Riding with Powertap Powercal

Today it’s my first day with Zwift:

I’ve started riding with my Powertap Powercal HR strap. I had low expectations because I read that things were not going to go adequately, but the experience was even worse than I expected.

I use a spin bike for indoor riding, so I have attached a cadence sensor also to my crank (considering that the speed sensor does not make sense because pedals turn at the same cadence than the main wheel, so it’s extremely easy to calculate the speed just by considering a circumference of the wheel for example: ~2 meters/revolution x 100 RPM x 60min = 12km/h). This speed is raw and imprecise, considering no gears and no resistance,  and also considering that in case that we offered a certain power with a different gear than the single gear that spin bikes provide, the RPM of the wheel could be way faster than the cadence of the pedal. 

But I’m saying this because of the cadence of the pedal in any Spin bike, is exactly the same as the cadence of the wheel, so it’s pointless to use a speed sensor in the wheel.

The calculation of the real speed could be done by estimating the % of the slope, the power, and the cadence: High power + High Cadence + Low % Slope = Extra Speed. 

I thought that the software considered this formula: Cadence x Power x Slope = Speed.

But I have discovered today, that the software it’s only using the Power output of the Powercal. Therefore, the bike moves even when I’m on the floor with 0 rpm cadence (so basically it’s saying me, that it doesn’t use the cadence for any calculation at all). 

Most of the power meters (pedal, crank or trainer), give a near exact calculation of the speed because the power-meters have a cadence sensor in the power meter installed and also, because power it’s only measured when there is movement in the meter, not by external factors.

But Powercal, obviously, does not have this cadence sensor because it only measures the power estimated, based on the HR variations.

So for example during a Zwift workout, the software said: “Go over 600kw for 10 seconds”: It was impossible: I could extremely increase the cadence of my bike, for example from 80 rpm to 130rpm and obviously this should have been a great variation in my HR (for example from 150 to 160 in 5 seconds). But if the software does not have any correlation between cadence and HR variation it’s impossible that the Powecal increases the total power output over 300-350 kw in that amount of time (10 seconds)

I understand that Powercal it’s not meant for Zwift, but I know for sure that there could be software solutions to offer a great experience by combining Powercal output with a cadence sensor for indoor spin bikes experience (or any other indoor bikes)

But being realistic, and considering there is not going to be a solution anytime soon, I would like to know if anyone has found an interesting alternative to have fun with Zwift while using this equipment.

Obviously, at this point I should discard all Zwift workouts because they are meant for adequate power-meters. I wondering alternatives right now, for example, doing an spinbike class while “having fun” riding in the Zwift landscapes. It’s not the full Zwift experience, but it’s an idea.

Any other ideas to share with me?

Thanks for reading!

All of your assumptions are wrong. Your cadence does not equal power, never has never will. Power meters on the crank arm can measure cadence but that is not used in any way to calculate power, they have a strain gauge build in that measures watts. Unless the stationary bike can handle a speed sencor and Zwift supports the wheel size or if it broadcast watts the stationary bike will not work correctly with Zwift. You could put on a set of pedals that measure power like PowerTap and that would make your stationary bike compatible with Zwift.


> Unless the stationary bike can handle a speed sencor and Zwift supports the wheel size or if it broadcast watts the stationary bike will not work correctly with Zwift.

As I said, the cadence in a spin bike it’s the same as the speed (or at least it’s calculable directly because it turns at the exact same cadence as the pedals). But as AFAIK in  Zwift you cannot configure the wheel size to estimate your speed from the cadence.

This is why speed = cadence = necessity to calculate the real speed in the game.

> All of your assumptions are wrong. Your cadence does not equal power, never has never will. 

I’m not saying that the cadence it’s equal power. I’m saying that when cadence = 0 the power = 0. Therefore (and just in this case) Cadence = Power. But in the scenario, I’m setting this is not true, because since the HR is never 0 the Powercal has some random power > 0 when cadence = 0, therefore there is a mistake from the beginning

But in the scenario, I’m setting this is not true, because since the HR is never 0 the Powercal has some random power > 0 when cadence = 0, therefore there is a mistake from the beginning that propagates in the low stances.

The Powercal do a good job calculating the power from HR variations. It’s an algorithm that works pretty well without big variances, in the spread from 100kw to 300kw approximately. 

If the workouts did not have big variances (as most HIIT workouts do), then the Powercal could work nearly good. But when the workout ask: “Go over 650kw in 10sec” it’s virtually impossible that the Powercal outputs over 350kw, because it’s impossible that your HR goes for example from 120 to 170 in 10 seconds that it’s what it’s needed to output 650kw in Powercal (not in real power-meter tho). 

It’s even difficult to stay over 350kw for long enough. The common power output for the Powercal goes in between 100-350 with an acceptable confidence I will say.

> You could put on a set of pedals that measure power like PowerTap and that would make your stationary bike compatible with Zwift.

As I stated before, I know that a set of pedals or a new crank/crankset with power meter integrated will solve this issue pretty well. But I’m wondering around the solution of Powercal + Cadence sensor scenario.

And a smart spin bike with all by default even better. But I’m talking about these specific tools to see if they could be feasible. 

From my point of view, there could be an easy approach considering that in Spin Bikes exclusively, the Cadence equals Speed for a given wheel diameter manually set (in this case). I acknowledge that It’s never going to be a real-bike scenario (without gears and automatic resistance, this it’s obvious), but it could be a good solution given the tools for low budgets.

Anyway, I was looking for alternatives to use this appropriately. I know that these modifications could never arrive, or are not easy to be implemented anytime soon.


Torque  is a measure of how much a force acting on an object causes that object to rotate.  Power  is a product of torque and rotational speed (i.e. cadence).


Zwift (and all biking) does not measure power based on cadence since you can pedal at different cadences and produce the same power. Power in cycling is messure (based on this thread) would be measured at ththe read wheel based on the speed of the rear wheel. You can messure power at the pedals, crank arm, crank and rear wheel, none of it is based on the cyclist cadence. I think you are trying to twist the thread to just troll, but that does not help at all. The OP is trying to get power readings based on cadence more or less alone and that is not possible the why he wants. The OP will need to add a power meter of some type or purchase anew trainer or stationary bike. The only way it could work is if the stationary bikes wheel size is support by Zwift and he attaches a speed sensor onto it.

“All of your assumptions are wrong. Your cadence does not equal power, never has never will. Power meters on the crank arm can measure cadence but that is not used in any way to calculate power, they have a strain gauge build in that measures watts”

Totes incorrect. The strain gauges measure torque, not watts. That’s why you can apply all your weight on the pedals at zero RPM and see zero power. Power is a calculation of force and rotation, so a crank based power meter uses cadence to calculate this. This is why SRM powermeters won’t work without their cadence magnet. 

OP does need a better setup though, he’s using a HR monitor based power estimator on a spin bike for Zwifting and intervals… Doesn’t matter what his cadence is, it’s not going to remotely accurate! And yes, Zwift uses power, not cadence, for forward movement.

Who are you anyways? I see you around on every question here, typically posting incorrect information and moderating. Someone asked about the disc wheel in game and you said it was for beta users only and it’s not coming back, only to have Jon Mayfield respond otherwise 😂

Jonathon Levie(Seattle)KRT,

Watts are what is used for power within Zwift and I did not want to add another measure to confuse the OP.

I am someone who has been thanked by a lot of people (including Zwift employees) for helping getting people setup and using Zwift. I do not post incorrect info and have helped many people. 

At that it had just been announced that the disk wheel would be coming back, before that Zwift stated that is was only given during beta testing and there were no plans on bringing it back. 

I see no issue with what I am doing.

Wow this went south pretty quick. 

Lets see if we could help Manuel. 

So It sound to me like you have 2 issues.

#1. The Zwift bike move even tho you are not pedaling. That is a short coming of your power meter, your hart beat and therefore it think you are generating power of some sort. Zwift does not require you to have a cadence meter you can see there are people riding with no legs moving. So just ignore that. Just keep pedaling. Sorry I dont have better words.

#2. The workouts ask you to produce high watts in short spikes that you are not able to reach. But I see you have a hard time reaching 350w for some time. I t sound like your in game FTP is set to high, lower your FTP to 100 and re do the FTP test using your HR power meter and see if the workouts are better.

Paul, you can explain how Zwift works without incorrectly explaining how power meters work. Simply put, Zwift uses power numbers it receives for it’s input, regardless of cadence.

Not related to Zwift, the way a power meter works is by using both a torque reading and rpm/cadence to calculate power. You keep saying the opposite, “Power meters on the crank arm can measure cadence but that is not used in any way to calculate power” and “Zwift (and all biking) does not measure power based on cadence since you can pedal at different cadences and produce the same power.” You can also pedal at the same cadence and produce different power… You can keep your power the same up a hill by riding at low rpm with lots of force, or with less force and a higher rpm, because watts are a calculation of both torque & rpm. A powermeter located on/near the cranks would calculate rpm from cadence, where as a Powertap Hub would be based on rpm of the hub. We both can agree this doesn’t apply to the OP’s question, he can spin as much as he wants but his power currently is based on HR.

OP has a potato setup which unfortunately isn’t suitable for what he’s trying to do on Zwift. A Powercal is not in the ballpark of being a useful powermeter. The device is essentially Saris’s version of Zwift power, but estimated based on heart rate. Since heart rate is slow to respond to any change in effort, all power estimates are delayed and wildly inaccurate, especially for short hard efforts. Attempting to hit power targets during 10 second intervals on a spin-bike with a HR pseudo-powermeter is equivalent to trying to run 4k Zwift on a graphing calculator. You’ll need to invest in better equipment.

There are plenty of pedal based powermeters that can be used on spin bike, and Zwift can also estimate power using one of the many compatible indoor trainers and a bike. http://zwift.com/get-started

@Gerrie, yes I know that the Power meter generates watts and this makes bike start moving incidentally. What I was commenting that there should be a feature implemented, that in case of cadence sensor is activated, and it’s outputting 0 RPM then the bike shall not move. Obviously in case that the user it’s using a cadence sensor, in other cases, it should take the speed of the speed sensor, regardless of the power, of the power meter, that it’s not as reliable as the cadence/speed sensor as we can see in this example.

This obviously should be meant, to include setup scenarios like this HR power meter, or other power meters in the future, that take in consideration external variables, apart from direct power). If Zwift developers are not interested in supporting this power meter at all then it should be stated clearly.

About the second thing, I’m going to check it, I’m not sure if I will be able to do FTP tests with this power-meter. This is something I’ve been looking in the past two days without clear conclusions yet.

@Jonathon, yes this is all true: Power meters have cadence sensors to calculate output power.

What I was saying, it’s that there could be some type of relation between the cadence and the power, for example, a multiplier. So as I stated in the beginning if the power it’s 150w and the cadence it’s 0RPM (0x) then the speed should be = 150x0 = 0 mph. But maybe this problem should be resolved by Powertap, not by Zwift, because it’s an issue of the HR sensor for this type of applications.

But I can’t agree with you in one thing: Now I’m trying to identify some setup where I can push it forward. I know that this power meter, it’s not the best at identifying instant variations of power. But I’, trying to figure out how to design new workouts that do not need big variations. Big interval training for example, that keep each interval for 2-3 minutes for example and without big jumps between intervals (or ascending/descending ramps sbetween each interval). Today I’ve found that there is an option to create personal workouts. I’m investigating in that way now to see if I can create something interesting

I think there should be more openness to trying new options. After some research, I’ve found out that this gadget could be really interesting for power training. It’s not the most complete training method, but better than going at constant 300w demo version.

This is some example of the possibilities it offers:


I understand that, for example, it’s impossible to race, but I was not willing to race near time soon. Maybe in 3-4 months, I will be planning to update to a power meter pedal system or see if I can upgrade my spin bike crankset with new power-meter cranks.

What I can’t understand it’s why someone has downvoted this topic, that it’s trying to introduce newer options to the software. Some people are completely awesome.

I understand your concern with the cadence, it sound like a relative easy fix for your situation to just ave a line of code saying _if,cadence = 0 , then speed = 0. _But for all other trainers that will be a huge issue. Because if your cadence sensor drop it signal you suddenly stop, it is bad enough that you stop when the power meter drop out now you double the chances of dropouts. What could work is if the manufacturer of the HR power meter will change their software so that you can set it where below a certain HR it send Zero power.

You should be able to do the FTP test with your HR power meter, it does not have sharp up and down segments, and the only thing that count is the 20 minute of free ride. 

I agree, that HR power meter look to be a nice gadget. it opens up a door to more fun training. It is not a power meter or a smart trainer, but for someone with a spinning bike it can make staring at the wall more interesting.

Yes but this it’s also easily solvable: If cadence = N/A then, not consider cadence for speed calculation :slight_smile: We should never want that new feature do create worse scenarios, shouldn’t we?

> You should be able to do the FTP test with your HR power meter, it does not have sharp up and down segments, and the only thing that count is the 20 minute of free ride. 

Ok, I’m going to see it. It’s there any documentation or manual of how to properly do it? Or just go into workouts -> ftp test -> and ready to go?


The cadence thing will only be noticeable when you start the game and when you are done, the rest of the time you keep pedaling all the time so it should not be to bad.


You can look at this, https://whatsonzwift.com/workouts/ftp-tests#ftp-test-shorter.

But it is easy you follow the workout, don’t worry to much if you don’t make the target values in the beginning the important part is the 20 minute free ride section. go as hard as you can for that section. 

After you know your ftp you will find the workouts are more realistic. The workouts are based on %FTP.

You don’t need a FTP number to just ride in the game, this is also fun, trying to keep up with other people or even going faster.


This web page also will help you get started, http://titaniumgeek.com/zwift-user-manual-unofficial-running-updates/


 You can find all the workouts on this page, https://whatsonzwift.com/workouts/ 

you should be able to find some that work for you.

Ok, thanks for the manual. I’m going to try this test to see if I can get an starting point.

Now I’m still also looking for specific workouts that may work fine with the sensor. Essentially workouts without the need of big fluctuations.

Hi Guys, 

I am a beginning rider and I would like to use my spinning bike. I have contacted tech support but their response was to try to use the power cal. They respond quick but my feeling is that there is not a whole lot of documentation about spinning bike support. From previous forums I read that the power cal is unreliable, so I am not sure if this is the best option. I really like the software and I would love to use my spinning bike but I am having so many hiccups with the set up and I cant get it right. I will greatly appreciate your help. 

Here is my current set up: 

Bike: Spinner Ride from Spinning ( manufactured by precor)

Sensors: Wahoo Speed and Cadence 

I placed the wahoo speed sensor on the flywheel and I can see the rider moving faster but I am unable to control the speed accurately. ( The cadence is placed next to the pedal and cadence seems to be working fine) 


I just finished racing and when the software said to reach 200, I will reach it and shortly after that it will jump to 400. I will then try to slow it down and it will quickly go down to 160.


I am trying to simply understand how to improve the ride without getting into intricate calculations/technicalities. I will greatly appreciate your help




The problem of the spin bikes (as I’ve stated and you may see by yourself), is that the revolutions per minute (RPM) of the flywheel are exactly the same as the RPMof the pedals. 

This is a big problem for a simple reason: you cannot measure reasonably the power because you don’t have enough references. Zwift uses a formula to calculate the power using the RPM of the wheel + RPM of the the pedal (cadence). Considering that if you pedal slower than the revolutions of the wheel (in a classic bike with gears), this gets interpreted by Zwift, that the resistance of the gear is big, therefore you are putting more pressure in each pedaling, therefore, more Power.


The RPM of the wheel it’s 100

The RPM of the pedal it’s 60

This means you are going at 200W (not real, just an example)

If you change the pedal to 70 (but the wheel it’s still at 100RPM), this means that the force you are putting on the pedal is less (lower gear) therefore the power goes down to 150

If you keep your RPM at 60 but the wheel goes to 120, then your have increased your gear, therefore more power maybe to 250W

But on a spin bike, this is impossible. If you increase your cadence, the RPM of the wheel automatically directly proportional increases. So the power measures are straightly lineal. If the Speed/cadence algorithm it’s incredibly faulty, then with only 2 variables (that are exactly the same), the algorithm becomes even more faulty. 

And the biggest problem is that the gear in a Spin bike it’s fixed BUT the resistance it’s variable through an independent knob. In a normal bike the resistance can be approximately extrapolated from the gear difference between the wheel RPM + pedal RPM. But in the spin bike you can be pedaling at 100RPM (therefore 100RPM in the flywheel because let me remind that RPM of the flywheel == RPM of the pedals) with a really strong resistance (therefore a HUGE power) and you can be pedaling at 100RPM with no resistance (therefore nearly 0 power). So the system can be saying that your speed it’s 30mph with 250W but the real power it’s 20W (because of the absence of resistance + the inertia of the flywheel), so you will see your dummy going really fast in the game, and you doing barely any effort on the bike. And vice-versa: You can be going up a hill at 10mph with only 100W in Zwift, but your real power its 300W (because you have put a really high resistance). Zwift will consider that the slope of the hill it’s too tilted for only 100W, therefore, your dummy will be stuck in the hill without moving.

The only solution that exists here, its calibrating your spin bike to match your break resistance with the power you put in each pedal. This will be the most awkward setup ever because it will mean that you won’t be able to touch your resistance knob ever (or you will break the “equation”). This is technical, but essentially what you have to do, it’s to rent some power pedals, and using two different power measure software (for example Zwift and a the Garmin app for the pedals), move the resistance knob until the power of the speed/cadence sensor in Zwift matches the power you are putting on the pedals in the Garmin app.

And never touch that resistance knob again. The worst part is that if you have a mechanical resistance/brake (not an electromagnetic), the mechanical part will wear down over the time and you will need to recalibrate periodically: So awkward as I’m saying.

So briefly and in conclusion with that setting and a spin bike, it will be really bad experience in Zwift. 

Same I could say with Powercal. Even powercal from my point of view, it’s way better than a Speed/Cadence setup (in a spin bike), because at least, they have designed an algorithm to analyze the mean of the power through a session (normalized power). So despite Powercal it’s really shitty for in-session power adjustments (and it’s virtually impossible to put the power over 400W for more than 5 seconds or to stabilize the power under 150W), you can see approximately your mean power in a full session with certain accuracy.

You won’t be able to do even a proper FTP test with a Speed/Cadence setup in a Spinbike but with the Powercal a FTP test it’s possible, because these tests measure your average power over big periods of time (5 minutes per increment). For anything over 2 minutes over 150W, PowerCal it’s extremely good. And the FTP tests are exactly this: ~20 minutes testing over 150W.

Currently I’m not using Zwift. I’m using a different software with my Powercal that fits adequately with Powercal, and maybe in the future, when I can purchase some type of real powermeter (probably some pedals like Vector or Powertaps, because the cranksets do not fit my square type cranks), I will come back

Zwift power estimates works because these trainers have a set resistance curve. If the spin bike/trainer has adjustable resistance then Zwift can’t assume a power output at a given speed unless the resistance is set correctly as specified in the link attached below. The resistance curve isn’t based on cadence from what I can tell.