Insightful discussion.

Anyone have experience with 12sp connex chains on Ultegra components? they come pre waxed with their chain link which I will use anyways, so it's the cheapest option for waxing.



Anyways, I'm still curious about ordering the groupset from aliexpress.

I contacted anrancee store and they say

I guess what this means is they declare lower value and technically the shipping you pay (40-60€) covers customs for that.

I'm led to assume that YFH shipping (which I think is just standard carrier for the respective country..?) still has a higher chance of being sent to your local customs office where you will have to pick it up and show proof of purchase...

I guess by going with UPS Express there's less chance of complications, since UPS will handle the fees and then bill you afterwards..?

I bought a 105 groupset and an Ultegra groupset, both Di2, from the Anrancee store, with YFH Shipping. I didn’t have to pay anything when it arrived. The shipping method is similar to DPD. I bought them from Spain, and I’m really happy with this store and their service.

My concern with that is the double cost as I'll need another once the bike is built and funds are a little tight. Obviously buying the wrong frame is also more expensive. The 56 is quite a lot smaller which scares me!

You can get a pre-purchase bike fit. They basically replicate the bike's geometry on a special fitting bike, and from there they determine the exact frame size and all the other measurements, such as saddle height, stem length, handlebar width, and more.

That way, you can buy the bike with the correct dimensions and set it up with the right number of spacers from the start. It saves you both time and money because you won't have to keep experimenting and swapping components. I think it's your best option if you're not sure about the right measurements.

That's a LOT of testing. Nice work on that. I will point out our testing scopes are very different, so it's impossible to conclude the issues I had don't apply to your unit too.

I've had a look through the data but can't find any peak power tests >1000W. Did you perform any sprint testing? This was one critical issue I had with their meter. Sprints caused issues with cadence sensing and residual torque (meter not returning to zero point after peak power efforts). This isn't unique to XCadey. I've replicated this with Elilee EK01, and again on that unit with 1202 firmware (after they'd published a press release claiming it was solved).

Was there any outdoor testing performed with the XCadey Gen 2? Indoor there's a lot less external variables - less temperature drift, less bike movement/vibrations, less stop-start pedalling which can also indicate lag/delay with meters. I can't see any coasting in the datasets you've posted.

Here’s what I think is happening with the XCadey and a few other spider power meters I’ve tested lately: The hardware simply isn’t capable of reliably measuring cadence in all scenarios, especially during fast acceleration peak-power road sprints. Garmin ran into the same issue with the Vector 3/Rally x00 series, and only solved it once they added a gyroscope in the Rally x10. This wasn't covered/known as most review tests simply didn't perform any high peak power testing. On top of that, these budget spider designs appear to use materials and/or bonding glues that return to a non-zero torque state after high torque efforts. This was confirmed within the XCadey app. That behaviour isn’t unusual for spider-based meters, but it does require good auto-zero logic to correct itself when coasting or back-pedalling.

Finally, I don't believe I was sent a faulty unit by chance. I suspect they're all going to fail due to the hardware design. Someone has posted ~1000W sprint data in another thread here that appears to replicate the cadence/power reporting issue. XCadey has made no attempt to send another "non-faulty" unit to retest, which I'd expect given their response. I can literally retest within an hour of receiving a new unit, I can have a video published a few hours later. I've heard nothing from them for almost two months (25th Nov 2025).

I noticed that Xcadey has released a new firmware version. According to the release notes, they have fixed the cadence measurement issue as well as the power spikes recovery errors that occurred after short, high-power efforts.

These are the release notes:
https://xcadey.com/2026/06/16/xcadey-gen2-spider-power-meter-v109-stable-firmware-released-battery-life-extended-to-350-hours/

In case you'd like to take a look.

Any ideas on the best way to calibrate this? I can see a calibrate option on my old Garmin edge 800 but pressing it does nothing.

I'm having the same issue with the Coros Dura. I press "Calibrate," but it never completes the calibration and eventually returns an error. I also can't calibrate it through the Magene app, and it's reporting lower power values than expected. It's off by about 40-60 watts.

Since I installed it about a week ago, I haven't been able to calibrate it successfully. Does anyone know of another app that can be used for calibration?

I've spun the cranks backward five times, and the readings remain the same, so it might be a defective unit. I've compared it against two other power meters and a smart trainer, and they all show very similar values, except for this one.

Tengo los dos juegos de ruedas, tanto las cyclone de radios de carbono como las topgun con radios de acero. La topgun es igual de rigida que cualquier otra rueda con radios de acero, la cyclone en cambio si que es mucho mas rigida como cualquier rueda con radios de carbono. Las topgun son las de 50mm de perfil 24mm interno y 30mm externo. Adjutno las fotos de peso de cada una con fondo de llanta instalado.

With the build now complete, the bike weighs 6.68 kg, including the two bottle cages, Garmin mount, and pedals. The only remaining upgrade is a set of Heppe thru-axles; for now, it is still using the original ones.

Components Used
Ullicyc T4 Ultra integrated handlebar (380 × 110 mm)
Kore Cyclone wheelset with a 50 mm rim depth, 23/24 mm internal width, and 30 mm external width  3.2 mm carbon spokes
RideNow TPU inner tubes
Continental GP5000 28 mm tires
BigRock 165 mm crankset
PassQuest Aero chainrings
Xcadey Gen2 power meter
Lightweight Goldix 12-speed cassette (11–34T), fully steel construction
Enlee ultralight 9 g bottle cages
Vertical bottle cage adapter
Shimano 105 Di2 groupset
Riderace ceramic pulley wheels
UDH-style direct mount derailleur hanger
Walgun ceramic bottom bracket
Lightweight Riderace Center Lock lockrings
Titanium bolts throughout
Supacaz-style handlebar tape
Onirii pedals
Shimano MT800 160 mm brake rotors
Elita One saddle

ll right, thank you. I was aware that both of you are using SRAM, but surely there must be some people who have built their bikes with Di2. Naturally, I asked Alisa first, but after waiting almost three days for a response, the answer I received was: “We do not provide this with this frame.”
That was quite a disappointment.

At first, I wanted to purchase a mount, but unfortunately I ordered the wrong one(27,2mm size instead 31) and therefore couldn’t determine whether a standard mount would work. I have now had one 3D-printed, which fits perfectly.
If anyone is interested, I can share the STL file.

I also believe that my rubber seatpost cover looks excellent—possibly even better than the original one, don’t you think?

I'm waiting the frame. Could you share the STL file so I can print the Di2 battery holder?

They said the black decals dont exist yet and are in production. So if you really dont like the white decals and they are under a clear coat, then just wait.
@fhold: Can you tell us if the decals are under a clear coat? if not then i guess you could heatgun them away

https://es.aliexpress.com/item/1005010279504846.html?gatewayAdapt=glo2esp