For long distance endurance racing there are two schools of thought when it comes to powering your GPS/lights/etc. I’m a firm proponent of hub dynamos which I’ve used with success for racing the Tour Divide and the Transcontinental races. There are two main suppliers of hub dynamos: Son, which I run on my Tour Divide bike, and Shutter Precision (SP) which I use on my Transcontinental bike. A new company—Velogical—has released a new dynamo which rather than being in a hub is a rim dynamo. This has two main attractions over hub dynamos:
- You don’t need to build a new wheel to use the dynamo; and,
- The dynamo only needs to be activated when you need power so there is no drag the rest of the time (hub dynamos typically have drag of about 2-5 Watts)
I was very intrigued with the Velogical system so I purchased one. This is what it looks like mounted on my bike with the dynamo engaged.
Being an engineer and having three dynamos, I decided to test the Velogical against my Son and SP dynamos to see the relative drag of the different dynamos. While at it, why not test a few wheel/tyre combinations as well? Here’s what I found …
The Test Method
Originally I intended to use the ‘Aerolab’ function from the Golden Cheetah software. Although designed to investigate the aerodynamic drag, it can also be used to estimate the rolling resistance by cycling a circular route with a power meter. However, in spite of my best efforts I was unable to get reliable and repeatable results so I sent to ‘Plan B’: a coast down test.
I started on a small rise and let the bicycle roll down the hill, marking the spot where the bicycle stopped rolling. By holding the same line, and keeping the same position on the bicycle, any changes in the distance travelled would be due to the hub drag, tyres or wind. So by using the same tyres and windless conditions you would have just the drag effects. These are only a few watts—as shown from the controlled tests on a range of hubs shown below which is from http://www.ctc.org.uk/file/public/feature-hub-dynamos.pdf.
The Test Site
Fortunately I live in rural New Zealand so didn’t have to travel far to find a suitable site. Just turned left out of my drive way.
The photo to the left shows the road section. I started at the top of the photo and it was 250 metres to the driveway on the right (which is the entrance to the Playhouse cafe).
I marked off the 250 metre spot and then put down a tape measure. For each run I’d try and hold the same line and keep the bike moving forward until it literally would be about to fall over. That spot was marked with chalk on the road and the run number.
The Test Rig And Wheel/Tyre Combinations
I used my Transcontinental Race (TCR) bike. This has a Luxos U dynamo light which let me do runs with and without the dynamo on to see the relative effect of using power.
I had the following combinations to test:
- SP8 Disc with Hutchison Sector 28 mm tubeless tyre
- SP8 Non-Disc with 25 mm Continental 4 Seasons tyre and Slime tube
- SON Disc on Stan’s ZTR rim with 25 mm Continental 4 Seasons tyre and Slime tube
- Easton non-dynamo with 25 mm Continental 4 Seasons tyre and Slime tube
- Shimano C50 non-dynamo with 23 mm Specialized tyre and latex tube
The SON and Stan’s ZTR were off my Tour Divide bike so I swapped my usual mountain bike tyre for the Continental. The Shimano C50 was off my time trial bike and was expected to be the fasted.
To ensure that there was no bias, each wheel and tyre combination were weighed. As would be expected, the mountain bike SON/ZTR were the heaviest so that combination required no weight correction. The other dynamos wheels were about 127 grams lighter, and the non-dynamos about 424 grams lighter. I had two water bottles which weighed 127 and 424 grams respectively. The coast downs were done with the appropriate bottle—or none for the SON.
All tyres were set to 80 psi, except the Shimano which were 100 psi.
The Test Conditions
To ensure still conditions, I did the runs early one morning—starting at 05:00. I got 5 sets of runs on Day 1 and 4 on Day 2. I had attempted another set of runs but the wind was so variable the results were meaningless (after Day 1 I got a portable anemometer from Ebay to check wind speed).
I was unable to get all the runs done in one day due to the wind coming up. To normalize the data I used runs of the Easton without the hub dynamo on both days. Theoretically, the only difference would be due to the mass density of air (RHO) which is a function of air pressure, temperature (5 C and 0 C on Days 1 and 2 respectively), and dew point. I used the Android App ‘DS Barometer’ for the air pressure (1022 hPa on Day 1; 1018 hPa on Day 2), the relative humidity from the Mapua weather station (which is at sea level and I could see Mapua from the test site) (92% and 89% on Day 1 and Day 2 respectively) to calculate the dew point from www.dpcalc.org/ (4 and -1). Golden Cheetah has a RHO estimated under Tools and this gave a value of 1.27615 kg/m3 on Day 1 and 1.29551 kg/m3 on Day 2.
This means that the mass density of air was 1.5% heavier Day 2 than Day 1, and this for the same position on the bike the aerodynamic drag would be 1.5% higher. This does not mean that during the coast down the bicycle will travel 1.5% less distance since the contribution of the aerodynamic drag decreases with the cube of velocity. As your speed decreases, the rolling resistance becomes the dominant force impeding motion. The figures below (from here) show the relative contributions of the different components.
Good sites for learning about bicycle aerodynamics are here and here. If you are super keen you can also read the book I wrote in 2003 on predicting vehicle operating costs which has a detailed discussion on mechanistic modelling of forces opposing motion. But the web sites are much better .
The Test Results
The table below shows the summary data for the 50 successful coast down runs over the two days. A number of runs were rejected because of issues like vehicle interference or clearly taking a different line. The maximum speed on the coast down runs was typically about 23 km/h.
With the exception of the Velogical, a set of runs were made with a specific wheel/tyre combination, then the wheel was swapped out. For the Velogical, a run was made with the dynamo off the rim, then the next run was done with the dynamo on the rim. By running in pairs like this an changes in ambient conditions were factored out.
The means were statistically valid with 90% confidence; and all but the SON and Shimano-Veological valid with 95% confidence. The table below shows the comparisons of the mean distances from the various pairs of runs.
What do these tell us:
- With and Without Lights On: There was no significant difference.
- With and Without Hub Dynamo:
- On Day 1 there was no significant difference when using the same tyres–i.e. for the tests of the Velogical and the SON which, along with the non-dynamo wheel all used the 28 mm Continental 4 Season tyres.
- There was a significant difference between the non-dynamo wheel with the Continental 4 Season against the SP dynamo wheel with the Hutchison Sector tubeless. This was counter intuitive since the mass of the slime tyre and tube was greater than the mass of the tubeless fluid which means that the inertial resistance should have been greater for the non-dynamo wheel. However, because I did not know the mass of the rims that could be part of the factor—but not likely since the SON wheel definitely had the heaviest rim. Unfortunately I ran out of time to test my SP with the Continental 4 Season tyres … a future test.
- With and Without The Velogical Rim Dynamo:
- On Day 1 there was no significant difference with and without the rim dynamo.
- On Day 2—even allowing for the 2.5% difference in conditions reflected by the additional resistance for the control—there was a significant difference for the Velogical. Why? I had not installed the dynamo correctly! With all my testing, the dynamo slipped and so it was no longer perpendicular to the rim. When you mount the Velogical you need to ensure that it is running perpendicular to the rim or there will be additional forces. The mounting clamp uses a 5.5 mm nut and while I had a 5 and 6 mm socket, I could not locate a 5.5 mm socket so had to use pliers for tightening the nut (I still cringe when writing this as my late father would clip my ear if he knew I did this). At some point during all the tests the dynamo appears to have shifted and that affected the resistance.
- My Aero Wheels With Latex Tubes Did Poorly: All the tests available show that there are reductions in rolling resistance with latex tubes. So how can a super fast set of aero wheels do so poorly—i.e. no additional distance over a regular wheel with a Slime tube? Tyre size. All the other tests were done with 28 mm wide tyres and my aero wheels run 23 mm tyres. As discussed here, the narrower the tyre, the higher the rolling resistance. Since rolling resistance was the dominant force for most of the coast down, the latex tubes were not sufficient to offset this.
Firstly, this was a lot of fun. Yes, 50+ coast down tests, the time setting up the wheels, the aborted runs, etc. took a lot of time, but for a geeky engineer what better way to spend your time at 05:00 on freezing mornings
My broad conclusions are this:
- There is no significant difference with the SON and SP hubs in the coast down test. This is consistent with the lab testing you can find here.
- The Velogical when set up correctly does not have any higher drag than the SON or SP hub. The manufacturer gives very clear instructions and you need to be careful to get it perfectly aligned. If you don’t do this properly, and it slips you will probably have much higher drag. Follow their instructions carefully and you should be fine.
With regard to the Velogical, a few other observations:
- They have fantastic customer support. The first system they sent got lost in the mail and it was replaced with no hassles. Provided good advice on setting it up etc. They do think TCR racers are a bit crazy. Probably right
- Since the unit is only on when you need it, you will save probably 2 watts over having a hub dynamo for most of your riding.
- The greatest advantage is your ability to fit it to your bike without having to get a wheel rebuild. This saves a lot of money as it costs about the same as the SON and SP hubs.
- You can also move it between bikes, but if you want to do that it would be best to get multiple mounts and just swap over the dynamo head unit.
- The system will not work with all types of lights. It works great with the Luxos U but you will need to check with Velogical that it will work with the light you have. It does work with the Pedalpower + Universal Cable.
- It has less power output than the SON and SP so it may take longer to recharge your GPS units.
- The only disadvantage is the noise. There is a ‘whirring’ sound as it spins against the wheel.
I’ll leave the Velogical on my TCR bike as a backup unit just in case my SP fails—most likely if I crash the bike! It’s only 95 grams so won’t notice the weight and if I need it I’ll be glad I had it. Afterwards I’ll move it to one of my other bikes. It’s a nice unit.