Using some novel subtractive manufacturing techniques (I drilled a hole in the shell) I was able to connect a 2.1mm DC plug to the battery.
Unfortunately the battery case is not weather safe anymore - I'll need to plug some holes if I want to road test this.
Then I discovered that the "neon" LED tube I bought for testing purposes bends the wrong way to weave between the spokes. So I zip tied it. Maybe not the best solution but it got the job done.
[#]bikeTooter #diy #3dprinting #electronics
=> View attached media | View attached media
=> More informations about this toot | More toots from yantor3d@mas.to
[#]bikeTooter #diy #electronics #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
A slow day at work gave me plenty of time to iterate on the design for the case, pushing the asymmetry I added last night by moving the inner screw point to the corner so that it overlaps the neighboring shell.
I don't think I can make this any simpler - not without switching to print-in-place fasteners. All six pieces print flat on the bed, although if I really had to, or wanted to, I could model a support structure and print them on edge, Slant 3D style.
[#]3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Oh hell yes!
I stumbled across American Bright LEDs, a domestic supplier of, among other things, LED rope lights. DigiKey and Mouser are among their resellers. So it would have been possible for me to get all the electronic components for this project from a single supplier!
Although I would still have gone somewhere else for the battery cells - Mouser doesn't carry 21700 cells, DigiKey charges $10 each.
=> More informations about this toot | More toots from yantor3d@mas.to
I managed to wire up the LED rope light so it can plug into the new battery prototype and do a wheel test.
As an aside, I'm curious how this stuff is made - it looks like a series of LEDs and resistors connector by bare wire, but it's embedded in a clear PVC tube, with a center void.
At speed the LED rope lights look basically the same as the "neon" LED strip in a diffuser. The difference is the rope light has 360° visibility, while the diffuser depends on the profile.
[#]bikeTooter
=> More informations about this toot | More toots from yantor3d@mas.to
The rope light is durable and easy to install - it's thick enough that it's a friction fit between the spokes. It's waterproof, and there are multiple domestic resellers in the US.
But its expensive, and I've had a hard time finding stores that sell it by the foot, or in 16 ft/5m lengths. And the stores that do sell it by the foot don't sell all the colors. I've found one exception but they don't list prices, so I'll have to reach out and see what they'd charge.
[#]bikeTooter
=> More informations about this toot | More toots from yantor3d@mas.to
I'm beginning to see why nobody has ever tried something like this as a commercial product yet.
LED rope light retails for about $2 / ft in 150' spools. Assume $1 / ft wholesale. It takes about 15' for an adult bike. A waterproof DC power cable set is $2 each - assume $1 wholesale.
Add in the two pin connectors and some wires, and you have about $20 in parts. Rule of three says you sell that for $60.
That's just the lights. Don't get me started on what it costs to build a 12V battery...
=> More informations about this toot | More toots from yantor3d@mas.to
But on the flip side, I remember hearing one of the guys at the bike shop once remark that there's no margin in the bikes themselves, but in the accessories.
I don't know that every cyclist would be willing to drop $100 a wheel on lights when there's a "good enough" option for $15 a wheel.
But as the saying goes, "don't by cheap tools". You can get a cheap cordless drill for $30 and it will last a year, or a name brand one for $300 and it will last a decade.
=> More informations about this toot | More toots from yantor3d@mas.to
"But Ryan, your dad bought a circular saw a garage sale for $5 in the 90s and it still works!"
There's a difference between a saw used once in a while for DIY weekend projects or community theater set construction vs a saw used every day by someone in a trade.
Likewise, there's a difference between the accessories you put on a kid's bike so their helicopter parent will let them ride their bike around the block vs the accessories you put on a bike you ride to work every day.
=> More informations about this toot | More toots from yantor3d@mas.to
Took a ride in the rain to give the latest iteration of the lights a field test. The lights stayed on and the inside of the battery stayed dry.
I need to get parts for a second battery in the front wheel, but that will have to wait until the new year.
I'm finding that PETG doesn't like to have things glued to it. Or maybe I should try to glue the foam spacers on vertically, rather than horizontally along a concave surface.
[#]biketooter #3dprinting #diy #electronics
=> More informations about this toot | More toots from yantor3d@mas.to
I reached out to the lighting company that had SKUs in all the different colors. Turns out they don't do custom orders at this scale (15'). Also marine rope light is way over spec-ed for what I'm after. But, the sales rep did point me at a domestic supplier of RGB LED strips, so that's good.
=> More informations about this toot | More toots from yantor3d@mas.to
I'm just a boy, standing in front of the internet, asking it why omni-directional marine grade RGB LED 3/8" rope light isn't a thing.
Best I can do is 5mm RGB LED strips in a flexible silicone diffuser tube, but I've yet to find a domesticat supplier.
=> More informations about this toot | More toots from yantor3d@mas.to
AdaFruit has a cool, dual edge RGB LED strip that would be amazing on bike wheels, but there's two problems.
It's only 1 meter long, and it takes 5V. I need a little over 2 meters per wheel, and 12V.
As much as I wanted to avoid sticking the LEDs to the wheel rim - weaving them between the spokes is better long term - I may have no other option...
=> More informations about this toot | More toots from yantor3d@mas.to
I just had a terrible idea.
I really, really want marine style RGB LED rope lights - bulbs spaced 1" apart in a clear tube for 360° visibility.
I may be able to just make this myself... looking at the solid color LEDs I have, it's just a bunch of LEDs in series-parallel with a resistor.
Edit: ugh, to be able to wire RGB LED bulbs in series you need the six pin variety...
=> More informations about this toot | More toots from yantor3d@mas.to
!!!
I found a shop that sells neon RGB LEDs that meets all my requirements!
✅ 12V DC
✅ Vertical bend
✅ Narrow profile (0.4" / 10mm)
✅ IP68 waterproof
✅ Domestic supplier
Only drawback is the cost - an 8' length is $134. Each. You could get the supplies to make these for $80 plus shipping from China.
But these would be perfect for bike wheel lights!
[#]biketooter
=> More informations about this toot | More toots from yantor3d@mas.to
As I'm continuing to spin my wheels (see what I did there?) about the options for wheel lights I had an "oh, right, duh" moment.
You don't need to do a continuous strip of lights around the rim. It's just what I want for my bike.
You could do a trio of strips along the spokes. You could do individual lights on the spokes. You could program an array of lights to animate patterns.
The important thing is that the battery will last longer than most of the off the shelf options out there.
=> More informations about this toot | More toots from yantor3d@mas.to
Having 12V available to power my hypothetical wheel lights has opened up so many options for repurposing existing products.
Boat lights. Car lights. Motorcycle lights.
So many options when you're not limited to a couple AAs or a watch battery.
=> More informations about this toot | More toots from yantor3d@mas.to
As I continue to my search for the ideal source for 12V bike wheel lights I've come across a company that sells wheel lights for motorcycles at any length in roughly 2" increments, from 2" to 197".
They're not "neon" LEDs but the ability to order custom lengths makes up that.
https://www.boogeylights.com/low-profile-led-light-strips/
=> More informations about this toot | More toots from yantor3d@mas.to
One of my Christmas gifts for my sister was a tune up of her bike. While at the shop I eyeballed the hubs and they all seemed to be a fairly consistent size and shape. All look to be within plus or minus 10mm of each other.
=> More informations about this toot | More toots from yantor3d@mas.to
It occurs to me that the concept of adding lights to a bike for safety instead of aesthetic is probably so American (and Canadian) on a concept that even someone bilingual and fluent in Dutch and English could not explain it to someone in the Netherlands.
"What do you mean the bicycle lane in the street!? And there's no barrier!?"
=> More informations about this toot | More toots from yantor3d@mas.to
The excitement of the hub battery / wheel lights project has waned a little as I've been on vacation and haven't been able to do anything but think about it.
I could keep it to myself. I could release it open source. I could sell the model and the plans. I could gather the parts and sell kits. I could build and sell whole units, battery not included because shipping.
Each level involves more effort, more risk, more return.
=> More informations about this toot | More toots from yantor3d@mas.to
Keeping it to myself is the easy option. It's out there and if someone is motivated enough, the could recreate the results from scratch.
Releasing it open source means a lower barrier of entry for someone to get it for themselves, and also they can change it to fit their needs.
=> More informations about this toot | More toots from yantor3d@mas.to
Selling the model and a set of plans under my own name comes at a risk; incorporating an LLC comes at a cost. And there's the question of whether anyone would buy an empty battery case and a set of instructions that then tells you to go buy more things that can't/shouldn't be all gotten on Amazon.
=> More informations about this toot | More toots from yantor3d@mas.to
Assembling kits with all the components (sans battery cells) and selling those suddenly makes me the bottleneck. I'd have to maintain inventory of the parts, and either print cases on demand or do bulk orders from a print farm. Then pack and ship and all that jazz.
Building and shipping full batteries (again, sans cells) adds another several hours of work to each unit and increases the number of things that could go wrong. And increases the amount of my time I need to charge for.
=> More informations about this toot | More toots from yantor3d@mas.to
I always tell myself not to monetize my hobbies, because then they become chores, but I feel I have something cool here; I just can't see how to make it scale to profit.
I have a good job (for now) and I'm doing okay (for now), but I see the appeal of the capitalist myth. Invent an awesome product, sell it for a profit, retire early.
The imposter syndrome is loudly telling me I don't deserve to be well off, especially when my sisters are struggling to make ends meet. What gives me the right?
=> More informations about this toot | More toots from yantor3d@mas.to
I could keep spiraling through this loop for dozens more posts, but Mastodon is not my therapist, so I won't.
I need to make sure that the hub battery even works before I commit to next steps - and that means building the a second one, installing the lights, and riding with it for several weeks or months.
My math says the battery should last up to 250 hours. I think my math is wrong. But even 20 hours of runtime would mean I would only have to charge once every week or two...
=> More informations about this toot | More toots from yantor3d@mas.to
A blindingly obvious realization about the battery lets me move one more component from "order online" to "get it at the hardware store"
Metal is conductive.
Screws are made of metal.
If I can confirm hardware stores sell individual 2.1mm DC extension cords, I can modify the case and write a BOM for the build that gets everything from the hardware store except for the battery and battery management system.
Progress!
=> More informations about this toot | More toots from yantor3d@mas.to
Actually I'm fine with the DC plugs being bought with the cells and BMS since that can still all be done at one site, assuming they're in stock. I've been trying to avoid a BOM that lists multiple suppliers.
Anyone know of a store* that sells metric fasteners AND lithium ion battery components?
=> More informations about this toot | More toots from yantor3d@mas.to
Fuck me in every orifice, clockwise starting from my left ear. There is a "store" that carries all the components necessary to build the battery, assuming the "screws as contacts" design.
Wal-Mart. Mother. Fucking. WAL-MART.
They have the metric fasteners.
They have the 2.1mm DC connectors
They have the lithium ion cells
They have the battery management system
They have the incidentals (wires, wire strippers, screw terminals, wire nuts...)
(Screams into the void)
=> More informations about this toot | More toots from yantor3d@mas.to
With the perspective of a good night's sleep and the last day of the year, I've been getting way ahead of myself, as usual.
I've created something new, something potentially useful, but I need to prove out it's usefulness. I got caught up in worrying about profit and production.
The lights arrive on the 2nd, as will the components for the second battery. I've got a few days off before I go back to work to do a few test rides. And a longevity test - up to 24 hours continous use.
=> More informations about this toot | More toots from yantor3d@mas.to
Get in loser, we're doing science.
[#]bikeTooter #diy #electronics #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Three hours later, and the light's are still on. So it would last for a commute, at least.
I'm taking a reading every hour with a multimeter - both volts and amps. I need some more components to wire in the power meter.
| Time | Volts | mAh |
| ---- | ---- | ---- |
| 00m | 12.44 | 16.7 |
| 60m | 11.36 | 15.4 |
| 120m | 10.45 | 13.87 |
| 180m | 9.28 | 12.26 |
The specs say these LEDs are 12V 2A, and the battery is 12V 5A so shouldn't it have already gone out yet? (Not that I am complaining...)
=> More informations about this toot | More toots from yantor3d@mas.to
An update - the battery died just short of four hours, reading 0.0 V at P+/P- and 8.4 V at B+/B-, which tells me that the BMS went into standby. I reset it by plugging it into the charge for a split second, and the lights came back on for a few minutes.
So, four hours at full power, give or take. Tomorrow, we'll try it at 50% power to see if it lasts eight hours.
=> More informations about this toot | More toots from yantor3d@mas.to
I also need to test it with the LED rope light; it has fewer LEDs and may last longer. I should also test it with three 12" segments to emulate what a spoke based setup would look like.
=> More informations about this toot | More toots from yantor3d@mas.to
I am, as they say, an idiot. Or more precisely, self taught at electronics and still learning.
The 12V LED strip is wired in series parallel, with three LEDs to a series. Each series draws about 20 mA. There are about 150 LEDs in the strip, so fifty serieses each drawing 20 mA, for about 1,000 mA.
Which explains why the battery only lasted four hours.
=> More informations about this toot | More toots from yantor3d@mas.to
So, that would mean that a 3 or 4 LED module on its own would last around 250 hours; three modules would last around 80 hours. A trio of 12" light bars would last 20 hours. In theory.
So, the trade off - do I want bright lights around the whole rim that need to be charged every day, or do I build a lower density light strip that can last for a week or two? Or settle for three line of lights that blur together into a disk? Or three modules of lights that blur together into a ring?
=> More informations about this toot | More toots from yantor3d@mas.to
Coming full circle, if I went back to the glow in the dark rims powered by four small sets of UV LEDs, a 12V battery could power that for days, maybe weeks.
But then I'm limited by color, and daylight. Although I should test how well glow in the dark filament does compared to glow in the dark paint...
=> More informations about this toot | More toots from yantor3d@mas.to
Test two underway - same setup as before, an 8' 12V LED strip powered by a 12V 5A battery in a custom 3D printed case, simulating a bike wheel rim light.
Prior results - About 4 hours runtime at 100% brightness
Prediction - About 8 hours runtime at 50% brightness
Science!
[#]biketooter #diy #electronics #3edprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Test two has exceeded the benchmark set by test one - after five hours at 50%, the lights are still illuminated, and do not appear to have dimmed. Although since the battery reads around 9V, I don't know if it will make it to eight hours.
Also of note - the factory length of the LED strips was about 98" (2.5m) - but I only need 70" for the bike rims. So in theory I should get 30% more battery life than what my bench tests have shown.
=> More informations about this toot | More toots from yantor3d@mas.to
Test two just passed seven hours.
And I am finding beauty in the math.
At factory length, the strips had 300 LEDs (120 / m). When cut down to fit my bike wheels (70"), it's about 210 LEDs.
I also have an LED rope light that has an bulb every inch - so cut down to fit my bike wheels it's about 70 LEDs.
So, in theory, the rope light should last 3x as long as the LED strip.
=> More informations about this toot | More toots from yantor3d@mas.to
Which one has more ✨️aesthetic ✨️ (poll in reply)?
[#]bikeTooter #3dprinting
=> View attached media | View attached media
=> More informations about this toot | More toots from yantor3d@mas.to
Which one has more ✨️aesthetic✨️ (photos in previous toot)?
[#]bikeTooter #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Test three has been underway all day, using the LED rope light meant for marine uses.
After seven hours, it has drained the battery from 12.4V down to 11.4 V - for comparison, the LED strip did that in an hour at full brightness, and in two hours at 50% brightness.
Which makes sense, as the rope light has approximately 1/5th as many LEDs as the strip.
I am suspect that the rope light could last 24 hours, but I am confident a smaller setup - maybe three strips along the spokes - could.
=> More informations about this toot | More toots from yantor3d@mas.to
Test three has hit the 24 hour mark.
Judging by the voltage reading on the battery and prior tests, it will last for a few more hours. But 24 hours of illumination would be enough for two weeks given my usual riding habits - a month if I only used it when commuting.
I'm pleased by the results, and looking forward to what "light bars" on the spokes will manage.
[#]bikeTooter #diy #electronics #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
I ran the lights for a few more hours tonight - they finally went out at around twenty seven and a half hours. Not bad.
I should probably stress test the cheap bike wheel lights I got that run on three AAAs to get a comparison, even if comparing a 5V strand of twenty micro LEDs to a 12V strand of 66 LED is probably silly.
=> More informations about this toot | More toots from yantor3d@mas.to
I have one last design change I'm considering for the battery case.
Currently it's printed in six separate pieces, which are then paired into three cases. Each case has half a hinge on either side that gets fastened to its neighbor to make a full ring.
Now I could make the hinges print in place, resulting in two assemblies that print in place. But that would force it to print flat on the bed, and each individual piece already has a lot of surface area, so it sticks to the bed until cooled...
=> More informations about this toot | More toots from yantor3d@mas.to
The advantage of the print in place hinges is it makes the design idiot proof - it goes from six parts that can be incorrectly paired to two parts that can only be assembled one way.
The disadvantage is the print probably can't be auto ejected; it's not something I can test at home. So that would close off the ability for me to bulk order them and offer DIY kits with all the parts (sans batteries).
=> More informations about this toot | More toots from yantor3d@mas.to
...actually, I just realized - making the hinges print in place would introduce more slop in the hinges because there would have to be a gap to prevent the pieces from melting together while printing.
So nevermind. Maybe I can add some finding features so that each half of a shell can only fit together with its counterpart.
=> More informations about this toot | More toots from yantor3d@mas.to
One thing that's always bugged me about the case is that it prints flat, which means it has about 4,000 mm^2 of surface area in contact with the bed. This makes it difficult to remove, even once the bed is fully cooled.
Printing it slanted means less surface area in contact with the bed - about 1,000 mm^2, and a uniform surface texture.
The slicer is complaining a little about the threads, but I think it will be fine. We'll know in about two hours.
[#]bikeTooter #diy #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
One minor change to the design is that I need to change from flat head screws to socket head screws - countersunk screw holes are a liability when printing at an angle because it results in unsupported overhangs. Switching to hex shaped counterbores solves the problem.
=> More informations about this toot | More toots from yantor3d@mas.to
The slanted print worked, although I need to adjust the interface at the bottom to use many short sprues instead of one long one.
I still need to figure out what to do about the countersunk screw holes, because I prefer flat head screws for the design. I guess I'll have to do some sort of support; maybe a sacrificial layer?
=> More informations about this toot | More toots from yantor3d@mas.to
The cells for the second battery arrive tomorrow, and I'm doing the final wiring on the neon LED strips for both wheels.
Soon...
=> More informations about this toot | More toots from yantor3d@mas.to
I need to find a better solution for wrapping the wires because overlapping segments of heat shrink looks tacky.
=> More informations about this toot | More toots from yantor3d@mas.to
Whose got two thumbs and forgot to post the first look photos in this thread?
This guy!
https://mas.to/@yantor3d/113808240398098366
=> More informations about this toot | More toots from yantor3d@mas.to
Technically speaking, the hub battery is at MVP. I've ironed out the major issues with the case, and with assembly and wiring. I've got a battery installed in both wheels and lights, too.
The outstanding issue is how to deal with variety of outer diameters on hubs. I've got a temporary fix - closed cell foam - but it seems like a jack. I'm experimenting with grip fins on the case itself and they work up to a point. I'm also going to try a flexible spacer.
=> More informations about this toot | More toots from yantor3d@mas.to
TIL that soldering 5mm RGB LED strips is a special hell. It can be done, but it's too easy to break the connections while installing them on a bike wheel.
CW: flashing lights.
=> More informations about this toot | More toots from yantor3d@mas.to
I've got five meters of RGB LED strip, so I'll try again tomorrow. Sunset is soon and I want to take the lights I have out for a field test.
=> More informations about this toot | More toots from yantor3d@mas.to
Field test one was a successful failure.
I set out around sunset. The lights were bright and the batteries were tight. But, both were dark when I got home.
I don't know when the rear light cut out, or why. My theory is either the battery wasn't charged or it shorted in the rain (because Portland).
The front light went out when I accidently went up and over the median in an unlit intersection of a side street. I assume it got jostled and maybe something shorted.
[#]bikeTooter #diy #electronics
=> View attached media | View attached media
=> More informations about this toot | More toots from yantor3d@mas.to
Both batteries reset when I got home and plugged them into the charger, so I assume the internals are still good.
Still, this got me thinking; the BMS I'm using is the DF Robot FIT0809 - because it's the only one I could find from a reputable supplier. It can be reset by connecting a charger, or by shorting P+ / B+.
Now, taking the battery off the wheel and opening up the case on the side of the road to reset it sounds stupid. But what about wiring in a reset button?
=> More informations about this toot | More toots from yantor3d@mas.to
My up and over with the median had me worried that one of cells had come out of the clips, or a clip had come free from the case. I'm pretty sure that didn't happen but now I wonder if I can come up with a more robust solution - maybe 3D printed clips in both halves of the shell, so the cells are totally immobilized?
Jarvis, start a new design file in the hub battery project for the mark VII iteration.
=> More informations about this toot | More toots from yantor3d@mas.to
Hub battery mark 7v1 has been sent to the printer. We'll see if it works in two hours and thirty minutes.
I've shifted from using metal clips (Mouser 534-247) to a 3D printed clip with spring clips (Mouser 789-209). This should completely immobilize the cells in the case, and will eliminate the need to secure the clips with glue, which has the added benefit of making it more repairable.
One downside of this design is it uses more filament - 60g vs 45g.
[#]bikeTooter #diy #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d huh, that could make mounting easier. Two loops, one on each side. No weaving back and forth, no LEDs blocked by spokes, all LEDs facing the right direction.
=> More informations about this toot | More toots from trouble@masto.ai
@yantor3d That looks to be sufficient performance to make the project very successful.
Pity such lights would be just as likely to get me a traffic ticket as save my life around here :(
=> More informations about this toot | More toots from LovesTha@floss.social
@LovesTha uh, what automobile centric hellscape do you live in that gives out traffic tickets for making your bicycle too safe?
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d can't have white lights pointing anywhere but forward. These lights would need to be orange.
I'm in Melbourne, Australia.
I may have been hyperbolic, but it is at least a coin flip. Some of it there are few cyclist deaths.
=> More informations about this toot | More toots from LovesTha@floss.social
@LovesTha Well, you can get LED rope light in all the colors, but I've never found it sold be the foot in anything but red, blue, or white. And it's normally sold in 150' spools...
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d If I recall correctly, you are using PETG for the parts? My experience printing with PETG on a PEI plate is that it is best to clean the plate with window/eyeglasses cleaner without ammonia and to remove the part when the bed is between 60°C and 40°C. I haven’t had any issues printing with PETG after I started doing this.
Hope this helps.
=> More informations about this toot | More toots from Siff@mstdn.social
@Siff Yes, I am using PETG.
I haven't cleaned the plate since I got it, and I've found that no matter what, the part sticks, even if I let the plate fully cool. The windex might work, but I have an eye towards delegating this to a print farm so I need to design for auto ejection.
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d The black (assuming that the screws are weather resistant)
=> More informations about this toot | More toots from Siff@mstdn.social
@yantor3d silver screws.
but gee whiz, glue some gears on it would you?
:]
=> More informations about this toot | More toots from RueNahcMohr@infosec.exchange
@RueNahcMohr ...I have been trying to think of something to emboss in the shell to reduce the surface area of the first layer. A gear motif might work!
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d 'random' surface jogs and embossed blitherish work too :]
=> More informations about this toot | More toots from RueNahcMohr@infosec.exchange This content has been proxied by September (3851b).Proxy Information
text/gemini
