hill, and then comparing the work done against gravity with the power meter integral. But there’s still a few percent road resistance, power train losses, etc.. It’s hard to measure down to the percent range.
And no, I’m not going to buy a second (or third, really) more expensive power meter to calibrate my first two…
If you really want to do proper testing, you need to compare it to a known, trusted power meter, e. g. a Quarq spider-based power meter or Assioma pedal-based power meters.
You are right we are skeptical, but — at least in my mind — for good reason. There have been plenty of power meter start-ups/Kickstarters in the past and I am not aware of one that produced a usable power meter that was widely available, which has held up to independent testing.
You weren’t all in on the IQ2 pedals? 
Shimano can’t even get their PM right so I don’t trust any PM until I see some of the big reviewers out there saying they good
It was Limits for me, Seemed stable enough indoors but was really wild outdoors.
I think I was giving them the benefit of the doubt far longer than others, but I wasn’t about to spend money on a random number generator.
Yup, same here. I was really waffling on the Rotor power meter I got for cheap two weeks ago, because it received meh reviews. But in the end I thought I wanted shorter (and needed new) cranks, and even as dumb cranks they’d be great.
Thanks, @OreoCookie . Being skeptical is good. I didn’t expect the Yamaha built-in power meter to be super precise. Its main purpose is to provide input for the motor control. If I remember correctly, it doesn’t even let you read out the power with the default handlebar display unit; I had to plug in a different display to make the data available over Bluetooth.
So, all one can conclude from my data is that both meters seem to measure similar within 10%, and the actual precision of the Cycleclick still needs to be verified independently.
I hope one of you guys with calibrated power meters gets hold of a unit soon. Did anybody else subscribe to the Kickstarter? Apparently there also was a Beta tester program before the first Kickstarter units shipped. None of you guys here were in that either?
In any case, it’s a real blast to have the power read out available on my non-motorized Canyon. I don’t think I need single-digit percent precision for my rides. It would be nice though if the calibration were reasonably constant over time.
At least personally I have doubts about the accuracy of the unit, until we get third party reviews I will remain sceptical.
We have seen many people come out with cheap power meter promises over the years and they have almost all not delivered, heck even Shimano still ships their power meter and it’s “accurate” readings.
I just hope @Aram has sent out review units but the account here seems abandoned so I guess we just wait and see.
Isn’t this supposed to be really easy to remove and put back on? Can’t you just put it on your buddy’s bike for a few minutes and record direct vs his Quarq?
I don’t think Aram abandoned his account. From my reading of this thread, he just seemed a bit tired of all the negative expectations (which are based on experience with other engineers and companies, not with him). Plus, right now I assume he and Shushanik (his wife and business partner) must be awfully busy dealing with Kickstarter communications, having to remold the ball joint part, and having to reschedule the remaining Kickstarter shipments. According to their most recent announcement, they had the next 200 shipments already sitting there, and those are now delayed and I assume have to be repackaged. As nice as it must be to see your invention become reality, the nitty-gritty of selling directly to customers must be exhausting.
Plus, there’s just not that much to discuss, if none of the people here have a unit (yet).
I’m surprised that none of the beta testers posted reviews, except for that one youtube video, which is not a review. (I don’t know whether it’s generational bias - i.e. me being old -, but I just don’t find video reviews all that enjoyable. It’s just horribly low information density, and when the density is higher for once, you end up scrolling back and forth to get all the details. What has suddenly become wrong with writing and pictures??)