4iiii’s has kicked off what will be a frenzy of power meter activity here at Interbike this week. In doing so, as the post headline implies, they’ve jumped onboard with the plans to further drive the price of power meters down into a larger cycling audience. They’ve introduced their latest product – Precision – which is their power meter offering that sits at $399US (with dual left/right sensor support at $749US).
I’ve had a chance to ride with the units on a ride, as well as to spend considerable time doing everything from installing the unit myself to calibration of the sensor. And, I did this all on a decked out power-meter laden bike. Just me, Precision…and three other power meters. Of course, a single data point of one ride doesn’t mean the product will be flawless (or failure bound). Rather, it’s just a first glimpse at things. And with that, let’s begin the walk through.
A Brief Overview:
The Precision is in many ways a lot like the Stages Power Meter, except with two key differences: It can be installed yourself, and it can be installed on both crank arms (versus just left-only with Stages). Well, that and being about 40% cheaper (or more, depending on how you purchase it).
Precision is all based around these two tiny pods. These prototype pods are slightly larger than your average running footpod, and the bulk of the size is really dedicated to the CR2032 coin cell battery contained within. The actual innards of strain gauges, accelerometers, and ANT+/Bluetooth Smart communications chips are quite small. However, the next prototype versions (and production versions), will significantly reduce the sides, effectively eliminating the sloped section (which currently contains additional wiring connectors for development logging capabilities).
These pods are then installed on the inside of your crank arms. You can install it in a left-only configuration, as seen lately on a number of power meters (Stages Power, ROTOR Power LT, Garmin Vector-S, Polar Keo Power Essential), or you can install it in a dual-sensor configuration.
In a dual sensor configuration you install it on both the left and right crank arms. This nets you a complete picture of your power output (watts), and thus minimizes issues with left/right balance.
These pods are installed by you (yes, you) using a calibration tool that’s included inside the kit. The pods are basically cemented on. You can remove the pod from the crank arm, but the process does destroy the pod. However your crank arm would be fine (Tip of the day: Would likely be cheaper to just buy a new crank arm).
Precision communicates with your head unit using either ANT+ or Bluetooth Smart. So it’s just a matter of which protocol your head unit uses. It’ll transmit all your standard power meter metrics on both sides, but I’ll dive into that a bit later on.
Now those that have followed the moves and players in the industry may find it interesting to know the person behind 4iiii’s Precision: Keith Wakeham. Those with good memories will remember it was Keith that about 2 years ago published plans for his open-source power meter design, which was a DIY approach to building a power meter. While that unit has no resemblance to today’s Precision, the learning from Keith’s efforts no doubt contributed heavily to it. Prior to that, he worked in the nuclear industry on strain gauges. 4iiii’s snatched him up last fall and they’ve been plugging away since. The power meter world is indeed a small place.
With that, let’s begin installation.
Installation & Calibration:
I want to first point out that the tools you see here are NOT consumerized yet. Which means that while the process is final, the actual tools that you’ll get in the package will be smaller, more user friendly and generally look ‘pretty’. They haven’t begin the manufacturing of those tools yet, since that’s about the last thing you’d do in terms of manufacturing.
And, despite how this process looks – it was actually silly easy. You can complete it in less than 5 minutes…at most.
For the installation I completed it on a single crank arm on the desk of a Fairfield Inn hotel. No fancy workbench was needed, and while I effectively had a nuclear engineer guiding me through it – there’s no need for that either.
Every kit comes with these tools (or again, a variant of them). they are:
1) A clamp
2) A two-part epoxy
3) A tool to squirt out the epoxy
4) A guide to fit the pod to the crank arm.
You can see all of them below:
To begin, you’ll need a crank arm, or two, depending on how many pods you’re installing. You’ll be using this little tool that helps to align the best installation spot for the pod on your crank arm.
The tool then holds the Precision pod in a perfect little grooved area. The pod’s backing is covered with a protective tape.
The first step is to take a pencil and simply mark the area on the crank arm where the pod will go.
We’ll then use the included sandpaper to lightly sand this area, which helps to increase the ability for the glue to hold.
Then we simply clean it off with a couple of swipes of swabs, which removes the sanded material. There’s also one more swipe with a gauze pad.
Now we’re ready to add a tiny little bit of glue. The glue gun will automatically mix the two epoxy’s perfectly, without any air bubbles.
You don’t need much epoxy.
You’ve got approximately 20 minutes until the epoxy starts to set. So you don’t have to worry about rushing. Heck, you could even go out and run a fast 5K.
Next, you’ll go ahead and remove the protective tape backing on the unit and place it in its tray:
Then, simply flip it over onto the crank arm. The tool settles into the pedal attachment area, keeping things snug.
Finally, we add a clamp. The clamp has a little rubber piece on the opposite side to protect against scratches.
You’ll leave the whole rig alone for 6 hours. Depending on the temperature outside (if hot), it can be a bit less. But really, just leave it alone. The photo you see throughout the post with the crank arm by itself were all taken the morning after with the completed unit.
Now, the piece I didn’t get to cover here was the calibration. The calibration will use a separate tool that has a ‘S’ type load cell in it. The tool comes in the kit, and is only used to calibrate. The tool will have three little grooves where you can hang a weight from it. The load cell can determine the weight within .01% accuracy, which then allows them to further calibrate the power meter. The user will place the weight at each of the three groves briefly, which in turn provides just enough differences in torque to properly calibrate the unit. The whole process should only take a couple minutes at most.
The unit is smart enough to determine whether it’s assigned to the left or right side, based on the torque and direction. Additionally, it’ll use private-ANT to pair itself up to the secondary pod if you bought that.
Connecting to a head unit, usage:
The 4iiii’s unit connects to head units using either ANT+ or Bluetooth Smart. To start, we’ll go with ANT+ since it’s the most dominant in the cycling world. For that, I just used an Edge 810. Of course, you can use any head unit that supports the power meter protocol.
Next, I’ll go ahead and have it search for a power meter to pair, where it’ll find the 4iiii’s on my bike:
Next, to prepare for a ride we’ll go ahead and calibrate it by setting a zero offset, similar to other power meters. This is typically done within the settings menu of the bike profile.
From here, you’re pretty much ready to go. While riding you’ll get total power displayed, as well as individual left/right power (assuming you purchased the dual sensor package).
All of this data is recorded by your head unit for later analysis in whatever apps you normally use, like Training Peaks or Strava:
Now, if you’re more of the Bluetooth Smart persuasion, you can pair with that as well. For example, the Polar V650 pairing with the 4iiii’s unit. Alternatively, you could use the Wahoo Fitness app on the phone which allows you to more easily send data to all sorts of 3rd party partners.
From a usage standpoint, it’s very similar to other power meters. You’ll complete a zero-offset at the beginning of a ride, but beyond that it’ll pretty much take care of itself.
Looking at the pod itself, there will no doubt be very valid questions on the adhesive and ensuring that it stays where it should, and that it isn’t impacted by weather or the accidental slip of the foot. All issues I’ll definitely be looking into as part of my in-depth review. In talking with 4iiii’s, it’s an area they’ve spent a lot of time in with respect to the specific glue chosen. Out of all the areas that they talked about where things weren’t quite finalized/firm yet, this was the one area they seemed very sure of.
The Test Data:
I started out this morning prior to sunset. Given my time constraints of the day I went with a single 40-minute ride for today, along with a secondary short section on the trainer. The ride was done on park roads in Lake Mead National Recreation Area. The same place I often do testing of devices when I stay in Vegas. The route is also used by a number of triathlons too.
Here’s the illustrated version of the power meters used in this morning’s tests:
As you can see, I’ve got the following units on the bike. We didn’t use my bike due to not quite having enough current generation prototype units to part permanently with. I did however bring my PowerTap to install, and we unboxed a set of Vectors that we installed on the bike together. The SRM and 4iiii’s unit was already in place. So I had complete control over two of the power meters, and they had prepared the other two.
SRM: Crank spider based
Garmin Vector: Pedal based
PowerTap G3: Rear wheel hub based
4iiii Precision: Left/right crank arm based
All of this data is then fed into two different spots. First is to conventional cycling head units, in my case the Garmin Edge 1000 and 810 for the Vector (1000) and PowerTap (810) data. And two additional 810’s for the SRM and 4iiii’s data. I have a few of them in my collection. These units all act as ‘backup’ of my data.
Second, it’s also fed into the much more important tool, the North Pole Engineering WASP ANT+ data consolidator. It allows me to seamlessly record everything as one coherent and complete picture. This removes the complexity of matching up files afterwards. The WASP then transmits to the phone via WiFi, where the data is recorded in real-time.
Without further ado, here’s what that ride data looked like at the ‘50,000’ ft view:
I excluded the SRM data, since for whatever reason the WASP wasn’t collecting that stream. So I need to manually re-process and normalize that data from the backup data files, which I’ll do later today.
Looking at the data however, you see at a high level things trend fairly well. You see that the Precision is a bit lower than the others. In talking with Keith, we saw that beforehand and believe that’s due to some minor drops he’s seeing with that particular pod, thus causing a lower overall value. You see one unexplained spike towards the very end, oddly enough as I was coming into the parking lot coasting.
Now, let’s dig a bit deeper into the differences between them. After all, I’m more interested in where they differ than where they are the same. This is where it’s important to remember that no two power meters will read identically – no matter how hard you will try. There are inherent accuracy limitations (usually 1%-2.5%), along with differences in placement. This means that a PowerTap will inherently read a lower wattage than a Garmin Vector – as Vector is closer to the source of the power (you). In order to make this graph make sense, I’ve tossed out those last few seconds of the ride (since that would hugely skew things and was an outlier).
First, here’s all three smoothed with a 10-second rolling average. I’ll use that as the baseline:
Next, we’ll look at the CP, which helps us to identify how they trended over chunks of the ride. I used Golden Cheetah to create this. Below you’ll see all four power meters (including the SRM). You’ll see that by and large they are pretty similar. We see slightly more variation at the top-end, with the 4iiii’s unit being low throughout (as expected from earlier notes).
You can see down at the bottom the average power values (be careful of looking at that), along with the max power values for a few different timeframes from 10-seconds up to 8-minutes. This helps a bit to make those values more relatable.
Finally, let’s look at cadence.
Eek. As you can see, this is a bit all over the map. And, to be fair to them – they expected that. The noted the night before the ride that they had just switched the component doing cadence to what they’ll use in the final production version and they hadn’t had a chance yet to optimize the algorithms there. Of course, in most power meters cadence is a key factor in determining power. Yet, I’m not as concerned about this particular component, only because accelerometer based cadence is at this point well understood in the cycling industry. So it’s mostly a case of them optimizing their new component algorithms as many other companies have recently done (Garmin, Stages, Power2Max, Polar, Brim, etc…).
I also did a short 9 minute test on a trainer, just to get some ‘smooth’ data to see what it looks like.
I’ll be uploading that data a bit later today.
Oh, and for those wondering about that gap in the middle around the 30 minute marker on the outdoor ride? That’s when the derailleur decided to try and establish a close and continuing relationship with the spokes on the wheel, thus injecting the chain down in the gap between the cassette and the spokes. It wasn’t pretty undoing.
You are welcome to download the raw data for your own analysis, all of it is located here. You may publish your own analysis of the data, though, I do ask you link back to the source post here. If you want a simpler version, you can just look at the Garmin Connect file online here.
Misc Details & FAQ:
Below are some random details and facts, along with what I suspect will be the most common questions.
Crank Compatibility (general): More or less anything with a flat surface. Conversely, I couldn’t however use my ROTOR crank arms on another bike, because those arms have grooves that the gauges wouldn’t be able to measure correctly. They’ve tested the last three generations of the Shimano 105, Ultegra and Dura-Ace cranks, as well as the Deore XT. Within SRAM, the Rival OCT, and then on the mountain bikes the X9. Within FSA the Energy crank arms.
Carbon Cranks: Thus far their testing “looks good” according to them, but they’re continuing to test cranks there.
Crank Length/Width: There are no inherent issues here with larger or oversized cranks. The only potential tricky area would be if your crank arm has some sort of weird bending or shaping that leave it no flat surface. For example, on the VERVE Infocrank (another power meter), it curves around and thus there isn’t any flat space to mount the pod. But that’s extremely rare (and likely not an issue for anyone in the world except me).
Chain and frame clearance: You’ll need approximately 9mm of clearance for the final production pod. Of course, where you install it will change the clearance required. Typically there is more clearance near the bottom bracket, and less clearance near the end of the crank arm. from a length standpoint, the exact length is still being finalized and will roughly be in the range of ~28mm wide by ~40mm long.
The Calibration Kit: This is sorta the funny special sauce. This is used to calibrate the sensor with hanging weights and a ‘S’ type load cell that will attach to your crank arm. The kit by itself costs $100, whereas it’s included in the $399 price point (but not the $349 pod package).
Temperature Compensation: The unit utilizes “active learning temperature compensation”, which is a fancy way of saying that they record zero offsets at different temperatures and slowly build up a database of different temperature compensation points. This will be triggered by you, and helps 4iiii’s get around the fact that the install isn’t being temperature calibrated at a facility elsewhere.
Battery Life: The unit utilizes a CR2032 coin cell battery – the same as most other sensors on the market. You can find it for a couple dollars easily around the world. 4iiii’s claims a battery life of about 200 hours. Obviously, I’ve not yet had a chance to test that.
Weight: Each pod weighs 10g. I have not yet personally verified this, though I have no reason to doubt it.
Weatherproofing: IP65 rated, which is slightly better than IPX7, as it protects against water jets versus just immersion. Thus allowing you to spray down your bike.
ANT+ Connectivity: The unit is able to transmit via ANT+ to any other standard cycling head unit that supports the ANT+ power meter profile. Today that’s every Garmin cycling and triathlon focused device above $200, the PowerTap Joule, Timex Cycling and Triathlon units, Magellan Cyclo and Up units, and many others. The unit supports sending the following metrics: Total Power, Left/Right Power, Torque Effectiveness, Pedal Smoothness. Of course, you must have a power meter head unit that supports those specific metric types. It broadcasts both Bluetooth Smart and ANT+ concurrently at all times (no impact on battery).
Bluetooth Smart Connectivity: The unit is able to transmit via Bluetooth Smart to any cycling head unit that supports the Bluetooth Smart power meter profile. Right now this is essentially the Suunto Ambit3, the Polar V800 (later this month), the Polar V650 (October), and then smart-phone driven apps and devices like the Wahoo RFLKT. The unit supports sending the following metrics: Total Power, Left/Right Power, along with some similar metrics for torque effectiveness. Of course, you must have a power meter head unit that supports those specific metric types. It broadcasts both Bluetooth Smart and ANT+ concurrently at all times (no impact on battery).
Pricing: The unit is priced at $399US for one pod (left or right only), or a dual option at $749US (both left/right pods). You can upgrade later for $349US with the PM-100 (pod without calibration kit). Pricing outside the US is TBD. The SKU name for the single pod is the PMK-100 (with calibration kit), and the left/right secondary pod is PM-100. The acronym comes from: : Precision Module Kit. Note again that you can install a single pod on either side. If a friend/shop/etc has the 4iiii’s calibration tool, then you can simply get the cheaper $349 SKU (PM-100) and use their calibration tool.
Shipping Date: 4iiii’s states they will begin shipping this year (Q4 2014). I should point out that in the 7 or so years I’ve been writing this blog only one power meter company hit their initial timeline estimates: Stages Power. They did so by having a similar plan as 4iiii’s in terms of waiting until the ‘last minute’ to announce.
Some summarizing thoughts:
At Eurobike, virtually every industry person asked me what I thought of left only power meters. My response was basically the same to everyone: I believe that left only power meters are in some ways a temporary thing. While Stages set the precedent, they did so not just in data collection location (left), but in price. But despite common thinking they didn’t put the price elevator at the lobby, but rather about 20-30 floors up from where their elevator can go. For now, that elevator has sat there and Stages has raked in the dough and outsold other power meters in some cases 10-20 to 1 (in talking with many local bike shops).
You saw other companies join in the left-only ‘thang’ by introducing their own variants at cheaper and more accessible price points. Then we saw a new entrant – Watteam’s PowerBeat – jump into the fray with promises of a $499 power meter for next summer. Meanwhile at the higher end, we’ve seen the price of Garmin Vector temporarily drop $400, while also introducing cheaper left only options. And we’ve seen SRAM’s Quarq restructure prices as well to lower them.
While some might think the end-goal here is a $500 left-only power meter, they’d be hugely mistaken. We’ll also be seeing more entrants in the lower-priced but higher data quality market by time the week is over, which starts to create a cyclical effect. Left-only units will drive current prices lower further, and left/right will take its pricing place. Then the cycle will repeat. It’ll repeat for the simple fact that companies like 4iiii’s and Stages have the manufacturing cost basis power to drive those prices incredibly low.
Of course, just like Stages – that’ll depend heavily on these new entrants (such as 4iiii’s) having a product that’s accurate across a wide range of riders and equipment, and a wide range of environments. And I suspect also like Stages, they’ll too have some initial teething bumps along the way. All power meter companies do.
My testing showed initial promise, not perfection, but promise. But it’s just that – a first ride. Once I get v2 prototype units on my bike I’ll be able to develop a much bigger bank of data to work from and establish whether or not the Precision and their self-install method is a viable option in the market. I’m actually less concerned about 4iiii’s ability to technically resolve any outstanding issues. It’s a beta product, if it was production they’d be selling them tomorrow. In my discussions with them about their outstanding technical issues, it felt to me like they had a very clear grasp on what had to be done, but it was more of a timing thing. They had been focusing on OEM installations (i.e. from a crankset company), and then changed gears to be self-install – so their attention got diverted a bit in the last month.
I also don’t have any concerns about manufacturing of Precision’s pods itself – 4iiii’s has a long history in manufacturing already. My only concern is that I don’t quite think they’ll get all this done by the end of the year. I suspect we’re likely looking at January-March of next year. Though, I suppose there’s likely not a lot of difference riding-wise between December and February (no matter which side of the equator your on).
For now though, it’s definitely an exciting time to be in the power meter market.
With that, thanks for reading! If you’re looking to burn a bit of time, here’s all my Interbike 2014 posts packed full of sports technology items.