Author Topic: Gps + magnetic speed sensor  (Read 2970 times)

quixoticgeek

  • Mostly Harmless
Gps + magnetic speed sensor
« on: 21 November, 2017, 10:50:01 pm »

Looking at cadence sensors for the Elemnt bolt, and notice they sell a bundle with a speed sensor, a cadence sensor and an heart rate sensor. Which is lovely and everything, but it made me think.

If you have a computer with gps, why would you add a wheel rotation sensor given the inaccurate nature of these sensors (what is the actual circumference of my wheel?).

What's the thought process here ?

J
--
Beer, bikes, and backpacking
http://b.42q.eu/

Kim

  • Timelord
    • Fediverse
Re: Gps + magnetic speed sensor
« Reply #1 on: 21 November, 2017, 10:57:09 pm »
The ultimate edge case is the indoor velodrome:  GPS can't get a signal, and even if it could, sampling error would make the measurement of distance travelled largely meaningless.

Lesser edge cases left as an exercise for the reader...

Pingu

  • Put away those fiery biscuits!
  • Mrs Pingu's domestique
    • the Igloo
Re: Gps + magnetic speed sensor
« Reply #2 on: 21 November, 2017, 11:01:30 pm »
It's not difficult to work out a wheel size which will give an accurate reading on the computer. My computer and GPS match each other pretty precisely.

I've not tried them in a velodrome, thobut.

fuaran

  • rothair gasta
Re: Gps + magnetic speed sensor
« Reply #3 on: 21 November, 2017, 11:11:08 pm »
Most of these newer speed sensors don't actually use magnets, instead its an accelerometer to measure the rotation of the hub.
They can automatically calibrate the wheel circumference using distance from GPS.

A speed sensor will be more accurate if you have a poor GPS signal.
Maybe also better for small changes in speed, ie quicker than the GPS can calculate. Or on a particularly twisty route, eg mountain biking, when the GPS can cut corners. Could also be better for stopped time, when the GPS position can wander a bit.

But for most regular cycling, it won't make much difference. I don't think there's not much point in a speed sensor for outdoor cycling (unless you ride through a lot of tunnels).

Aushiker

  • Cyclist, bushwalker, phottographer (amaturer)
    • Aushiker: Bicycling and Hiking in Western Australia
Re: Gps + magnetic speed sensor
« Reply #4 on: 22 November, 2017, 01:02:26 am »
But for most regular cycling, it won't make much difference. I don't think there's not much point in a speed sensor for outdoor cycling (unless you ride through a lot of tunnels).
Or do for example mountain biking with lots of tree cover or ride in a GPS signal free area [yep they exist in some odd places].

Also for indoor training they are handy. I have the Wahoo magnetless speed and cadence sensors on my bikes given they come in a bundle anyway.

PaulF

  • "World's Scariest Barman"
  • It's only impossible if you stop to think about it
Re: Gps + magnetic speed sensor
« Reply #5 on: 22 November, 2017, 06:08:15 am »
I use mine primarily on the turbo but it is also more responsive than a gps based speedometer. I know this is down to the sampling rate but using the Wahoo software and a RFLKT the lag is 15 seconds or so

Re: Gps + magnetic speed sensor
« Reply #6 on: 22 November, 2017, 07:16:54 am »
I tend to use footpld data for running now as it's usually more consistent in measuring run distance than gps where I run (trees etc). It may. E slightly more accurate, but the gps in current watches is usually pretty good. On the bike I just use gps outdoors or the turbo feed speed and cadence back to the watch in any case.

Re: Gps + magnetic speed sensor
« Reply #7 on: 22 November, 2017, 08:07:46 am »

what is the actual circumference of my wheel?


Step 1 - Lay a 3m tape out on the floor
Step 2 - Place bike with an identifiable wheel spoke pointing down, or stick a marker on the wheel, at 0 on the tape
Step 3 - Wheel bike until one revolution has been completed, read measure

Arguably deflection under weight may cause a slight reduction in effective circumference, doubtful that would be anything much more than a mm or two, or 0.1%


Re: Gps + magnetic speed sensor
« Reply #8 on: 22 November, 2017, 12:03:53 pm »
Step 4 repeat until you get a consistent reading
"Yes please" said Squirrel "biscuits are our favourite things."

quixoticgeek

  • Mostly Harmless
Re: Gps + magnetic speed sensor
« Reply #9 on: 22 November, 2017, 12:11:25 pm »

Step 1 - Lay a 3m tape out on the floor
Step 2 - Place bike with an identifiable wheel spoke pointing down, or stick a marker on the wheel, at 0 on the tape
Step 3 - Wheel bike until one revolution has been completed, read measure

Arguably deflection under weight may cause a slight reduction in effective circumference, doubtful that would be anything much more than a mm or two, or 0.1%

The accuracy is never as good as 0.1%. At least in my experience. Especially as tire pressure is never that consistent. Perhaps it's not helped by me sticking 100kg of kummerspek on the wheels...

I like the idea of the acceleratometer ones that auto calibrate with the gps...

I mostly asked to try and decide if the extra €129 for the bundle with cadence, speed and HRM is worth it over the base model... and then spending €29 for the cadence add on...

Will see how I feel come pay day...

J
--
Beer, bikes, and backpacking
http://b.42q.eu/

Samuel D

Re: Gps + magnetic speed sensor
« Reply #10 on: 22 November, 2017, 12:22:43 pm »
I don’t see the point in having cadence. What are you going to do with the knowledge that you’re pedalling at 88 RPM? (Which is anyway knowledge acquirable by counting while looking at a clock for 15 seconds.) It’s data for the sake of data. There’s quite enough of that in my life at the tail end of 2017.

But speed derived from wheel RPM is useful, because GPS receivers, while very accurate over a long time scale, are poor over a split second at the slow speed of a bicycle. Consequently, there’s a lot of data smoothing in the short term, leading to uncertainty about your instantaneous speed even in open terrain. For me, this becomes a minor problem when, for example, taking a pull at the front of a group and wishing to hold a steady pace.

There’s also the problem of the GPS not knowing when you’re stopped.

Unfortunately the accelerometer-based wheel-speed sensors are large, ugly, wireless devices that strap onto the hub (a beautiful part of the bicycle that I prefer to leave visible) and need their battery attended to.

quixoticgeek

  • Mostly Harmless
Re: Gps + magnetic speed sensor
« Reply #11 on: 22 November, 2017, 12:31:52 pm »
I don’t see the point in having cadence. What are you going to do with the knowledge that you’re pedalling at 88 RPM? (Which is anyway knowledge acquirable by counting while looking at a clock for 15 seconds.) It’s data for the sake of data. There’s quite enough of that in my life at the tail end of 2017.

Because I am a beginner and I find that having such info can help me with improving my pedal efficiency. I also take the approach that I'd rather have the data and not use it, than want the data and not have it. Also being the geek I am, managing datasets is kinda what I do for a living these days. Data is good, Data is useful, I like Data.

Quote

But speed derived from wheel RPM is useful, because GPS receivers, while very accurate over a long time scale, are poor over a split second at the slow speed of a bicycle. Consequently, there’s a lot of data smoothing in the short term, leading to uncertainty about your instantaneous speed even in open terrain. For me, this becomes a minor problem when, for example, taking a pull at the front of a group and wishing to hold a steady pace.

Just how short is your turn on the front?

Quote

There’s also the problem of the GPS not knowing when you’re stopped.

Unfortunately the accelerometer-based wheel-speed sensors are large, ugly, wireless devices that strap onto the hub (a beautiful part of the bicycle that I prefer to leave visible) and need their battery attended to.

Battery attention is the only downside I see. Yes the hub can be a work of art, and strapping something to it can detract from that, but ultimately while my bike is IMHO a thing of beauty, she is also a thing of utility. Having a sensor helps me ride her better, so the cost is worth it.

And anyway, I won't notice under the You Don't Want To Know™[1].

J

[1]Not entirely sure what it is or where it comes from, but it seems to have a near invisible coating on the streets of Amsterdam, yet the simple passage of a bike over it is enough for it to coat EVERYTHING with a thick layer that you can never quite get clean again. It's also the composition of the bottom of canals, under the 3m of water, there's 3m of bikes, then it's just You Don't Want To Know™ all the way down (until you get to the turtle...)
--
Beer, bikes, and backpacking
http://b.42q.eu/

Pingu

  • Put away those fiery biscuits!
  • Mrs Pingu's domestique
    • the Igloo
Re: Gps + magnetic speed sensor
« Reply #12 on: 22 November, 2017, 12:42:28 pm »

Step 1 - Lay a 3m tape out on the floor
Step 2 - Place bike with an identifiable wheel spoke pointing down, or stick a marker on the wheel, at 0 on the tape
Step 3 - Wheel bike until one revolution has been completed, read measure

Arguably deflection under weight may cause a slight reduction in effective circumference, doubtful that would be anything much more than a mm or two, or 0.1%

The accuracy is never as good as 0.1%. At least in my experience. Especially as tire pressure is never that consistent. Perhaps it's not helped by me sticking 100kg of kummerspek on the wheels...

I like the idea of the acceleratometer ones that auto calibrate with the gps...

I mostly asked to try and decide if the extra €129 for the bundle with cadence, speed and HRM is worth it over the base model... and then spending €29 for the cadence add on...

Will see how I feel come pay day...

J

Step 4: Enter wheel size measurement on computer
Step 5: Go for a ride of a known distance (say 10km or so)
Step 6: Compare computer distance with known distance
Step 7: Adjust wheel size on computer as necessary

Kim

  • Timelord
    • Fediverse
Re: Gps + magnetic speed sensor
« Reply #13 on: 22 November, 2017, 12:45:58 pm »
Unfortunately the accelerometer-based wheel-speed sensors are large, ugly, wireless devices that strap onto the hub (a beautiful part of the bicycle that I prefer to leave visible) and need their battery attended to.

They do mean there's nothing sticking out to get snagged on undergrowth though, which is likely to be an advantage off-road.  The cadence equivalent is a net win in terms of ugliness because you don't have to build up a stack of shims from the chainstay  /derailleur post to get within magnet range of the crank.

simonp

Re: Gps + magnetic speed sensor
« Reply #14 on: 22 November, 2017, 12:48:13 pm »
With tree cover or canyon effect, GPS speed can be all over the place. A wheel sensor will be far more consistent. I'm not sure GPS deals particularly well with acceleration, and a speed sensor can inform the GPS that the bike isn't wandering all around a car park while outside a control.

Kim

  • Timelord
    • Fediverse
Re: Gps + magnetic speed sensor
« Reply #15 on: 22 November, 2017, 12:50:25 pm »
Unfortunately the accelerometer-based wheel-speed sensors are large, ugly, wireless devices that strap onto the hub (a beautiful part of the bicycle that I prefer to leave visible) and need their battery attended to.

They do mean there's nothing sticking out to get snagged on undergrowth though, which is likely to be an advantage off-road.  The cadence equivalent is a net win in terms of ugliness because you don't have to build up a stack of shims from the chainstay / derailleur post to get within magnet range of the crank.


As for the usefulness of cadence displays, I find that it sometimes works to detect The Bonk, when early stages have impaired my sense of how I'm pedalling, but the bit of my brain that can read numbers off a screen still works.

Samuel D

Re: Gps + magnetic speed sensor
« Reply #16 on: 22 November, 2017, 01:02:48 pm »
Just how short is your turn on the front?

Oi!

Feanor

  • It's mostly downhill from here.
Re: Gps + magnetic speed sensor
« Reply #17 on: 24 November, 2017, 11:34:41 am »

what is the actual circumference of my wheel?


Step 1 - Lay a 3m tape out on the floor
Step 2 - Place bike with an identifiable wheel spoke pointing down, or stick a marker on the wheel, at 0 on the tape
Step 3 - Wheel bike until one revolution has been completed, read measure

Arguably deflection under weight may cause a slight reduction in effective circumference, doubtful that would be anything much more than a mm or two, or 0.1%

To improve accuracy, I'd advise marking the ground at the start, and doing 10 revolutions and then marking the ground again.
Measure the distance, and divide by 10.