Go Back  Bike Forums > Bike Forums > Electronics, Lighting, & Gadgets
Reload this Page >

Bicycle computer that computes power from speed and cadence?

Notices
Electronics, Lighting, & Gadgets HRM, GPS, MP3, HID. Whether it's got an acronym or not, here's where you'll find discussions on all sorts of tools, toys and gadgets.

Bicycle computer that computes power from speed and cadence?

Old 04-26-17, 11:58 AM
  #1  
Junior Member
Thread Starter
 
Join Date: Oct 2016
Posts: 5

Bikes: Dean Ti Custom

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Likes: 0
Liked 0 Times in 0 Posts
Bicycle computer that computes power from speed and cadence?

I'd love to find a cycling computer that uses BLE speed and cadence to compute power in watts. Does anyone know of something like that? I see computers like Cateye's Strada and Padrone use BLE to feed the computer but it appears they want a power source to display watts.
Velocedge is offline  
Old 04-26-17, 12:36 PM
  #2  
Senior Member
 
caloso's Avatar
 
Join Date: Jul 2002
Location: Sacramento, California, USA
Posts: 40,865

Bikes: Specialized Tarmac, Canyon Exceed, Specialized Transition, Ellsworth Roots, Ridley Excalibur

Mentioned: 68 Post(s)
Tagged: 0 Thread(s)
Quoted: 2953 Post(s)
Liked 3,106 Times in 1,417 Posts
I don't know of any computer that does that. And just wondering how it would know whether you were riding down hill or with a tailwind?
caloso is offline  
Old 04-26-17, 12:55 PM
  #3  
Senior Member
 
CliffordK's Avatar
 
Join Date: Nov 2014
Location: Eugene, Oregon, USA
Posts: 27,548
Mentioned: 217 Post(s)
Tagged: 0 Thread(s)
Quoted: 18430 Post(s)
Liked 4,548 Times in 3,379 Posts
Power is somewhat independent from cadence. So, one could put out 200W power at 50 RPM or 100 RPM, just applying a different amount of force at the pedals, or applying the force for a longer period during the stroke.

Strava provides post analysis power calculations primarily based on speed, I think.

Power and speed is complex. Just looking at average speed, it is not very good as one has to account for hills, wind direction, drafting, and etc. Some could be captured such as elevation, although GPS satellite elevation isn't very accurate, at least for cell phones. However, speed on a bicycle is complex and involves acceleration during the part of the stroke while the cranks are parallel to the ground, and deceleration when they are vertical, with the shape of the power curve dependent on the rider and whether standing or sitting. It is my belief that if one could get an extremely accurate speed calculation, one could derive factors such as drag, and determine power. However, the sensors to provide accurate enough of data might be as complicated as just getting a real power meter.

There have been a few attempts to use non-power information to estimate power. Here is a rather long discussion about Aerofly. Lots of speculation. There is also mention of a few other real-time power estimation systems in the thread, each with their problems.

https://www.bikeforums.net/road-cycli...sts-129-a.html

I've got the Arofly. It was interesting for testing, but makes enough errors that it wouldn't be particularly useful for training. Some of the errors are predictable. Some are less so. The device has a lot of problems with calculating power on hills, both ascents as well as descents under power.
CliffordK is offline  
Old 04-26-17, 12:58 PM
  #4  
Senior Member
 
wphamilton's Avatar
 
Join Date: Apr 2011
Location: Alpharetta, GA
Posts: 15,280

Bikes: Nashbar Road

Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 2934 Post(s)
Liked 341 Times in 228 Posts
I've been thinking about that. If you know the bike's instantaneous velocity and acceleration, and the magnitude of the vector from an accelerometer, you can derive the immediate slope of the road.

With a couple of data points, near enough in time that the rider position and apparent wind do not change significantly during the period, you could in theory solve for CdA and Crr (with Crr likely not changing much or at all). Perhaps during a deliberate coast (zero power) or else by utilizing the zero point of the power stroke. Knowing what those are dynamically, I think you might be able to solve for power and apparent wind speed. It would all be approximate of course, and a better application of these ideas would be measuring power and solving for instant CdA, but with enough computing power I think it is possible.
wphamilton is offline  
Old 04-26-17, 01:01 PM
  #5  
Junior Member
Thread Starter
 
Join Date: Oct 2016
Posts: 5

Bikes: Dean Ti Custom

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Likes: 0
Liked 0 Times in 0 Posts
Yeah, I understand all the variables... just looking for something in the ballpark. If you're attached to the phone/GPS then you have a little more information. Zwift does it (albeit without wind). I'm looking for something like that. Not perfect, just a benchmark figure that works in the long term scheme of things.
Velocedge is offline  
Old 04-26-17, 01:28 PM
  #6  
don't try this at home.
 
rm -rf's Avatar
 
Join Date: Jan 2006
Location: N. KY
Posts: 5,977
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 987 Post(s)
Liked 530 Times in 365 Posts
Zwift does it, using the known power curves for various bike trainers. The only (small) variable here is how much power is lost with tire flex. Otherwise, it's strictly how fast the trainer is spinning, measured using rear wheel revolutions per minute. (I've seen mentions that the fluid resistance changes a little as it warms up, though.) A trainer has no wind resistance or gravity effects at all, and those are huge when riding outside.

For instance, Kurt has this power vs speed chart. 400 watts at 25 mph, for example.


Last edited by rm -rf; 04-26-17 at 01:36 PM.
rm -rf is offline  
Old 04-26-17, 01:36 PM
  #7  
Senior Member
 
caloso's Avatar
 
Join Date: Jul 2002
Location: Sacramento, California, USA
Posts: 40,865

Bikes: Specialized Tarmac, Canyon Exceed, Specialized Transition, Ellsworth Roots, Ridley Excalibur

Mentioned: 68 Post(s)
Tagged: 0 Thread(s)
Quoted: 2953 Post(s)
Liked 3,106 Times in 1,417 Posts
Originally Posted by rm -rf
Zwift does it, using the known power curves for various bike trainers. The only (small) variable here is how much power is lost with tire flex. Otherwise, it's strictly how fast the trainer is spinning, measured using rear wheel revolutions per minute. No wind resistance or gravity effects at all.

For instance, Kurt has this power vs speed chart. 400 watts at 25 mph, for example.

Which is why this works on a trainer but not on the road.
caloso is offline  
Old 04-26-17, 01:53 PM
  #8  
Junior Member
Thread Starter
 
Join Date: Oct 2016
Posts: 5

Bikes: Dean Ti Custom

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Likes: 0
Liked 0 Times in 0 Posts
Yeah, again not looking for something perfect... if I did, I'd by a power meter. Looking for something to give me a "feel" for the power I'm putting out. Don't care if it's "real" as long as in the big scheme of things it all averages out. I can easily enter height, weight, age, sex, speed, cadence, etc. into a computer program and come up with a guesstament. That's all I'm looking for. So, going back to the original question, do you know of a computer that does this? If not, then no need to answer. Calso, your answer, "I don't know of any computer that does that" quite sufficient!
Velocedge is offline  
Old 04-26-17, 01:56 PM
  #9  
Senior Member
 
Join Date: Feb 2011
Location: Minnesota
Posts: 4,673

Bikes: N+1=5

Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 875 Post(s)
Liked 245 Times in 181 Posts
Originally Posted by Velocedge
Yeah, again not looking for something perfect... if I did, I'd by a power meter. Looking for something to give me a "feel" for the power I'm putting out. Don't care if it's "real" as long as in the big scheme of things it all averages out. I can easily enter height, weight, age, sex, speed, cadence, etc. into a computer program and come up with a guesstament. That's all I'm looking for. So, going back to the original question, do you know of a computer that does this? If not, then no need to answer. Calso, your answer, "I don't know of any computer that does that" quite sufficient!
Strava does the best, IMO. They seem to have captured data from lots of riders with power meters over the same terrain you're riding. They know the rider's weight, the speed, cadence and the grade so they can compute it. There is insufficient information in speed and cadence to even estimate the power.

J.
JohnJ80 is offline  
Old 04-26-17, 02:10 PM
  #10  
Senior Member
 
CliffordK's Avatar
 
Join Date: Nov 2014
Location: Eugene, Oregon, USA
Posts: 27,548
Mentioned: 217 Post(s)
Tagged: 0 Thread(s)
Quoted: 18430 Post(s)
Liked 4,548 Times in 3,379 Posts
Like I said above, the Arofly does it, more or less, and calculates cadence from its speed estimates. But it has errors for example assuming coasting when one is pedalling or visa-versa.

The thread I linked to above had a few other bike computers that do it too. Yeah, buried in too many pages. Each with their problems.

Strava provides watts, but only with post analysis.

If riding on reasonably flat roads, then I believe one could use pure speed for pacing just as well as using power. Just learn what it feels like, and your personal endurance at 15 mph, 20 mph, 25 mph, or 30 mph. So, use speed and what the ride feels like while riding, then look at Strava later..
CliffordK is offline  
Old 04-26-17, 02:33 PM
  #11  
Senior Member
 
bbbean's Avatar
 
Join Date: Jul 2013
Location: Missouri
Posts: 2,704

Bikes: Giant Propel, Cannondale SuperX, Univega Alpina Ultima

Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 678 Post(s)
Liked 431 Times in 259 Posts
Originally Posted by Velocedge
I'd love to find a cycling computer that uses BLE speed and cadence to compute power in watts. Does anyone know of something like that? I see computers like Cateye's Strada and Padrone use BLE to feed the computer but it appears they want a power source to display watts.
Any app can compute estimated power.
__________________

Formerly fastest rider in the grupetto, currently slowest guy in the peloton

bbbean is offline  
Old 04-26-17, 02:37 PM
  #12  
Senior Member
 
caloso's Avatar
 
Join Date: Jul 2002
Location: Sacramento, California, USA
Posts: 40,865

Bikes: Specialized Tarmac, Canyon Exceed, Specialized Transition, Ellsworth Roots, Ridley Excalibur

Mentioned: 68 Post(s)
Tagged: 0 Thread(s)
Quoted: 2953 Post(s)
Liked 3,106 Times in 1,417 Posts
Here's a comparison of Strava's estimate vs. actual power data on my local Saturday group ride. I picked a long flat segment (5 miles, dead flat). Same segment, same speed, same rider, same group, same bike. Only difference is on one ride I had my PM and the other is estimated.

Actual power:

AVG MAX
Speed 27.9mi/h 32.2mi/h
Cadence 88 110
Heart Rate 172 bpm 181 bpm
Power 255W 767 W
Elev Diff 7ft


Here's the estimate from another ride:

AVG MAX
Speed 27.9mi/h 35.3mi/h
Cadence 0 0
Heart Rate 175 bpm 186 bpm
Est Power 393W

This shows the huge difference that drafting in a group makes. It's possible that on one ride I spent more time on the front or that there was a huge tailwind, but I don't think either would account for a 50% difference between recorded and estimated power.
caloso is offline  
Old 04-26-17, 02:56 PM
  #13  
Senior Member
 
Join Date: Sep 2015
Location: QC Canada
Posts: 2,008

Bikes: Custom built LHT & Troll

Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 873 Post(s)
Liked 153 Times in 110 Posts
Originally Posted by Velocedge
a computer that does [compute power]
I've used this bike calculator. It is designed to compute speed from power and other variables, but can easily be used to compute power. Simply enter known parameters (weight, riding position, tires, wind, grade) and iteratively adjust power until your actual speed matches the predicted speed.

It gave me a reasonable ballpark figure, confirmed with a Stages PM.

---

Now, calculating power is somewhat like trying to calculate your heart rate based on speed, age, etc. We could probably predict that your heart rate will be around 120 on a level course at a leisurely pace, and 165 when you climb a 10% grade. But the reason why people wear an HRM is to get an accurate measure that will indicate by how much their rate improves as the are getting more fit, or whether they are pushing to envelope a bit too much, or to help pace a run, etc.

In the end, a calculator, because is makes many assumptions, may not be so useful.

I'd guess that if you are not an athlete, you probably ride at 160W minus your age (+/- 20W). Add 50-100W if you train hard.

Last edited by gauvins; 04-26-17 at 04:30 PM.
gauvins is online now  
Old 04-26-17, 03:07 PM
  #14  
Senior Member
 
Seattle Forrest's Avatar
 
Join Date: Mar 2010
Location: Seattle, WA
Posts: 23,208
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 18883 Post(s)
Liked 10,646 Times in 6,054 Posts
Originally Posted by caloso
I don't know of any computer that does that.
Any modern Garmin (that runs even the first version of ConnectIQ) can. You just wouldn't want to.
Seattle Forrest is offline  
Old 04-26-17, 03:22 PM
  #15  
Senior Member
 
Join Date: Feb 2011
Location: Minnesota
Posts: 4,673

Bikes: N+1=5

Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 875 Post(s)
Liked 245 Times in 181 Posts
Originally Posted by Seattle Forrest
Any modern Garmin (that runs even the first version of ConnectIQ) can. You just wouldn't want to.
Exactly. If you want a power measurement with *any* accuracy, you pretty much need a power meter.

J.
JohnJ80 is offline  
Old 04-26-17, 03:27 PM
  #16  
Senior Member
 
Seattle Forrest's Avatar
 
Join Date: Mar 2010
Location: Seattle, WA
Posts: 23,208
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 18883 Post(s)
Liked 10,646 Times in 6,054 Posts
Originally Posted by Velocedge
Yeah, I understand all the variables... just looking for something in the ballpark. If you're attached to the phone/GPS then you have a little more information. Zwift does it (albeit without wind). I'm looking for something like that. Not perfect, just a benchmark figure that works in the long term scheme of things.
Serious question: how close is good enough, how far off is too much?
Seattle Forrest is offline  
Old 04-26-17, 04:16 PM
  #17  
Senior Member
 
Join Date: Apr 2017
Posts: 363
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 148 Post(s)
Likes: 0
Liked 6 Times in 4 Posts
My Sigma ROX 8.1 shows power in max power(watts) and avg power(watts) I don't know how accurate it is as I never have used a power meter but it is fairly consistent in the stats it spits out. That model has been discontinued and now is model 23.16 The kicker is it's only sold in Europe right now. I called Sigma they give me that info and I ordered it from the Netherlands from a Sport Shop I found online. Total with the DHL shipping was around $150 It came in 5 days. I am putting it on my new Raleigh TAMLAND 1 and it's at the LBS right now and won't get it back until around next Friday so I can't give you a review on the new 23.16 computer.

Zman
Zurichman2 is offline  
Old 04-26-17, 05:52 PM
  #18  
Se˝ior Member
 
ItsJustMe's Avatar
 
Join Date: Sep 2005
Location: Michigan
Posts: 13,749

Bikes: Windsor Fens, Giant Seek 0 (2014, Alfine 8 + discs)

Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 446 Post(s)
Likes: 0
Liked 8 Times in 7 Posts
Not possible. How would it know if you were going up hill or down?

It could guess and assume a flat route but it'd be just a guess.
__________________
Work: the 8 hours that separates bike rides.
ItsJustMe is offline  
Old 04-27-17, 04:56 AM
  #19  
Junior Member
Thread Starter
 
Join Date: Oct 2016
Posts: 5

Bikes: Dean Ti Custom

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Likes: 0
Liked 0 Times in 0 Posts
Zurichman2, that Sigma ROX 10 and 11 looks like it does calculate power. Has altitude capabilities to help tell you whether you're going up or down too. Thanks for the tip!
Velocedge is offline  
Old 04-27-17, 07:12 AM
  #20  
Senior Member
 
caloso's Avatar
 
Join Date: Jul 2002
Location: Sacramento, California, USA
Posts: 40,865

Bikes: Specialized Tarmac, Canyon Exceed, Specialized Transition, Ellsworth Roots, Ridley Excalibur

Mentioned: 68 Post(s)
Tagged: 0 Thread(s)
Quoted: 2953 Post(s)
Liked 3,106 Times in 1,417 Posts
Does it have a weather app so it knows whether you have a tailwind or proximity radar so it knows whether you are drafting?
caloso is offline  
Old 04-27-17, 02:24 PM
  #21  
Senior Member
 
Seattle Forrest's Avatar
 
Join Date: Mar 2010
Location: Seattle, WA
Posts: 23,208
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 18883 Post(s)
Liked 10,646 Times in 6,054 Posts
Originally Posted by Seattle Forrest
Serious question: how close is good enough, how far off is too much?
This is an important question. You said it doesn't have to be exact, but how inexact is good enough?

You can take a permanent marker, write 160 w on a piece of tape, and stick it on your stem for a lot less than a new cyclocomputer. Total cost = $0.00 assuming you already have the parts.

If you have a Garmin or a smart phone, you can get free power guesses from Strava or Golden Cheetah.

Another option is you can spend maybe $100 or more on a computer that will give you a different wrong guess, and change it more often.
Seattle Forrest is offline  
Old 04-27-17, 02:42 PM
  #22  
Junior Member
Thread Starter
 
Join Date: Oct 2016
Posts: 5

Bikes: Dean Ti Custom

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Likes: 0
Liked 0 Times in 0 Posts
Well, could care less what the number is. As long as I have a number to compare from one ride to another (riding alone) then I have a benchmark. So it really doesn't matter if it's a "real" number or not. Just something to relate from one ride to another. Like Zwift, the watts or w/kg is just a number... doesn't mean anything real. But I have come to know if it's 3.4 or 4.2 w/kg what that means to my legs. Would like the same thing IRL.
Velocedge is offline  
Old 04-27-17, 03:05 PM
  #23  
Senior Member
 
Join Date: Feb 2011
Location: Minnesota
Posts: 4,673

Bikes: N+1=5

Mentioned: 21 Post(s)
Tagged: 0 Thread(s)
Quoted: 875 Post(s)
Liked 245 Times in 181 Posts
Originally Posted by Velocedge
Well, could care less what the number is. As long as I have a number to compare from one ride to another (riding alone) then I have a benchmark. So it really doesn't matter if it's a "real" number or not. Just something to relate from one ride to another. Like Zwift, the watts or w/kg is just a number... doesn't mean anything real. But I have come to know if it's 3.4 or 4.2 w/kg what that means to my legs. Would like the same thing IRL.
There are two issues here - overall accuracy and repeatability. Unfortunately, the way you want to do this gives you neither.

The problem is that it won't be repeatable unless you are on the same course in the same direction with the same wind and at the same body weight. All of those things can have large impacts on the power output and won't make for a repeatable number received from the computer. You just can't isolate a power measurement to just speed and cadence and get a decent, repeatable estimate.

I see significant variance using my power meter when I ride the same course due to wind, my aero position etc...

Power is the rate at which energy consumed. There are many factors which are not constant that effect that. If you want to know what that is then you need to measure it going in. There is not enough information from speed and cadence to estimate it and have it be repeatable. You either need lots of other rides over the same terrain to estimate and interpolate (i.e. Strava) or you need some other information (power, heart rate, wind information, etc...) to give you some basis for a reasonable estimate of work.

The suggestion to write a number on a piece of tape is not too far off from reality. There might be times when it's actually right, albeit random.

J.
JohnJ80 is offline  
Old 04-27-17, 03:51 PM
  #24  
Senior Member
 
Seattle Forrest's Avatar
 
Join Date: Mar 2010
Location: Seattle, WA
Posts: 23,208
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 18883 Post(s)
Liked 10,646 Times in 6,054 Posts
Originally Posted by Velocedge
Well, could care less what the number is. As long as I have a number to compare from one ride to another (riding alone) then I have a benchmark. So it really doesn't matter if it's a "real" number or not. Just something to relate from one ride to another. Like Zwift, the watts or w/kg is just a number... doesn't mean anything real. But I have come to know if it's 3.4 or 4.2 w/kg what that means to my legs. Would like the same thing IRL.
Write 160w on a piece of paper. Tape it to your stem. Make another one that says 140w for days when you're tired, and one that says 210w for really great days when you're killing it.

You owe me $180.
Seattle Forrest is offline  
Old 04-27-17, 03:54 PM
  #25  
Senior Member
 
Seattle Forrest's Avatar
 
Join Date: Mar 2010
Location: Seattle, WA
Posts: 23,208
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 18883 Post(s)
Liked 10,646 Times in 6,054 Posts
Originally Posted by JohnJ80
The suggestion to write a number on a piece of tape is not too far off from reality. There might be times when it's actually right, albeit random.
A broken clock is right twice a day.

The thing is, a power estimate might be right 90 % of the time, or it might be right 10 %, you'd never know either way if you're not measuring it. Also, an estimate could be wrong by a little, or it could be wrong by a lot, and again, you just don't know.

People are different, and we're after different things. For me, if I can't trust that the number is useful, then it isn't.
Seattle Forrest is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service -

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.