Endless Sphere

I just noticed your route map. You gotta stop by for a visit when going thru Costa Rica for supplies, repairs, or just plain fun. Looking at the schedule my multi-hybrid electric motor powered sport fishing catamaran should be in the water with all the bugs worked out, so that's guaranteed time well spent unless you're somehow allergic to what you catch.

Never worry about wrath on the forum. 10 years here and I'm certain that anyone with anything useful to share has nothing but best wishes for you on such an epic journey and wish they could do it too.

John in CR wrote: ↑

Jun 16 2018 4:09pm

I just noticed your route map. You gotta stop by for a visit when going thru Costa Rica for supplies, repairs, or just plain fun.

That sounds amazing. Thank you. I guess it's a good thing I'm coming the long way around!

John in CR wrote: ↑

Jun 16 2018 4:09pm

Never worry about wrath on the forum. 10 years here and I'm certain that anyone with anything useful to share has nothing but best wishes for you on such an epic journey and wish they could do it too.

I guess I did kind of imply that you guys are given to wrath. What I meant to say was that most of what I know about ebikes I learned from all of you and this trip wouldn't have been possible without you all.

solarEbike wrote: ↑

Jun 16 2018 12:32pm

Nice. Haven't seen that one before. My Nomex (aramid) honeycomb core panel in the video above came in at 1.2 kg/m² with 5.5 oz/yd² (186 g/m²) cloth on both sides and my second test was 1.0 kg/m² with a lighter cloth on one side. Given the Sunpower Gen III cell temperature coefficient of power is -0.29%/ºC, I would need to see a significant cell temperature drop to offset the extra weight. It's all about the Wh per kg for me.

Asuming 2m² of solar panels the extra weight of an aluminium backside is 2.5kg. In the falt you need proably an extra 1 (one!) Watt to compensate for the extra rolling resistance. the slightest difference in air resistance would be much more significant.

2m² of solar panels could produce 320W under optimal conditions. An extra 10K would give you an extra 2,9%, which is around 9 Watt. And I assume the difference could be as high as 20-30K if you have air flo under your panel.
Of course you don't usually have 1000W/m² of solar irradiation, so the average gain is lower.

The much more significant advanatge that I assume(!) is the thermal conductivity between cells.

Picture shows a modul using (half) sunpower cells on front of 10mm polycarbonate "double skin sheet" with an thermal conductivity of 3W/m*K (used for greenhouses to keep the heat in). I wanted to use it because it is lightweight and quite rigid.

As you can see even under a moderatly sunny and warm day in German some cells will get significantly hotter than others and if you short cut the modul (worst case) some cells get dangerously hot. A modul with PE front side would have been destroyed in this test. It would be significantly worse in really hot climate.

Anotther factor is thermal expansion of differnet materials. Cells are from silicone, frame usually is aluminium. PC as a back side would have been a bad idea from that perspective, too. I don't know about thermal expansion of your Nomex backside.

Cephalotus wrote: ↑

Jun 17 2018 1:05pm

Asuming 2m² of solar panels the extra weight of an aluminium backside is 2.5kg. In the falt you need proably an extra 1 (one!) Watt to compensate for the extra rolling resistance. the slightest difference in air resistance would be much more significant.

This is great stuff. Thanks for sharing.

Designing for optimal solar conditions and flat terrain is fine if your goal is multi-day trips on a route with no clouds and no elevation changes. I've made similar weight trade-offs myself such as when I doubled the size of the battery on my commuting bike. The bike was much heavier but it had twice the range so it didn't matter that it was heavy. I never need to carry the bike up a flight of stairs or lift it over my head to put it on the roof of a car.

For year-round touring, the priorities are different. I was reminded of this on my most recent 2 day camping test ride. The weather was much less sunny than the forecast had indicated and there was a very steep 3000 foot (1000 m) climb near the end of day 1 and again at the start of day 2. My battery was empty, clouds had rolled in and my gearing was insufficient for pedaling up the hill. I ended up having to get off and push for nearly three hours. It was brutal.

Given the choice between carrying a 2 kg heatsink and having a bike that's 2 kg lighter, I know what I would choose in that situation.

If you get a chance to measure the steady-state cell temperature of a PV module with a thermally conductive support panel and a module with a thermally non-conductive panel, please post the results. Ideally, both scenarios would be tested while stopped and while moving to be helpful to your average solar ebike designer.

My guess is that the benefit of the heat sink will be lower when moving due to the cooling effect from airflow on the front side being the same in both cases. But I love it when I'm wrong. I always learn more that way.

Cephalotus wrote: ↑

Jun 17 2018 1:05pm

Anotther factor is thermal expansion of differnet materials. Cells are from silicone, frame usually is aluminium. PC as a back side would have been a bad idea from that perspective, too. I don't know about thermal expansion of your Nomex backside.

The Nomex core material in my panel will flex with the top and bottom skin surfaces and those are made of carbon fiber and epoxy which has nearly zero thermal expansion. The Sunpower cells themselves are interesting in this regard. Thanks to the metal backing, they act like bimetalic strips and curve into a cup shape in the hot sun when the metal backing expands at a greater rate than the silicon cell material. Anyone designing their own solar panel needs to plan for this in their design. For ultralight encapsulation, it helps to bond the entire module surface to the support panel.

I swear this hill was crazy steep. How come it never looks that way on video?

solarEbike wrote: ↑

Jun 17 2018 5:36pm

Given the choice between carrying a 2 kg heatsink and having a bike that's 2 kg lighter, I know what I would choose in that situation.

Oh, I'm a fan of reducing weight, too.

This is my street legal speed pedelec which goes 45km/h (53km/h without v-max reduction) and weights just 19.5kg with fenders, pannier, Supernova M99 light, license plate, 550Wh battery and so on (18kg is possible)...

If you get a chance to measure the steady-state cell temperature of a PV module with a thermally conductive support panel and a module with a thermally non-conductive panel, please post the results. Ideally, both scenarios would be tested while stopped and while moving to be helpful to your average solar ebike designer.

I will do such tests under "stopped" conditions with a cheapish 18W PE test module (also Sunpower cells). Maybe I can add a fan to simulate some air movement?

In my opinion the concept of a solar bike is useful if you want to drive significantly faster compared to a bike with no motor at all. If you ride fast, let's say 30km/h+ air resistance is much more important than weight and most bikes and especially many solar bikes have VERY poor air resistance.

Not my site, but interesting to read:

https://www.cyclingabout.com/how-much-d ... -you-down/

https://www.cyclingabout.com/speed-diff ... g-results/

https://www.cyclingabout.com/touring-ty ... e-testing/

https://janheine.wordpress.com/2012/05/ ... -bicycles/

Cephalotus wrote: ↑

Jun 17 2018 6:13pm

This is my street legal speed pedelec

That's a sweet ride. My sincerest condolences on your loss. Rest in peace, Bionx.

Cephalotus wrote: ↑

Jun 17 2018 6:13pm

I will do such tests under "stopped" conditions with a cheapish 18W PE test module (also Sunpower cells). Maybe I can add a fan to simulate some air movement?

Or maybe wait for a very windy day? You would get a good laminar airflow over both the top and bottom surface. Cheaper than a wind tunnel.

Cephalotus wrote: ↑

Jun 17 2018 6:13pm

Not my site, but interesting to read:

That last one was new to me. I've come across Alee's articles before but I think this is a good time to re-read them as I'm contemplating trying out some aero improvements along the lines of a minimal tailbox fairing which would also protect my gear from sun, rain and prying fingers.

The hotspots in your infrared camera shots are consistent with my experience buying at least 5 different semi-flex panels from eBay, Amazon and AliExpress over the years. They were mostly junk. Solbian in Italy makes a high quality flex panel with genuine SunPower cells. Here's an article they wrote on testing their product against competitors. Normally, I would be wary of a company publishing in-house test results with such small sample sizes and no third party validation but in this case they're consistent with my own experience. https://www.solbian.eu/en/blog/testing- ... rises-n195

energyi wrote: ↑

Jun 15 2018 3:03pm

would recommend anyone looking into doing a similar project to also consider a bifacial module, can generate electricity from both sides.

I assume you meant bifacial modules made with glass like this one? At over 10 kg/m², they don't fit my needs. However, I want to thank you for getting me thinking about an idea I had abandoned a while back: extra cells which face down while riding and fold out for charging while stopped. These would have their own MPPT charge controller.

I've been making carbon fiber/honeycomb sandwich panels lately to test the feasibility of using these for my trailer. The panels are crazy light and stiff but the edges are very fragile. I'm planning to bond 12mm carbon fiber tubes around the perimeter to add side impact protection, rigidity and to connect multiple panels together. These tubes could be modified to create a continuous hinge ("piano hinge") as shown in the video below. If they can be locked in place while open, I could possibly even ride with them deployed when I'm on roads with no traffic and not much wind.

solarEbike wrote: ↑

Jun 17 2018 7:55pm

That's a sweet ride. My sincerest condolences on your loss. Rest in peace, Bionx.

No big deal. I have enough BionX hardware to last for the next 15-20 years and also the sweet BionX Development software

It's quite complicated to build a street legal speed-Pedelec over here, there are not so many options.

I'm all for saving weight in my everyday Pedelecs and bicycles, because low weight make them nice and easy to handle and carry. Often I do carry them the steps 4 stores up to my flat, sometimes wearing a suit.

with a solar powered bike low weight is beneficial of course, but handling is shitty anyway. I personally are planing more into a foldable solar generator (my moduls are 55cm x 60cm each) and I was thinking abot adding shoulder straps and a load carry system on the trailer, so it could be possible to carry the solar trailer incl. luggage for small distances on your back.

There could also be the possibility to add a Packraft to the system. Mines weights around 4kg and can carry 250kg

So with all the extra stuff to carry around (battery and motor around 8lg, solar panels + charger around 10-15kg, trailer around 8kg) there needs to be a significant benefit over riding just a normal bicycle.

I assume travaling at 30km/h air resistance becomes a much more important consideration compared to travaling around 15km/h average.

When looking at the suntrip vehicles I wonder to see so little optimizing of air resistance. Just look how some people mount their luggage. You can just see the wasted watts in drag.

The hotspots in your infrared camera shots are consistent with my experience buying at least 5 different semi-flex panels from eBay, Amazon and AliExpress over the years. They were mostly junk. Solbian in Italy makes a high quality flex panel with genuine SunPower cells. Here's an article they wrote on testing their product against competitors. Normally, I would be wary of a company publishing in-house test results with such small sample sizes and no third party validation but in this case they're consistent with my own experience. https://www.solbian.eu/en/blog/testing- ... rises-n195

Thnaks for the link. I already did know the site because it is one of very rare ones with IR pictures of flexibe panels showing the microcracks. I'm super interested in that topic, but sadly nobody at the Suntrip seems to do some IR thermal or luminescens pictures of the modules before and after the trip.
I bet you could learn a LOT with such pictures and it is not expensive to make them.

The Solbian moduls are quite expensive. Afair I paied 180 Euro for one 50W panels (payed for 4 of them and got 2 refunds because of cracked cells). I bought modules with ETFE front cover (no reflections) and black fibreglass backside. I thought this is a better idea for stealth camping not having the withe moduls.
But maybe using a white aluminum backside would have been the better idea, not that a plan those aluminium backside which imho is much better than my first idea with 10mm Polycarbonate.

Using thermal insulation on the backside of your modules will increase Delta T by almost factor 2. So if the standard module will get up to 70°C with 30°C air temperature, with a thermal insulation on the back you would almost hit 110°C under such conditions. And this is average temperature and not hot spots. Imho this is to hot and will degrade the cells and materials used quickly.

This is way I believe aluminium is a better idea.

PS: Your sun tracking solar bicycle trailer is VERY cool! If you tilt the generator how does it handle side winds?

Quick test today to get an idea of how much energy I can expect from additional cells mounted on my trailer so they face the ground while riding but can be deployed to face the sun when stopped as shown in the video above. I was hoping for around 10% of full sun but It's even less than that. I think the downward facing cells are a dead end for me.

However, I'm really liking the piano hinge and plan to incorporate it into my main panel so that it can fold up into four sections for transport and to protect the cells when needed or maybe even fold up the panel during extreme wind conditions.

Same 60 SunPower cells under the same temperature and irradiance conditions.

Cephalotus wrote: ↑

Jun 18 2018 2:22pm

with a solar powered bike low weight is beneficial of course, but handling is shitty anyway

I don't know about any other solar bikes but my trailer handles like a dream. I can tuck into corners at high speed. With a fully loaded bike, I can barely tell it's even there except for some rattling sounds I can't seem to eliminate. I've had it up to 36 mph (58 kph) on a steep, narrow, twisty downhill with traffic right behind me and I felt I could have easily gone faster. I have speed limiting enabled so my proportional regen was kicking in and keeping me from going faster.

Cephalotus wrote: ↑

Jun 18 2018 2:22pm

I assume travaling at 30km/h air resistance becomes a much more important consideration compared to travaling around 15km/h average.

That's what I'm thinking, also. My target cruising speed for solar ebike touring is 20-25 mph (32-40 kph) so the right aero improvements should have measurable results, I would think.

Cephalotus wrote: ↑

Jun 18 2018 2:22pm

When looking at the suntrip vehicles I wonder to see so little optimizing of air resistance. Just look how some people mount their luggage. You can just see the wasted watts in drag.

That was my impression, too. I'm seeing a lot of what looks remarkably like drag parachutes and air brakes.

Cephalotus wrote: ↑

Jun 18 2018 2:22pm

I bet you could learn a LOT with such pictures and it is not expensive to make them.

What kind of equipment are you using? I see that FLIR now makes one for Android and iOS for US$ 200.

Cephalotus wrote: ↑

Jun 18 2018 2:22pm

PS: Your sun tracking solar bicycle trailer is VERY cool! If you tilt the generator how does it handle side winds?

Thanks. I've done 500 miles (800 km) of road testing, much of it along open water of San Francisco Bay. Most the time, I can ride with it fully tilted and it makes no difference in handling. Cross winds were an issue only once and I just flipped a switch and the panel parked in a horizontal position as shown below. In this position, even strong side wind was not an issue thanks to the low profile and height (about 24" or 60 cm) combined with the mounting point on the bike being the rear axle.

solarEbike wrote: ↑

Jun 18 2018 10:54pm

I don't know about any other solar bikes but my trailer handles like a dream. I can tuck into corners at high speed. With a fully loaded bike, I can barely tell it's even there except for some rattling sounds I can't seem to eliminate. I've had it up to 36 mph (58 kph) on a steep, narrow, twisty downhill with traffic right behind me and I felt I could have easily gone faster. I have speed limiting enabled so my proportional regen was kicking in and keeping me from going faster.

Is your trailer custom made?

I have a Tout Terrain Mule and it also rides(!) excellent.



You barely notice that it is there. Problem is all the situations wher you do NOT ride the bike, lets say carrieing it over stairs. Here a Trailer (and I asusme any Trailer) really sucks and this is where lightweight bikes shine. I can very easily carry the bike shown above in one Hand and have one Hand left do to other things, like open doors or carry luggage or wahtever. Add a Trailer to that and everything become difficult. This is why I think that it doesn't matter so much if your bike + trailer + stuff weights 44kg or 48kg for example. It's not fun to carry anyway.

While riding the weight also doesn't matter except for going uphill.

That's what I'm thinking, also. My target cruising speed for solar ebike touring is 20-25 mph (32-40 kph) so the right aero improvements should have measurable results, I would think.

There was a posting some days ago about the Suntrip 2015 winner who averaged 174 miles per day, but it took him 14(?) hours riding time per day. Let's assume 12 hours real riding time. That's 15mph average and ha had a recumbent and a very large solar array and he wanted to wind a race, not make a nice trip.

But I also hope 30km/h+ is doable, because on a trip I would prefer riding for only 6-7 hours per day and having time to do and visit other things, too. So 6-7h * 30km/h = ca. 200km. This is more than I do ride on a trekking bike without motor.

At 30km/h I need around 4-7Wh/km (depends on air resistance and human power and wind), so for 200km/day this is an average of 1100Wh/day. More than my planed 150W array can produce per day
Maybe 150W + 150W stored would be an option for 100% solar, but this adds even more weight and you have to isnatll a second solar generator on every stop.

Another idea is to carry 1050Wh of batteries and charge around every 3rd day using grid power.

That's around 5hx150W (sunny days) x 3 days = 2250Wh

2250Wh solar + 1050Wh stored in battery = 3300Wh in 3 days.That's 1100Wh per day.

If it rains I'm not so motivated to ride 200km/day anyway.

That's my plan.

What kind of equipment are you using? I see that FLIR now makes one for Android and iOS for US$ 200.

https://therm-app.com/therm-app/

But a cheaper one will do.

For IR luminescence (to see the microcracks) I use and old Sony F717 + a small magnet which costs around 50 USD used.

Thanks. I've done 500 miles (800 km) of road testing, much of it along open water of San Francisco Bay. Most the time, I can ride with it fully tilted and it makes no difference in handling. Cross winds were an issue only once and I just flipped a switch and the panel parked in a horizontal position as shown below. In this position, even strong side wind was not an issue thanks to the low profile and height (about 24" or 60 cm) combined with the mounting point on the bike being the rear axle.

This Looks to be one of the best solar trailer designes I have seen so far. Did you build the trailer yourself or is it based around a commercial product?
Does it offer suspension? (difficult to see)

Cephalotus wrote: ↑

Jun 19 2018 4:14am

Did you build the trailer yourself or is it based around a commercial product?
Does it offer suspension? (difficult to see)

The trailer is my own design. Using BOB trailer hardware to attach to the bike otherwise scratch built. There's a lightweight elastomer suspension using a Cane Creek Thudbuster with about 1" (25 mm) of travel. I'll increase the travel a little for the next build.
The last part finally arrived today for my new and improved motorized panel tilt mechanism: an adjustable clutch to protect the drive assembly from overload in case of fall, crash, limit switch failure, etc. No idea if this will work but the idea is that it fits entirely inside the main boom tube (1.5" / 38 mm ID). Claimed torque of the motor/gearbox is 30 KG-cm (2.9 N-m) which is easily 2-3 times what's needed to move the panel thanks to the torsion spring to take the weight of the panel off the gearbox at 75° tilt. Add a shaft-mounted potentiometer, some photosensors, sprinkle some Arduino on it and you have automatic single-axis tracking inside a water and dust resistant housing. 1.55 lbs (703 g) for the parts shown here.

The clutch looks like a good idea. Seems like crashing or tipping over is only a matter of time so some protection against damage will pay off quickly.

Getting more out of the panels would be nice but it looks like you already have things pretty much optimized. When you are stopped, it might work to add some reflector panels on the sides of the panel. Something super light like aluminized mylar that can fold up and get stored under the panel when moving.

fechter wrote: ↑

Jun 21 2018 8:34am

When you are stopped, it might work to add some reflector panels on the sides of the panel. Something super light like aluminized mylar that can fold up and get stored under the panel when moving.

I picked up some reflective Mylar at Tap Plastics several years ago for testing out just this idea but never got around to trying it. I still want to try it just for fun but I don't have high hopes for using it on the road. If I actually overcame the problem of mounting the reflectors so they evenly illuminate the solar panels and don't blow away in a gentle breeze, I worry I might actually cook the panel. The panel isn't suitable for concentrated photovoltaics. Cell temperatures get up to 170F (75C) on a hot day in full sun without any reflectors.

One of the Sun Trippers tweeted about this just a couple of days ago. Says he saw a 10% increase using what appears to be an emergency foil blanket.
https://twitter.com/ArachneLabs/status/ ... 0722649088

Here's how the solar trailer performs under challenging conditions. Instantaneous readings of solar, motor and pedal power are displayed across the top of the screen. This is about 2-3 hours before sunset on a hilly, twisty, tree-shaded road. I picked this route and time of day not because they're representative of typical touring conditions but because watching a rear view camera going in a straight line with no shading would be boring.



Huge thanks to Justin, Teklektic and everyone who worked on the solar firmware for the CA over at Grin. I've logged over 2000 miles (3000 km) with various solar setups over the years but I've always used two CA's and mostly had to make do with quick glances at the CA's while riding and jotting down totals at each stop. The combination of video and telemetry overlay is a game changer for solar bike design.

I hope this helps others who are looking to build their own solar rig.

I've uploaded the log data to the Grin Trip Analyzer here: http://www.ebikes.ca/tools/trip-analyze ... rip=vQbjaL

Note: The Wh totals are a bit out of step with each other. The solar and motor Wh are from the start of the trip while the human Wh are from the start of the video (3 km into the trip). This is because the log contains Ah for solar and the motor controller but only instantaneous human watt readings. I power cycled the Analogger and CA at the start of the video to make it easier to sync the data with the video.

Tools used: CA3 with solar firmware, modified GPS Analogger with Hall sensor for solar charge current, Sempu Gen3 bottom bracket (uncalibrated), GoPro Hero 5 Session for video capture, DashWare for telemetry data screen overlay (running on a 12" MacBook under VMWare), Final Cut Pro to add the yellow markup and to edit the footage down to 5 minutes (from about 25).

I wasn't sure I was ready to speak on camera about my project but when Justin shows up with a camera and says "Hey, wanna make a video?" you say "Yes." Shot at Maker Faire Bay Area in May 2018.

Very nice video. You've got me seriously thinking about a solar trailer for some long distance touring.

A little progress update on my attempt to create a practical aero tail box fairing for long distance touring. Here's my sketch:


The idea is that all of my gear fits behind me and the cover keeps out the sun, rain and sticky fingers. And, if I'm lucky, extends my range on long days by a small but measurable amount.

The primary inspirations for this approach are the Kamm tail designs of Wunibald Kamm in the 1930's, John Tetz's excellent work on his foam velomobile and the aluminum Alleweder velomobile. That last one was of particular interest because it has almost no compound curves so when the sheet is removed it unrolls to a flat shape. That's where the practical part comes in as I need everything to pack down to a small size for taking commercial flights across those inconvenient oceans.

Over the past week or so, I built a cardboard prototype to help me think through some of the design challenges and to do some road testing. It's a little rough around the edges but the idea was to do a quick iteration and generate more ideas. The next build will conform more closely around the shoulders and the lower front half will conform to the Ortlieb recumbent panniers which I don't have yet.

Here's how it attaches to the bike. The tube plays triple duty as support for the fairing, hitch for the trailer and attachment point for light but bulky items like the sleeping bag so they don't stick out on the side past my arms and shoulders.

Here's the foam material I plan to use for the final build. It is cross-linked polyethylene foam (XLPE), a closed-cell rigid foam with decent mechanical strength and UV resistance. I'm using a sheet that is 1/4" (6mm) thick, 4 lbs/ft^3 (64 kg/m^3) density. The piece shown below weighs 2.95 lbs (1.340 kg) and when cut down to size should be about half of that. As you can see, it does an excellent job of holding its shape with minimal supports.

Last night, I finally attempted to quantify the aerodynamic gains and the results are inconclusive. I don't have access to a wind tunnel or even an indoor track (anyone in the SF Bay Area in a position to help with that? ) so I did a series of coast down tests from 25 mph (40 kph) and power usage tests running at 22 mph (35 kph) on motor only, using a CA3 on closed-loop speed throttle (PID tuning hell) with an analogger. My venue was mostly flat but not entirely free of wind so I think that ruined the results. I repeated the coast-down test 4 times with and without the box and ran the same stretch of road in both directions at constant speed.

The results? Constant speed test averaged 326 watts with the box and 347 watts without which appears to be a 6% reduction in power but on the coast down test I traveled on average 5% farther without the box. I plan to repeat the test with the foam version in the near future.

Finally, here's a gratuitous sunset shot because, you know, Instagram...

I see you're using CAD (cardboard aided design). One of my favorites too.
After seeing Justin's trike flipped over by the wind, I guess the material and design should be able to withstand this. A cross wind might do funky things to the steering. You might be able to test for this with the cardboard version. I know from experience when there's a strong cross wind and a big truck goes by that interrupts the wind, it tries to suck you into the traffic lane.

The idea has been around for a long time but you don't see many in use.

The egg fairing:

fechter wrote: ↑

Jul 18 2018 2:12pm

After seeing Justin's trike flipped over by the wind, I guess the material and design should be able to withstand this. A cross wind might do funky things to the steering. You might be able to test for this with the cardboard version. I know from experience when there's a strong cross wind and a big truck goes by that interrupts the wind, it tries to suck you into the traffic lane.

I've been riding it around for a bit, including some gusty places like around the Port of Oakland and it seems to do ok in moderate cross wind. The whole thing pops off easily and rolls up so I could stop and roll it up under heavy wind conditions. If I decide later that it's not working out, I can always get rid of it while on tour since it's not structural and my setup will work without it.

When stopped, I can park the trailer at a 90 degree angle to the bike which makes it much harder to tip over. For overnight stops in windy locations I may even stake it down with some guy lines or take the foam into the tent with me for extra padding and insulation from the ground.

Love the egg fairing. I have a whole folder of aero inspirations but haven't seen that one before.

Looks like a good surface for some flexible amorphous solar cell ... Just thinkin' out loud ...

wturber wrote: ↑

Jul 18 2018 3:04pm

Looks like a good surface for some flexible amorphous solar cell ... Just thinkin' out loud ...

I hear you. And I tried to talk myself into it. I really did. Here's one concept I dreamed up and one I actually built a few years back.


Ultimately, I found that the thin-film stuff doesn't come close to top-end monocrystaline silicon on a Wh per kg basis, particularly when you consider that you need twice the surface area and twice the structural support hardware.

Also, curved solar panels have a much bigger power output penalty than you might imagine by just looking at them. All the cells seem to be getting about the same amount of sunlight, right?


Well, not really. Using my calibrated irradiance meter above, I find that even relatively small tilt angles off perpendicular to the sun result in 5-20% drops in irradiance. Consider that each cell or row of cells on a curved panel has a different irradiance level and different temperature so different Maximum Power Point voltage and current. All cells connected in series are limited to the current of the lowest performing cell in the string and parallel strings of cells on the same charge controller are held to the same voltage so none of them are at their MPP voltage. As a result, the total output is lower than the total combined irradiance would predict.

The curved panel above has a nominal 50W rating so it should be putting out around 40W on a warm, sunny day but I would typically get only 30-35W. That's a big penalty if I'm hauling a solar panel around on a bicycle. Absolutely flat is the way to go. I'll leave the curved experimenting to others without fear that I'm missing out.

The Wh per kg metric also makes it possible to evaluate the merits of design options like the motorized tracking system or attaching an aluminum heat sink to your cells. Both give you more Wh so you should definitely add them, right? Well, not necessarily. They also add significant weight so you need to consider both weight and energy production together.

For example, Raf is hauling two huge 1mm thick aluminum plates around with him (source) which, by my estimate, weigh 5.3kg, excluding the aluminum angle to keep the panel from flexing and other mounting hardware. By comparison, my new carbon fiber/Nomex honeycomb panels weigh 2.8 kg less for the same total cell area. That means I would have 275W more than him (66% more watts) if I match his panel weight and cell type. Let's say 40-50% more if we consider the additional mounting hardware those additional cells would require.

I have no doubt the heatsink is increasing his Wh compared to an identical panel with the same 120 cells mounted on a thermally insulating support structure. But is it 40-50% more over the duration of the trip? ...including all the riding hours with good airflow over the tops of the cells?

Wh per kg is not a metric you'll find on any PV module spec sheet because it depends on your mounting solution, riding strategy and weather at your location. It's not easy to measure but it's the one that matters.

solarEbike wrote: ↑

Jul 18 2018 2:00pm

Last night, I finally attempted to quantify the aerodynamic gains and the results are inconclusive. I don't have access to a wind tunnel or even an indoor track (anyone in the SF Bay Area in a position to help with that? ) so I did a series of coast down tests from 25 mph (40 kph) and power usage tests running at 22 mph (35 kph) on motor only, using a CA3 on closed-loop speed throttle (PID tuning hell) with an analogger. My venue was mostly flat but not entirely free of wind so I think that ruined the results. I repeated the coast-down test 4 times with and without the box and ran the same stretch of road in both directions at constant speed.

The results? Constant speed test averaged 326 watts with the box and 347 watts without which appears to be a 6% reduction in power but on the coast down test I traveled on average 5% farther without the box. I plan to repeat the test with the foam version in the near future.

Very intersting. At 35km/h 326W is about 9,3Wh/km and this with zero human power input.

Was this incl. your solar trailer or without?

Would be interesting how much drag the trailer adds.

Cephalotus wrote: ↑

Jul 18 2018 6:46pm

Very intersting. At 35km/h 326W is about 9,3Wh/km and this with zero human power input.

Was this incl. your solar trailer or without?

Would be interesting how much drag the trailer adds.

Honestly, I didn't even think about the trailer because I currently can't attach both the tail box and the trailer at the same time but it's a very good point. The trailer's presence in the slipstream could very well cancel out any gains from the tail box. I'll be sure to include the trailer next time and try it with and without the trailer if I have time and patience.

I've uploaded the data for the 35 kph run with the tail box here: http://www.ebikes.ca/tools/trip-analyze ... rip=O65ZAY
Without tail box here: http://www.ebikes.ca/tools/trip-analyze ... rip=PEND5z

I excluded all data points below 32 kph in my analysis to account for acceleration/deceleration at the start, end and mid-point turn-around. Those points are included in log I uploaded above so the average watts don't match what I reported.

You can also see the reduced power on the second half of the run when I'm going in the opposite direction and the slight headwind becomes a slight tailwind. The effect looks about the same with and without the tail box so I'm hoping it all cancels out.

You can also see there's still quite a bit of speed and power fluctuation which is not ideal for this kind of test. This was the best I could get after 2 hours of PID tuning. There's room for improvement for next time.

wturber wrote: ↑

Jul 10 2018 2:55am

Very nice video. You've got me seriously thinking about a solar trailer for some long distance touring.

I'm getting the solar long-haul bug too