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
The mark 7v1 worked a little too well; I had to (carefully) destroy the case to get the battery cell back out with just one spring clip installed.
Oops.
=> More informations about this toot | More toots from yantor3d@mas.to
Iterating some more on the hub battery case I've embraced the kind of geometry only practical with 3D printing to solve some of the issues caused by switching to the leaf spring battery clips; namely that it means there needs to be a wall at either end of the battery, 8mm wide, 1.6mm thick, and 12mm tall, more or less centered on the cell, which is 21mm in diameter.
So rather than waste a lot of filament building up to that height, I'll model an arch across the case.
[#]3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Eight billion people on the planet, and not one of them runs a business selling 3/8" LED rope light at custom lengths in all colors.
Capitalism was a mistake.
=> More informations about this toot | More toots from yantor3d@mas.to
One of the many advantages to working from home is I can take breaks and work on personal projects.
I've printed both halves of a mark 7 iteration and it's amazing. The cell is fully immobilized by the case itself, and the flying buttresses that the battery clips are on provide cable management for the wires.
I need to re-add all the finishing touches, and tweak the shape of the flying buttresses, but it's very close...
=> More informations about this toot | More toots from yantor3d@mas.to
The downside of being close having a finished design is then I need to make a decision about what, if anything, to do with it above and beyond my own use.
=> More informations about this toot | More toots from yantor3d@mas.to
I ordered tiny buttons for the reset switch and they are, in fact, tiny.
=> More informations about this toot | More toots from yantor3d@mas.to
Field test two of the hub battery was also my first day biking to work in the new year with a high in the low 30s and fog.
A wire came loose in the front battery during morning commute. But the rear battery held up for the whole 26 mile round trip.
The vibe of riding with the wheel lights on at night is like the HUD indicator in co-op video games that highlights your allies.
[#]bikeTooter
=> More informations about this toot | More toots from yantor3d@mas.to
The more I spin my wheels on it, the more I realize why this sort of things doesn't exist as commercial product.
A 6' LED strip with an RF mini controller connected to a 12V S3 battery with a custom case has a frustratingly high number of supply chains; up to five if you're shopping around for the best prices: a print farm, a battery supplier, an electronics supplier, an LED supplier, and a hardware store.
=> More informations about this toot | More toots from yantor3d@mas.to
After repairing the front wheel battery I've identified a second fault - the Keystone Electronics 247 battery clips have a bad habit of permanently bending to the point where they no longer contact the battery.
Nothing some folded up tin foil can't solve.
So my switch to the Keystone Electronics 209 leaf spring contacts is needful.
=> More informations about this toot | More toots from yantor3d@mas.to
Tonight's commute confirmed my suspicions that the current iteration of the hub battery I have on my bike is very susceptible to a hard shock, like going off a curb.
I'm not sure if something in the BMS can detect the drop and shut down, or if it jolted some of the wiring and caused a short.
Either way, the next iteration fully immobilizes the cells, and I now know to do a drop test.
=> More informations about this toot | More toots from yantor3d@mas.to
Whelp, any notion I had of trying to set up and sell full kits - batteries and lights - just went out the window.
The CC LED Glow Ride Plasma bicycle Wheel Lights Kit includes batteries, chargers, lights with remote for two wheels for $129.
The whole reason I went down the road of trying to build my own was that in their original product, they used zip ties to attach the battery to the hub, and a weird proprietary plug.
=> More informations about this toot | More toots from yantor3d@mas.to
Now that they're using standard 2.1mm plugs and an RF controller, their offering is basically what I've been building towards.
My battery has a larger capacity, though. So there is that.
=> More informations about this toot | More toots from yantor3d@mas.to
Thinking about it some more, it's not clear from the photos of their LED controllers are an inline dimmer switch or an RF receiver; given the photo doesn't show a remote, I'm going to assume the former.
=> More informations about this toot | More toots from yantor3d@mas.to
I have a hot date tonight.
With a soldering iron.
=> More informations about this toot | More toots from yantor3d@mas.to
For most of the prototyping phase I've been filling the gap between the battery case and the hub with closed cell foam, because it's squishy and doesn't require a precise fit. But it always felt like a temporary solution.
Now that I've settled on a case design (only took seven finals...) I want a more elegant solution.
So I've printed to spacers - one with standoffs, one with grip fins - to close the gap. We'll see which one works better.
Black PETG really doesn't photograph well...
=> More informations about this toot | More toots from yantor3d@mas.to
Something interesting I just noticed; 3/8" LED rope light costs, on average, two to four times what 1/2" LED rope light does.
Wat.
Still unable to find a place that sells 1/2" by the foot, though; the best I could do was somewhere that sells it in 10' increments.
=> More informations about this toot | More toots from yantor3d@mas.to
So what did you do tonight?
[#]bikeTooter #diy #electronics #3dprinting
=> View attached media | View attached media | View attached media | View attached media
=> More informations about this toot | More toots from yantor3d@mas.to
If you're wondering why the plugs are spliced in using wire nuts, it's because I'm not good at soldering multiple wires together.
=> More informations about this toot | More toots from yantor3d@mas.to
I've been researching wholesale suppliers for the components for a complete wheel light kit - consisting of a hub battery and a 6' length of LED rope light with an RF controller.
In theory, I could get the cost per unit for materials down to $30. But that doesn't factor in the cost of overseas shipping, and is vulnerable to tariffs.
Without the rope light, I can get all the components domestically*. But, I have never been able to find a source for 1/2" LED rope light by the foot.
=> More informations about this toot | More toots from yantor3d@mas.to
So it would be difficult to mass produce in a cost effective way; my benchmark is what CC LED charges for their wheel kit.
But the other side of the coin is that it feels wrong to release it open source without being able to cite a source for 1/2" dimmable LED rope light by the foot - best I know of is a company that sells it in 10' increments.
=> More informations about this toot | More toots from yantor3d@mas.to
Correction, there's a supplier in Arizona that sells it by the foot, but their listing doesn't say if it's dimmable. I'd have to reach out and ask.
And they sell a connector with pigtail ends? Amazing.
https://www.ledlight.com/led-rope-light-1_2-diameter.aspx
=> More informations about this toot | More toots from yantor3d@mas.to
I made a mistake while building the second battery - I accidentally melted one of the battery clip mounting arms with the soldering iron.
While a new one is I printing I can show the clever solution I came up with for the wires that connect adjacent cells and the B1 and B2 terminals on the BMS.
Rather than splice wires at the terminal, or at the battery contact, or with a third wire, I stripped the wire in the middle and soldered that.
[#]diy #electronics #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Good things come in pairs.
[#]bikeTooter #diy #3dprinting #electronics
=> View attached media | View attached media
=> More informations about this toot | More toots from yantor3d@mas.to
Assembling battery two and rebuilding battery one took all day because I made a few mistakes.
I melted one of the shells. I forgot to add enough slack to the wires to allow the case thirds to rotate the full 120° degrees needed to install them. At one point I literally had my wires crossed - with the wrong wire soldered to the B- and B1 terminals.
All fixable mistakes, but it highlights the fact that this is still a prototype and has many unknown unknowns.
=> More informations about this toot | More toots from yantor3d@mas.to
Good news - the latest iteration of the battery case passed the drop curb test - raising the wheel up about six inches off the ground and letting it drop. In prior iterations this would cause enough vibration that the BMS would shut off and enter safe mode for some reason.
Neither spacer ring worked in the rear wheel so I went with foam for now. The grip fin spacer is on the front wheel - we'll see how it does...
=> More informations about this toot | More toots from yantor3d@mas.to
The capitalist in me says I should productize this and sell it.
The communist in me says I should release it open source.
The realist in me says I should get a burrito.
=> More informations about this toot | More toots from yantor3d@mas.to
Wait, do I actually want the effort of running a small business or is the novelty of thinking about logistics a form of mental stimming?
=> More informations about this toot | More toots from yantor3d@mas.to
The latest field test was a success - the lights stayed on whole time, even with all the bumpy suburban stroads in Beaverton.
Now, four meters of 12V LED isn't much against the noontime sun on a clear day, but they make me visible when in shadows cast on the bike lane from buildings and trees. And if I use a flashing pattern I think it's more conspicuous than a steady glow.
Still, these are best suited for times without the sun.
[#]bikeTooter #diy #electronics #3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
I spent the rest of afternoon fiddling with a spreadsheet trying to estimate the cost of materials for making and selling hub batteries and wheel lights online.
Don't give me that look. I have weird stims, ok!?
=> More informations about this toot | More toots from yantor3d@mas.to
I have managed to locate domestic suppliers for all parts, at wholesale or at least resale prices. Even at materials plus time, these would cost quite a few slices of avocado toast.
But if I'm right about the longevity of rope light at 25% brightness, the lights could last for up to 100 hours. That's gotta be worth something to someone. And even at 25% brightness, twelve feet of LEDs are plenty conspicuous at night.
=> More informations about this toot | More toots from yantor3d@mas.to
But, I still think I could sell them cheaper than it would cost someone following a set of DIY plans to source all the parts individual, if only by virtue of not having to pay shipping for multiple orders from different stores.
You try finding a (US based) store that sells everything on the parts list at wholesale prices! Even mouser doesn't have evening I need.
=> More informations about this toot | More toots from yantor3d@mas.to
Trying to parse various rules on shipping products that contain lithium ion batteries. On top of everything else, that would be the biggest logical hurdle to selling these.
Since the cells will be contained in a product it might be as strait forward as adding a UN3481 label to the box?
I feel like I should just, y'know, email the various companies and ask them...
=> More informations about this toot | More toots from yantor3d@mas.to
I think that since this counts as a battery case, and the cells are < 20 Wh each and < 100 Wh in total, and can be packed with or contained in, then they can ship normally as long as they have a UN 3481 sticker - no "dangerous goods" surcharge.
But I've e-mailed UPS to clarify. I did not e-mail FedEx or USPS because I cannot find a contact e-mail on their website and, as a millennial, I am physically incapable of making a phone call.
=> More informations about this toot | More toots from yantor3d@mas.to
While I've been running the numbers to see what it would take to make these in small batches, past a certain threshold it would be better to contract a print farm to print a bulk order of the cases.
That throws a different set of wrenches into the soup because manual ejection is expensive. So I need to reduce the surface area of the first layer.
The go to strategy for this is to chamfer an edge and print the model at a 45° angle. But that's not possible with a model shaped like a bagel.
=> More informations about this toot | More toots from yantor3d@mas.to
So, I modeled a support block, and experimented with how many sprues it takes to hold things together while printing, but cleanly and easily separate after.
The first print at 45° had deformation in the screw holes - the slicer complained about overhangs a bit - so I'm trying again at 60°, which eliminated the overhang warnings.
With this support block, and a tapered cutout in the bottom, I've reduced the surface area of the first layer from ~4,000 mm^2 to ~500 mm^2
Good enough.
[#]3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Two unit tests, zero integration tests.
I've never had adhesion problems with my prints. Quite the opposite; they stick too well.
[#]3dprinting
=> More informations about this toot | More toots from yantor3d@mas.to
Printing with the model at a 45-60 degree angle is (allegedly) better for mass production because there's less surface area on the first layer and so it's easier to auto eject...
...but I notice that it takes almost twice as long.
=> More informations about this toot | More toots from yantor3d@mas.to
Showed a coworker the a photo of the wheel lights and he all but pre-ordered a set right then and there.
That's one, at least.
=> More informations about this toot | More toots from yantor3d@mas.to
Riding at night with the wheel lights on looks amazing in the reflection on darkened windows as I pass buildings.
I need to stop and get a picture next time.
=> More informations about this toot | More toots from yantor3d@mas.to
Well that's disappointing.
After checking literally every store I could find, I found one that sold short lengths of 1/2" 12V dimmable LED rope light.
It arrived today.
120V.
Maybe it's an honest mistake, but it's not a good look. And I liked their prices because they were selling at a little over $1 a foot - most other suppliers I found charge $2 a foot.
=> More informations about this toot | More toots from yantor3d@mas.to
Correction - I did not test the rope light properly. It is 12V, just in a 120V box for some reason.
=> More informations about this toot | More toots from yantor3d@mas.to
The list of tools I've used while developing these wheel lights now extends into my sewing kit.
An awl works on plastic, too.
=> More informations about this toot | More toots from yantor3d@mas.to
Maybe I should just go all in and invent a bike wheel that has the lights and battery built in...
=> More informations about this toot | More toots from yantor3d@mas.to
My goal for this weekend is to get the LED rope light cut to length and wired up. Once that's done I'll have two sets of wheel lights to do A/B comparisons with.
It's been a struggle to find LED rope light that is "dimmable" ... but I've been reading up on PWM (pulse width modulation) and it is my understanding that any LED strip or rope light is dimmable using PWM since it toggles lights on and off at a high frequency to give the appearance of dimming?
[#]diy #electronics
=> More informations about this toot | More toots from yantor3d@mas.to
I should also test the relative visibility of different lengths of LEDs on the wheels.
I figure 12", 24", 36", and 72" are good lengths to try, as it takes about 72" of LEDs to encircle the rim of a bike will.
=> More informations about this toot | More toots from yantor3d@mas.to
Well, we learned some things today.
1/2" rope light is barely able to fit; 3/8" is better but it retails at like $9 a foot.
1/2" rope light is not as bright as HD LED strips in a neon diffuser. I'm not an electrician but I suspect it is because it's 40 LEDs per meter as opposed to 120 LEDs per meter.
We'll ride with it for a week or two, see how it feels compared to the "neon" LED strips.
[#]bikeTooter #diy #electronics #3dprinting
=> View attached media | View attached media
=> More informations about this toot | More toots from yantor3d@mas.to
...now I'm wondering if I forgot to turn off one of the lights in the garage when I took the photo of the neon LED strips, or if my phone decided to auto balance the lighting for me...
=> More informations about this toot | More toots from yantor3d@mas.to
The sun has set. Time for a test ride.
=> More informations about this toot | More toots from yantor3d@mas.to
A three hour bike ride around Portland in mid January on a Saturday night.
Yes, I'm single, how could you tell?
I noticed the lights were showing signs of voltage drop by the time I got home, which is odd. I'll bench test the lights tomorrow and see if these LEDs are less efficient than the other rope light I used earlier.
[#]biketooter
=> More informations about this toot | More toots from yantor3d@mas.to
It's Sunday, so I'm recharging the small accessory lights on my bike and helmet.
And I discovered the charging port on my rear light is bent and won't accept a USB plug anymore. So now it's a $30 paper weight.
I might be able to tear it apart and repair or replace the charging port, but I would have to 3D print a new case for it.
=> More informations about this toot | More toots from yantor3d@mas.to
I've been going round and round with trying to decide what kind of lights to use.
"Neon" LED strips - an LED strip in a silicone diffuser - has the best look. But it's hard to find top bend, three sided diffusers, and 120 / meter strips draw a lot of power.
Rope light - LED bulbs in series in a clear PVC tube - has the best runtime. But it's nigh impossible to find in 3/8", and 1/2" is barely able to fit in between the spokes.
So I decided to take another look at boat/motorcycle lights
=> More informations about this toot | More toots from yantor3d@mas.to
The two best options I've are from Plash Lights and LED Glow.
Plash has both neon LED strips that are only .39" (10mm) thick and regular LED strips that are .49" (13mm) thick - both rated IP68 and available at various lengths.
A 6' length is almost $70.
https://plashlights.com/products/mini-led-neon-flex-rgb
LED Glow has "million color motorcycle kits" that you can buy "replacement" strips for - including a 6' length, the right size for a bike wheel.
A 6' length is about $40 - free shipping at $50
https://www.ledunderbody.com/replacement-strip-for-advanced-million-color-motorcycle-kits/
=> More informations about this toot | More toots from yantor3d@mas.to
Both would be amazing options for bicycles... provided you:
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d So the LEDs don’t use 120V…. Where’s the step-down happen? (And could you cut that part out?).
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@GeekAndDad Turns out, I am an idiot. The box says 120V, but I re-tested them and they are, in fact, 12V. Derp.
They still forgot the connector pins.
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d great news!
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@yantor3d 💡
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@yantor3d 4 lithium cells (3.7V) should work?
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@GeekAndDad Three, actually - their maximum voltage is 4.2.
And this whole thread has been about building a custom bike wheel hub battery pack to hold three 21700 cells.
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d Well cool!
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@yantor3d Nice but this jumped out at me
“Recommended installation for a boat is fully embedded in marine RTV silicone. You do not want it moving as over time something will break. Fully embedded in sealant and it will last 10+ years. The clips are fine for applications where there is no vibration or movement.”
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@GeekAndDad Good thing I won't be using clips, eh?
In my experience, jamming rope light or neon LEDs up between the spokes along the rim is very secure.
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d One thing about the generic rope lights and whatnot that I found myself wondering about when sleep eluded me last night is UV resistance of the tube/plastic parts. The motorcycle and boat lights ought to have that consideration already factored in during material selection (hopefully). Sorry if I w already mentioned this (shingles vaccine has my brain more scrambled than usual today 🤪).
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@yantor3d Should use less energy with fewer LEDs per foot though which seems good?
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@GeekAndDad that is the theory. When I bench tested the rope light several weeks ago it lasted about 27 hours on a single charge - the LED strip lasted 6 or 7.
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d 🤔
=> More informations about this toot | More toots from GeekAndDad@mastodon.social
@yantor3d can you set them to amber not white? White is for front, red for rear, amber for sides.
=> More informations about this toot | More toots from mjr@masto.bike
@mjr No, these are single color LEDs. If I had RGB LEDs I could set them to different colors. Or I could get single color LEDs in amber.
Let me guess, there are strict laws about bike lights in your country?
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d yep. Also strong tradition of those light colours/directions, to the point where showing white to the back attracts unwanted attention, and even joggers put reds to the back
=> More informations about this toot | More toots from mjr@masto.bike
@mjr Understandable. This is just a prototype; I've been going with what I can find. If I ever release this commercially or open source I'll make sure to acknowledge tradition.
=> More informations about this toot | More toots from yantor3d@mas.to
@yantor3d the two ways to control led brightness is pwm where the human eye can't perceive flashing faster than ~200 Hz (aka persistence of vision) and (constant) current control, but the latter is hard unless you use a dedicated chip.
The problem with pwm is that it doesn't photograph well. Long exposure times show the lights as a series of dashes.
Now I wonder how camera only navigation systems deal with that, especially since most light sources flicker!
=> More informations about this toot | More toots from trouble@masto.ai This content has been proxied by September (ba2dc).Proxy Information
text/gemini
