Introduction
Following a successful initial experiment with two solar cargo bikes in 2020/2021 (Julien and his family’s cargo bikes – Designing a bike solar!), it was time to solarize my daughter’s bike, now that she’s old enough (10 years old) to ride a solar bike without too much risk – which is why I’ve only electrified my 7-year-old son’s bike so far.
The basis of the setup is her everyday bike, which can be converted to solar in less than an hour. The solar version was completed in May 2025 and then dismantled at the end of August after approximately 1,500 km, mainly during our summer trip combining sections of EuroVelo 6 and 1.
Detailed technical description
◦ The bike
Woom Now 6 (Austrian children’s brand) in 20/26-inch front/rear with a schoolbag holder at the front, but above all, weighing 12.8 kg. This fairly high frame leaves enough room to integrate solar structure supports on the roof.
The compromise between dual school/travel use lies in the very raised position, which is comfortable and safe for everyday use in the city but not very suitable for long-distance solar use: too much weight on the buttocks and an unnecessarily high solar structure. I should have changed the handlebars before the trip to lower the hand position.
This would have helped relieve the postural pain that appeared on long stages – a 15 to 20 cm height difference between the saddle and the handlebars being far too much in this case.
The only modification to the frame is the drilling of the down tube to install the threaded rivets holding the battery holder.
However, it was a clear error not to have taken into account the thinness of the down tube: repeated vibrations due to unpaved roads and unpaved paths got the better of this mounting method, which had been tried and tested on other frames with much thicker tubes.
Additional tape support allowed me to complete the loop without tearing out the rivets, but it would have been more It’s a good idea to print custom PETG fasteners or use Grintech’s “double bob.”
Second mistake: using the bike’s kickstand thread on a version of the bike three times heavier than the original. After 15 days of use, repeated mounting and dismounting of the kickstand under load got the better of the four small weld points, which eventually failed. Decathlon’s very Suntripian solution, the €7 telescopic walking pole, allowed us to continue our adventure while still being able to recharge during breaks.
◦ Motorization
Recovery of two 2020 Grin Bafang G311s (250W category with gears) from previous cargos, each of which had suffered quite a bit (damaged casing on one, advanced rust due to a faulty seal on the other). By salvaging the functional parts from these two identical engines, I was able to rebuild a fully operational one.
The choice of mounting on the front wheel was imposed on me, but ultimately proved to be a good one to better distribute the front/rear weight balance, which is heavily weighted towards the rear with the saddlebags and the raised position.
This small engine no longer needs to prove itself and has proven itself perfectly suited: reliable, lightweight, low-noise, and with enough torque at (very) low speeds to tackle the steepest climbs, despite its power being legally limited to 250W in the CycleAnalyst. On a mountainous route, the solution would obviously not have been suitable, but on a river or along the Atlantic coast, no problem.
◦ Electrical
Also taken from our stock of batteries used daily, namely a 48V 13S7P Li-ion from 2018 in a Hailong case equipped with a standard BMS.
However, I didn’t measure the cell balance before leaving, and given that our daily charger is set to about 80% of the nominal capacity—to avoid full charges that unnecessarily strain the cells—I suspect the battery has poorly balanced sets or tired cells, which significantly reduces the energy capacity.
This should be around 400 Wh rather than the nominal 640 Wh, but sufficient compared to the 564 Wh consumed over the longest 106 km leg. Combined with 160 Wp solar power (or about 640 Wh solar per day on average in summer), it never dropped below 45 V, always leaving a minimum of 30% capacity remaining.
The Baserunner controller (inverter) is located in a neoprene pocket attached to the front of the bike, thus outside the battery box, for two reasons:
• Avoid soldering a cable along the controller into the narrow battery tray for solar injection or DC/DC power supply
• Better ventilation of the controller outdoors
Also install a 4-port Anderson 30A distributor to distribute the 48 V between the battery, controller, MPPT, and DC/DC: a practical solution and proven, also ideal for detecting potential faults and offering reliable and robust power connections (when a minimum of care is taken in the connector design…)
◦ Controller and on-board computer
Once a CycleAnalyst, always a CycleAnalyst! This is the fifth setup based on this, so I didn’t even consider another solution, even though this application, based on a 1970s microcontroller with 100% analog regulation and an LCD display, is clearly no longer cutting-edge.
◦ Pedal sensor
This particular point requires a dedicated paragraph here, because, compared to a conventional solar bike (adult with luggage), there are two important differences:
• A much lower total rolling weight of around 60 kg, therefore greater sensitivity to variations in electric assistance (drastic difference in perceived assistance between 50 and 100 W on the flat, for example)
• Use by a child who does not yet understand the concepts of power and torque
For economic reasons, I first installed a conventional PAS with a Hall sensor, but on test weekends, my daughter tended to use a lot of assistance to let herself be carried along by the bike.
Installation of a torque sensor to encourage the rider to exert less pressure on the pedals. Unfortunately, it comes with 175mm cranks with a specific press-fit, compared to the original 150mm cranks on the bike, a size much more suitable for children. Furthermore, it has now been proven (among other things here) that shorter cranks allow for more efficient pedaling.
Furthermore, due to the particularly low bottom bracket on this children’s frame, she even had a tendency to scrape her sole against the ground when pedaling with shoes that were too soft, hence the need to shorten the cranks.
According to internet forums, a semi-divine level of DIY is required to properly tap new pedal threads—when crank arm forging lends itself to it, which is the case here—and left- and right-hand tap sets (pre-cut, cut, and finish per side, i.e., 6 in total) are also prohibitively expensive.
Feeling adventurous, I attempted drilling with a size 13 drill bit on a drill press and threads with two basic pedal threading taps. Result: pedal axles perpendicular to the cranks and no loosening over 1,500 km.
The problem therefore seemed to be solved with sufficient ground clearance and a cranks preventing hip swaying on the saddle. On the other hand, her naturally weaker muscle strength, combined with a shorter lever arm, forced the torque sensor’s strain gauges to work in their lower detection zone, and even with significant multipliers in the CycleAnalyst (up to x6), she rarely exceeded 150W, which wasn’t enough on some climbs.
After two days of traveling, realizing that this solution seemed too physical for her, I set the torque sensor in the CycleAnalyst to single PAS (like her brother), and the two children finally managed to find a compromise between minimal pedaling and battery conservation, each achieving a consumption of 5.5Wh/km at an average speed of 18km/h by the end of the trip, which is quite respectable.
◦ Solar equipment and/or solar trailer
- Photovoltaic panels
Even though the range of semi-flexible monocrystalline panels has expanded considerably compared to previous years – especially with a dramatic drop in peak wattage prices – it has not been easy to find panels of the right size for this application.
The best compromise was (imperfect tense because the product was no longer available at the time of writing) proposed by Greenakku:
• Semi-flexible module with a structured front panel – advantageous for mobile use in my opinion because it is less dependent on the orientation towards solar irradiation
• 2 80Wp modules, or with the empirical 4Wh/km/day – summer use & on the roof – approximately 640 Wh/day in production compared to the 440 Wh consumed on average per day over 13 days
• 84x54cm per panel:
• 1.7 meters in total length on this bike corresponds to good shade without significant overhang towards the front or rear
• 54cm width provides good sun and weather protection while remaining very maneuverable.
• 1.7 kg / module, or 3.4 kg above the head, about 6 kg in total with the structure, making a stationary load easily handled by my daughter (once launched, the weight is much less critical)
• €70 each or 88 cents/Wp (just for reference, industrial and residential modules are in 2025 at 11 cents/Wp in Topcon bifacial glass/glass structure, i.e. technically much more advanced than these…) but this price remains more acceptable than the 180 cents/Wp that I paid in 2020 for the Sunpower 170Wp
• Sunpower cell in IBC (integrated back contact) guaranteeing better crack resistance
• Series assembly of two panels, i.e. 160Wp compatible with the classic 300Wp Suntrip MPPT: 50% of margin compared to the nominal power, a welcome reserve considering the overflowing optimism of Chinese sellers regarding the electronic capabilities of their products
• Addition of 30A Anderson connectors to facilitate assembly/disassembly and troubleshooting on the go, as they are standard on all my bikes
• Mounting using self-adhesive Velcro strips as on previous bikes, an easy-to-apply and inexpensive solution that greatly facilitates proper positioning on the frame and disassembly for transport
- Solar regulators
Recovery of a controller re-configured to 54V – a small margin of 0.6V compared to the theoretical maximum since regeneration is not possible on a geared motor – to avoid testing the functionality of the BMS. Battery 7 years old.
- Supporting structure
Two major improvements compared to my 2021 assemblies:
• The use of a 3D printer, allowing me to optimize the weight of the supports based on the specific mechanical constraints of this use
• The use of standard Kipp industrial joints, inexpensive and with a 30mm diameter, compatible with 1.5mm thick 6060 T6 aluminum round profiles and offering a weight/mechanical stability ratio interesting
Sizing the structure remains the most time-consuming step—but not the most expensive compared to the electronics and the electrical transmission chain.
Given the significant weight of a children’s bike, especially when it’s positioned high up and thus prone to involuntary sideways tipping, priority was given to developing the lightest possible solution.
Total cost: approximately €160
• Industrial fittings: €50 with delivery
• Aluminum tubes: €50 to the local dealer
• 3D parts: approximately €15 for PETG in various colors + a few euros in energy Because it takes about 100 hours of printing and filament drying time
• Self-adhesive Velcro: about €15 at the hardware store
• Spray paint: €17 at the hardware store
When it comes to 3D printing, PETG (like mineral water bottles) offers easy processing, good mechanical strength (much larger elastic zone than standard PLA), and fairly good thermal resistance/UV resistance (so it doesn’t melt in the summer sun). It can be printed at fairly high speeds, an important criterion given the size of the parts.
No complicated calculations (as I’m not familiar with Freecad’s FEM module), but rather rough sizing, which proved surprisingly stable throughout the entire journey.
Another advantage that became apparent during a fall—fortunately minor—against a pole: the PETG supports absorbed the impact energy by breaking one after the other, whereas a much more rigid aluminum structure on the axle longitudinal would surely have transmitted more energy to my daughter who might have injured herself. Makeshift repair with tape to finish the trip without any problems then simple definitive repair by re-printing the damaged parts. And curiously, the Velcro attachment prevented the front panel from being completely ruined, and it didn’t lose all of its production following the fall.
Feedback
◦ Strengths
Compared to my first solar bikes, I’m much more satisfied with this one, as I was able to pay more attention to the finishing touches since I didn’t start from scratch. This makes the overall look much less “DIY,” with many people I met even thinking it was a commercial kit. It was also important to me that my daughter could use the bike safely and enjoy riding it, both goals achieved. This bike was also presented to his fourth-grade class, and aside from the general interest shown by the students, even the little boys who were quick to compare their dad’s car expressed their desire to try riding a solar bike. Mission accomplished 😉
◦ points for improvement
Aside from my mistakes with the battery and kickstand mounting, this project proved perfectly suited to its intended use. So, once the necessary repairs have been made, no major modifications are planned before the next solar-powered installations.
◦ Statistics
As in my previous article, below are the statistics for my daughter’s trip:
No errors in the actual fuel consumption per km Negative ending: my daughter “saved” her little brother twice, who was a bit greedy for assistance. So she recovered her brother’s almost empty battery so he could finish the stages. Over the 1000km covered, only one “safety” mains recharge was used. was necessary following the breakage of a panel in the fall against the post.
My next bike / or the next evolutions
The fabulous world of recumbent bikes is catching my eye… but given that it’s taken me 3 years between two solar-powered projects, it’s definitely a long-term prospect!