Solar trailer build

Spring is just around the corner, and today is dry. I finally could resist no more, and so the first of my Sunpower 110W solar panels has finally made it out of its box, onto a heavyweight piece of plywood and into the wind and occasional sun. As the picture shows, for 'open' systems, adding solar input is quite simple. Just a panel, a suitable charge controller and plug in to your battery's charging port. You can connect to the output port as I have here, but better to go via the proper route to let the BMS do its stuff.

 

Is it worth it? That depends on how far, how fast, how off-grid you travel. Not much good on the crowded London streets, but on the open road for wild camping tours in Scotland I'm expecting it to be quite useful.

 

I use Anderson powerpoles for most of the connections, and standard solar MC4s to connect to the panel.

 

That plywood might make the odd local journey on the trailer, but my aim is something much lighter. It is a start!

 

That's great! So you're charging at about 4A? I've noted the parts (cheers!), in case I fancy having a go at this sort of thing sometime. Are you planning to add the other panels you mentioned, to this charging system? How many? Will the panels fit inside the trailer? Will you need a black belt in Origami, to make them fit inside trailer? Where do snarks be? Will you need a bigger net?

Edited February 28, 20224 yr by guerney

That's great! So you're charging at about 4A? I've noted the parts (cheers!), in case I fancy having a go at this sort of thing sometime. Are you planning to add the other panels you mentioned, to this charging system? How many? Will the panels fit inside the trailer? Will you need a black belt in Origami, to make them fit inside trailer? Where do snarks be? Will you need a bigger net?

In good sunshine, in northern UK, panel output will get up to 80% or so of the 'rated' 110W, and by the time it is 48 voltified by the charge controller, each panel will provide a max of about 2A. I'll start with a single panel, but the aim is to have 3.

 

I would like to end up with a mounting system allowing the three to be stacked when necessary for compactness, and fully extended when riding. One panel area is not an unreasonable trailer size in any situation, whereas 1.7m long by 1.15m wide is a bit of an inconvenience!

 

This roughly 300W of panel seems about the going rate for solo solar bikes.

 

Need a sensible balance between weight added and power/energy gained, as I live in hilly parts and my little motor has to get all the weight up them.

...and plug in to your battery's charging port.

Can you do that with Shimano batteries, just connect +/- volt to the charge port? I thought there'd be some kind of smarts like there is with Bosch to prevent you doing that?

Solar panels...

Scotland...

I think I've discovered a flaw in your plan ... lol

Can you do that with Shimano batteries, just connect +/- volt to the charge port? I thought there'd be some kind of smarts like there is with Bosch to prevent you doing that?

Yes, all sorts of problems there. I am on a two bike approach for that reason. I have careful plans to 'supplement' the Shimano battery rather than expect to charge it directly from solar, and I am building a TSDZ2 conversion onto an old mountain bike, which will have no solar issues, just 'where does the battery go?' and 'can I fix a trailer tow hitch to that?' kind of problems instead.

 

I have read online of successful parallelling of non-Shimano batteries with the Shimano one, adding a second BMS in parallel to the Shimano one, and also various expensive scenarios of bricking a battery by allowing current to flow backwards through the discharge port.

 

So lots of care, ideal diodes, datalogging, hacked chargers and above all only risking my oldest battery when anything potentially harmful is attempted for the first time.

 

It might be a long slow journey!

I know sod all or very little about solar so a few Q.

Panels of a similar size/dimesions can have a different wattage rating Y or N !!!

 

For instance the one in Matthews pic is 150w and the one in the ebay link says 1000w and looks to be of a similar size. So is the ebay one advertising a bit false on the wattage ?

1000W Flexible Solar Panel Kit 18V Controller Dual USB Battery Charge Camping | eBay

 

Or are they like ratings for 18650/21700 lion cells and can have a different rating depending on the quality of the panels solar cells used.

I know sod all or very little about solar so a few Q.

Panels of a similar size/dimesions can have a different wattage rating Y or N !!!

 

For instance the one in Matthews pic is 150w and the one in the ebay link says 1000w and looks to be of a similar size. So is the ebay one advertising a bit false on the wattage ?

1000W Flexible Solar Panel Kit 18V Controller Dual USB Battery Charge Camping | eBay

 

Or are they like ratings for 18650/21700 lion cells and can have a different rating depending on the quality of the panels solar cells used.

Mine is 110W, they do make a very nice 170W but I could not source them in the UK at a sensible price. They ought to be 1.5 times the 110W, since they are exactly 1.5 times the number of cells, but I could not get below 2.25 times as much. I could have had them from Grin at 300USD back in December, but worried about the transport from Canada.

 

The lightweight semiflexible panels, which are the only sensible choice for bike applications, share with batteries 'you get what you pay for'. Cheap will disappoint, quality will last ten years plus without much loss of output. Cheap ones fail because the cells break, not necessarily visible to the eye.

 

So, must be Sunpower cells, look them up for why, and I found genuine Sunpower panels in the UK for close enough to the same price as Sunpower cells assembled by another company, so I paid the bit extra.

 

That listing is a nonsense: not even a typo of 100W, the panel is so small probably barely 20. Work on 150 to 200W peak per square metre of panel area. Anything seriously outside that is nonsense.

 

The panel power is what's called peak output, formally written Wp, and is output in standard lab conditions. Real world in the UK, expect 80% of that in good sunshine when facing the sun, and in the overcast sky here today, I am averaging about 10%.

 

For low cost solar, including these panels, Bimble Solar.

Thanks for confirming, I did wonder if the photo cell quality differed. Hence asking.

 

The outlay for moving or wild off grid power isn't cheap so the alternative for simple home off grid use just simply wouldn't recoup any outlay costs.

Would the noise of a wind turbine like this scare off snark? 11kg doesn't include the mast - not a lot of detail in the advert, and it'll likely be highly impractical transport and setup-wise for hunting expeditions in the deep Scottish windy winter wilds. I'd need planning permission for use on my roof, a mate was refused such recently for his commercial property.

 

https://www.amazon.co.uk/hongmei1-Horizontal-Generator-Certificated-electricity/dp/B09NLMYHTK/ref=

 

The one shown near the beginning of this video looks quite similar:

 

Thanks for confirming, I did wonder if the photo cell quality differed. Hence asking.

 

The outlay for moving or wild off grid power isn't cheap so the alternative for simple home off grid use just simply wouldn't recoup any outlay costs.

Yes the costs and benefits have to make sense to each potential user.

 

It is a very different story costwise for fixed systems as the normal glass topped panels are much cheaper, and do not have the same issues with cell breakage. A new 100W glass panel is about £50, but a secondhand 250W panel is also about £50. Compared with my Sunpower panel, that's one tenth of the cost per W!

 

The other biking comparison is with having a bigger or second battery.

 

If you want to ride twice as far as your battery can take you (I know you are a special case as you have many batteries!), then you can buy another battery.

 

Or (on a decent day admittedly) you could tow one of these panels and get about 50 or 60W from it, which is half of what a gently ridden ebike needs, which means your range is doubled.

 

The costs are similar.

This wind generator looks easy to lug around on a bike trailer based getaway, to somewhere dark and windy. 600W peak is hard to believe:

 

https://davicifurniture.com/product/gototop-wind-turbine-600w-12v-wind-turbine-generator-red-lantern-vertical-wind-generator-5-leaves-wind-turbine-kit-red/

This wind generator looks easy to lug around on a bike trailer based getaway, to somewhere dark and windy. 600W peak is hard to believe:

 

https://davicifurniture.com/product/gototop-wind-turbine-600w-12v-wind-turbine-generator-red-lantern-vertical-wind-generator-5-leaves-wind-turbine-kit-red/

Small wind turbines are not much use, especially little vertical axis ones like that. Conventional horizontal axis ones on a hill, yes. On a roof, pointless. Not enough wind speed, too much turbulence.

A little bit of afternoon sunshine pushed power up to around 60W, and 100 minutes worth of measurement shows 43Wh gained, so an average of 25.8W. Overcast is consistently 10 watts or so.

 

Blue sky, 110Wp panel pointing directly at the sun, 95W output this morning. Low ish temperature helps, but even so, quite a bit higher than I was expecting. This might work quite well!

 

Evening update:

 

Dull first thing, then perfect blue sky day for 6 hours, from 11am to 5pm, although the hill got in the way towards the end. Total of 560Wh, with 484 of that in those 6 sunny hours, so 80W average in the sun and 70W average all day.

 

That is with manual tracking i.e. moving the panel to face the sun every hour or so. Maximum instantaneous power I noticed was 98W. This is 20% higher than I expected to see, so a good benchmark test.

 

Two or even three panels in the summer is going to quite an experience!

Edited March 1, 20224 yr by matthewslack

A little bit of afternoon sunshine pushed power up to around 60W, and 100 minutes worth of measurement shows 43Wh gained, so an average of 25.8W. Overcast is consistently 10 watts or so.

 

[ATTACH alt=20220228_163510.jpg]45910[/ATTACH]

 

That's a very interesting looking beadboard! So you've got an Arduino, programmed it - is that a memory board addon card at the top with sd card for recording readings? Programmed in Python? Did you have to write a program from scratch or was there something suitable floating about on the interwebs? Do you have a circuit diagram? No worries if you've put this together without one and you don't have one to hand.

That's a very interesting looking beadboard! So you've got an Arduino, programmed it - is that a memory board addon card at the top with sd card for recording readings? Programmed in Python? Did you have to write a program from scratch or was there something suitable floating about on the interwebs? Do you have a circuit diagram? No worries if you've put this together without one and you don't have one to hand.

Arduino nano, which has an avr328p processor. I've been messing about with avr since 1999, well before I heard of arduino.

 

This one is quite basic code except that I use a timer interrupt rather than the arduino time related functions to get precision timing.

 

There is code out there for all the add on boards, and good tutorials for how to use them, and that is where I start with each project because it is really quick to get going.

 

This one has the micro SD board on the spi port, tiny ssd1306 oled display on the i2c port, two LEM LTC6 current sensors, three resistor potential dividers for voltage measurement and a TL431 voltage reference on 6 of the ADC channels.

 

The LEMs were available rather than chosen, very good and hall effect so isolated from the circuit unlike a shunt, really easy to use, but consume about 30mA each, so in future I'll move to a different way of measuring current.

 

The code samples all 6 ADCs at 125Hz, calculates power each time from each related pair of V and A sensors, accumulates Ah and Wh, and reports to serial port and logs to SD card at 1 second intervals.

 

The code is too messy to publish just now, and I'm still working on what goes on the tiny display. It's good enough for this stage of the project.

 

There are two channels, YELlow and PURple, for solar and battery, and the sum of them is what the motor is using. There will be a larger display in the future, driven by the serial (USB) output.

Solar panels...

Scotland...

I think I've discovered a flaw in your plan ... lol

A better idea might be a large roof with a gutter, the rain water then driving a turbine, suitably geared. The added benefit is you would stay dry.

I remember little water turbines in burns were quite common when we went on holiday in the Highlands in the 60's. Three or so little wooden blades with used tin cans at the end. I think they generated enough power for a weak light bulb or two. (no LED bulbs then)

The underpinnings are taking shape. The base rectangle is an old Burley Solo, with a spare towbar put in backwards and joined to the original for extra length. The framing for the panels is reclaimed 25mm aluminium square section tubing from an old aviary, joining plates are new 2mm and 1.5mm aluminium sheet 6082T variety, all riveted except the four leg ends are bolted to the Burley brackets. Blue pipe to deflect any unintended interactions.

 

There will be weight on a platform or box underneath which will need to be sufficient to counter aerodynamic tendencies, and low c of g to avoid rollover.

 

1670mm long by 1160mm wide, for three 110Wp Sunpower modules, which will be secured by double sided outdoor mounting tape and bolts in the provided grommets.

 

More bolts and bracing to add, and the towbar may get a little longer.

 

Well you haven't been slack mathewslack! Very nice build! Solar panels are not cheap and appear to be fragile - can you add downforce using skirts on the sides and some sort of aerofoil underneath like an F1 car, to stop it taking off in the wind? Maybe curve the panels, or keep a hinged middle panel slightly lower while the trailer is in motion, so that they're slightly concave when pointed skywards, thereby reducing lift? When stationary, you could put them all back in the same plane?

Edited March 16, 20224 yr by guerney

Trial of two battery setup.

 

For my Shimano solar bike, I am using a generic battery in parallel with the Shimano one, with Schottky diodes linking the positive cables to the feed to the motor and controller. The normal Shinano communication wires remain in normal place from the Shimano battery to the rest of the system. The solar charging will feed into the generic battery.

 

The operating principle is very simple: whichever battery is at the higher voltage drives the bike. Most of the time this will be the solar charged generic battery.

 

I have just completed a 16km, 200 metres of climb test ride of the parallel battery setup without the solar connected. The generic battery was nearly full, 40.8V, and the Shimano one about 60%, 37.8V at the start. My two channel datalogger was watching the output from the batteries. Finishing voltages at rest after recovery time were 38.1 and 37.5V respectively.

 

The result was as expected, but of course I needed to see it to believe it. 85Wh taken from the generic battery and 8Wh from the Shimano. Those few are when under load up hills, the voltage of the generic battery is pulled down to that of the Shimano, and they then share the load.

 

The Shimano display only notices the Wh taken from the Shimano battery, which is a nice easy reference to what's going on. So 16km in level two assist, took indicated range down from 64 to 61! In normal times that would have gone from 64 to about 30.

 

The charging side of things is already tested. Now I have to complete the trailer build and make everything robust and weatherproof.

Tales of two batteries, two charge controllers and one EU.

 

Generic batteries

 

I recently bought a new cheap 48V 20Ah Psppower 21700 Samsung cells claimed, but haven't looked inside, for my conversion bike, which tested at 17.5Ah and 840Wh on my charge/discharge rig. It is no doubt 13S4P with cells named 5000mAh, giving the 20Ah nominal capacity. I don't know whether only 17.5Ah down to 3.0V per cell is normal for any Samsung 21700 cell or not. I was a bit disappointed, but it was a £293 delivered battery.

 

I even more recently bought a secondhand hardly used Greencell brand 36V, 13Ah, 468Wh claimed battery as the generic second battery for my Shimano running on the sun project. It tested at 473Wh and 12.9Ah, so near enough exactly as labelled. Very pleased with that for £153!

 

Boost charge controllers

 

Solar bikes usually have higher battery than panel voltage, so need a boost charge controller. There are also no known (to me) 36V normal charge controllers.

 

I bought three cheap Elejoy mu400sp controllers which in my early experiments worked as expected. These can be programmed with a set output voltage, and they then do 'maximum solar power' followed by constant voltage (CV) charging to that voltage.

 

I had a glitch last week, which seemed to be charge controller related, and with my inaugural long solar trip getting rather close, and no quick source of new Elejoy, I ordered with fingers crossed one of the 'gold standard' Genasun controllers from the makers in the Netherlands. It arrived today, in an opened inner box within an opened outer box, but as it is solar equipment and zero tariff, and was exactly what the labels outside said, came with no delay and no added taxes or fees. Phew! How far it has travelled with the boxes open, I have no idea.

 

I wanted one anyway, but at €235 plus €39.50 for p&p I needed a very good reason!

 

The Elejoys turned out to be fine, that issue is not fully understood but the workaround is to charge via the battery output rather than the charge connector. The BMS wasn't allowing the solar input to switch charging on for some reason. Having now observed on the Elejoy LED displays the CV behaviour at top of charge, I am content for now that overcharging should not happen as the BMS balancing is in play, and I set the finish voltage a bit below 42V as well.

 

One will only get near full capacity at the minimal discharge of 0.2a, typically bikes discharge rate though is anything from 13% of the controller rating to 100% so depends on the controller type & rating used and the PAS level used.

A current controller will be more frugal then a speed controller.

 

The generic Sam 5000mah cell used will be the 50e, typical capacity though can vary from 4750 - 4900 mah though mooch did get just over 5000mah on the pair he tested but that was down to 2.5v at 1a.

4600 - 4750mah is more typical with a 1a load to 3v.

 

What average current load is being applied to the Sam 21700 battery ?

 

For capacity/range I look at the typical load I will apply to a particular cell then search for discharge graphs of the cell at the current share I expect it to operate at to see the expected mah I can expect. Multiply that by the cell count in parallel to have an idea of total battery mah and range I may need.

Typically I use base line of 3.3v so not to deep discharge the cell and nearly all my BMS cut power at approx.3.2v.

One will only get near full capacity at the minimal discharge of 0.2a, typically bikes discharge rate though is anything from 13% of the controller rating to 100% so depends on the controller type & rating used and the PAS level used.

A current controller will be more frugal then a speed controller.

 

The generic Sam 5000mah cell used will be the 50e, typical capacity though can vary from 4750 - 4900 mah though mooch did get just over 5000mah on the pair he tested but that was down to 2.5v at 1a.

4600 - 4750mah is more typical with a 1a load to 3v.

 

What average current load is being applied to the Sam 21700 battery ?

 

For capacity/range I look at the typical load I will apply to a particular cell then search for discharge graphs of the cell at the current share I expect it to operate at to see the expected mah I can expect. Multiply that by the cell count in parallel to have an idea of total battery mah and range I may need.

Typically I use base line of 3.3v so not to deep discharge the cell and nearly all my BMS cut power at approx.3.2v.

Discharge resistor is 26 ohms or so, a 2kW panel heater repurposed. Current varies from 2A at full charge down to 1.5 at 39V cutoff. I expected a bit of a drop from 5000mAh, but the curves did not suggest as much as I got! However, per Wh it is cheap ish, and for my low power riding, it should do. I'd just rather have an extra 2.5Ah in the tank! I may open it up and see what cells I really have, but I won't be sending it back unless it fails.