v3 features: 1) add a sensor to detect a battery in the hopper - use that to trigger the motor to run - motor runs for a few loops and if no detection of the charger - stops. - allows you to just always leave it there and on - doesn't do anything until it is needed - then when you drop a battery in, it runs and charges it. 2) add multi-polarity support / simplest way would be 2 relays - one which allows voltage measurement - so when the battery touches, voltage detected, and polarity as well. then second relay energized to change polarity if required / then first one to switch over to charger. - same proc you currently have - you can then dump them any which way - wont matter. There is probs a better not relay way to do this. 3) bigger hopper space on bottom and top for the cells. - more cells 4) custom pcb (see if jlc/pcbway will sponsor you for this :D ) - and move the PCB to the bottom as additional weight when empty and out the way/neat. - add a small fan that switches on when charging as well just to help out the charger IC 5) change to DC motor to simplify circuitry. - move the motor to the back with a belt (or even inset it a bit so it is hidden) - minimize size For the charger IC - they are all pretty common, so you can just get the IC without the circuit board and then you can tap into it directly - it will probs also give you more information like state etc... So you could even add a little LED on the front to show it's states.
there is some nice ideas here, only thing I would add is you can potentially get away with just having 2 sets of contacts oriented in reverse to each other if reverse polarity protection is a thing on the charger module (or if you design your own) i.e. the motor rotates the battery to 1st set of contact, if reverse polarity protection are triggered (i.e. not charging) than just move to the 2nd set of contacts
Re 1) You can add IR diode & receiver - that way you can (a) align the hole to the hopper so it waits in "open" position and (b) you can immediately resume operation when battery drops in. I'd say that the drum either should be a little more "funnel shaped" so the battery has time to fall in - or better yet you should wait a little bit when the opening is up because this way you can easily pinch the battery and possibly rip your motor off... Or shear the gears. 6) Add continuous loop so you can store the batteries inside & keep them charged 😁 NGL, the electronics part was a bit painful (poor D1 mini 😅) for a person that desings circuits, but love your design - nice one 😎
May I suggest to add: 5) A thermistor for the battery, if it get past a certain temperature stop charging to avoid possible fire hazard 6) Charged batteries can can back inside and be recharged automatically:)
love your point about uncounciously finding a solution when given time. theres a lot to be said for sleeping on a problem especially if you hit a mental block
Actually, an arduino is an overkill too. The LED on the charger, is enough to drive a FET to run a DC motor, but it would then keep turning the motor when the hopper is empty, but that could be meidated with photo transistor.
@@Xanthopteryx I am not a fan of "more fun" if you want stuff to work reliable over a long time, keep it simple. How do you even want to implement a Hall-Sensor as far as I know they can only detect magnetic field, what a battery doesnt have?
I use a bunch of these IKEA batteries for my Xbox Series X controllers, and I love the idea of modding this design into a miniature Xbox that charges batteries for the giant Xbox 🥰 One thought to avoid having to glue the stepper motor mounts on - could you just integrate them into a mounting plate with thru-holes lined up to the existing threaded inserts on the main body?
That's not terrible, i like it. I also actually like that you went ahead and just handwired it because honestly too many vids are just PCB manufacturer ads.
@@Mellow_labs I think the point must be explained in an email from pcbway, because 'tubers always seem to realise how much better pcb's are after they get their sponsorship offer 😆
Honestly in the film industry we use a crazy amount of rechargables which all get left too long in their charges. Imagine being done with your shoot, instantly dumping the empty ones in this system and it just dumping them out when done, it truly is genius! I can see this being a product in the future.
Instead of having the V1 servo hold pressure on the contacts it could release a spring that keeps the contacts on and when powered can compress the spring. Kind of like truck air brakes.
Heh, my first thought on how to improve V1 would be (like you thought) a rotating drum, but I would've solved the connection issue with a normally-extended solenoid rather than a servo, although the solution with the compliant mechanism is fairly clever. Of course, like other comments have stated, making it able to be left on, and have some sort of sensor in the input hopper would be ideal (as well as some bits to disallow sticking foreign objects in the drum), and for the output, I'd imagine making the bottom open so you can set it on top of variable size output hoppers would be a good idea, so if you don't need much space, you can choose a smaller one, or if you have a lot of batteries to store, a larger one.
That's a major design flaw. Either design your magazine as to not allow the battery to be inserted incorrectly using a keyway. Or add a polarity switch. @@Mellow_labs
I hate how the motor is sticking out on the side ^^ Maybe you can fit the motor inside of the spinning barrel? Or use a belt? (Could be 3d printed) I really like the compliant mechanism! Maybe add some status leds on the front. Once the "collector" is full, the device will have to stop charging. Maybe you can add a button which you press when you take out a battery to tell the device to charge another battery (or two if you press twice)
Great work. Now make it fool proof by installing something like a volt meter, thats reads the current and 2 charging positions, so it does not matter, how you put it in. And also add a switch, so it turns of, when there is no battery in it
spring load the pins (like in regular battery chargers) and then just have a regular motor with a small plastic mechanism that compresses the spring and pushes the charged battery out in a timed motion, once the next battery falls in, release the spring and let it charge, much less need for complex electronics
A thought I had other than checking polarity is to also check whether the battery is a dud. If so then have it put into a bad battery hopper. This way you only have known good batteries coming out.
Now add barcode or similar to the batteries so they have individual IDs, and then use charge statistics and analyze statistics to send to your computer via bluetooth so you can see stats for each battery, like charge time, health and more.
I’ve had that kind of issue with esp’s when they don’t get enough low impedance power. I have a separate 5.1V power supply when programming, and like 200uF of capacitance on the 3.3v rail and now I rarely have issues
About that stepper motor just buzzing with the D1 mini - it looks like only 3 of the 4 LEDs on the driver board are lighting up, so that could indicate something's not working with the 4th GPIO pin chosen, which would result in only 1 out of 2 coils in the stepper being energized, which would definitely just make it buzz without progressing the rotation. Maybe that pin wasn't actually set as an output? Also, at 9:06 it looks like you're missing another wire to share the ground between the D1 mini and the driver board, which could also cause weird issues like this. Awesome project btw!
Crazy idea hat on: use a bimetal temperature switch as the charging contact that applies pressure and another that act as the barrier for the next battery. Use resistor to heat trigger them to open. So when the batteries is approaching its nominal charge a resistor near the bimetal contact engages heating it and releasing tension. Then when the battery falls the resistor losses its nominal voltage and if there is no load o. The charger engages the upper resistor to bend the bimetalic strip connected to the barrier. No motors all electrical. Crazy hard tolerances to hold batteries so don't do it horizontally rather do it at 10-15 degs off vertical so it can fall but its not fighting gravity. If you need to bounce ideas around on how to do it talk to the classic pinball community. That hardware if filled with stuff like that.
Interesting to say the least. And some times things not working as expected makes you learn so much more to fix the problem(S). Now that you have that working, make it automated to charge the batteries even if you put them in the hopper the wrong way. Check the polarity and if the battery is reversed, swap the charging polarities around. Than no need to watch how the hopper is loaded... Just an idea to further the automation...
Possible problems with the ESP: - You may be using the same pins as the UART port. - Your serial to USB converter is not well connected and it can't reset and put the ESP in boot mode correctly. For the second one, short pin 0 and GND, reboot and it will be ready to flash.
So the ESP‘s do that, even if they’re brand-new and nothing is connected to them, and only when it comes to installing micropython, something like WLED and ESPhome work fine
@@Mellow_labs which wemos d1 mini (or clone) do you use? I had somewhat similar issues with a lot of the newer d1 minis which use the newer ESP8266 EX chip. It claims to be 100% hardware and software compatible, but I tried to use them with multiple bigger projects (GBS-C scaler and wled) and they did not work at all. Some minor projects with arduino worked fine though. I then ordered some d1 minis which use another ESP8266 versions and they worked fine. Maybe you're using the ex version as well? Wouldn't be surprised if they don't work with micropython as well.
The first idea for the new version would have worked if you rotated the terminal (of the connector) by what ever degrees would be required for it to slide onto it instead of directly 90° into the strip.
I wonder if it would be an improvement to add what i think is called a Geneva mechanism, like they have in movie projects that use film, to add a mechanical way to put the battery in the same place every time easily. that (or at least something mechanical) was my first thought for how to control the batteries' movement through the charging section
Brillant. Love how you keep finding ways that it could have been done more simply, too. BTW, I think the D1mini has a divider on the analog input already, so I think it's set to divide down from 3v3 (but check me on that, it might be 5v). You could have used esphome instead of micropython, but then you might be tempted to make it much more complicated....
It does, and it’s not clearly stated in the documentation I found out the hard way after making my own divider and trying to figure out why I’m not getting the correct values.
@@Mellow_labs haha yeah, plus if one happens to be buying clones (as I may have done from time to time) then who knows what they might change from the original!
Just found your channel and I have a few minor suggestions: 1. I noticed your room lights flickering in a couple of your closeup shots. I recommend that you film in 25 or 50fps. to avoid that. Or even better, make your own lighting system and use constant current drivers for the LED’s. 2. When you film over-the-shoulder shots, always lock the focus on what you are working on. That way, the autofocus won’t hunt between your shoulder and the thing you are working on. (You could also benefit from choosing a higher F stop on the camera. This will allow more “wiggle room” in the ‘in-focus’ distance.) I look forward to seeing more videos from you! Cheers from Denmark Bo
Instead of having the springy plastic, just orient the terminals so the battery slides on the spring (a bit like most AA battery holders, but two sides?)
this comment has 2 functions: first it drives engagement 🙂 second, you said you like the soldering of the prototype boards, can you provide some hints or a dedicated video on how to do that properly? That is something that always hinders my projects, I have no idea how to do them properly, BUT i have to confess that I did not look that far into it. So maybe that is an idea, or not.
for the standoffs, a way better approach would be to just add thread inserts to the plate and screw the stepper in with a not glued standoff and screw long enough to reach the threaded insert
Slick. Photographers might be interested in this. You could use the drum assembly, rotating one way then the other, to distribute a hopper of cells to two of these to increase throughout.
great video and awesome idea but as it only charges 1 battery at a time its pretty limited to the usefulness,. what if you adapted itso that the next version could charge multiple batteries at a time as most things require 2+ batteries these days so it would be a lot more useful. maybe add a tiny screen to show the amount of charge in the batteries.
You need a detector for the hopper, so the motor dont spin indefinately when its empty. also you should watch some videos of micro soldering. It will help alot soldering like they do with flux, solder joints will be much prettier and you save alot of time when de-soldering and cleaning off PCB's. Its quite easy really. And get solder with lead in it, its much better suited for home soldering as the lead free stuff takes to high temperatures, hardens to fast and is much more fragile. You are not mass producing stuff so no need to respect RoHS.
D1-Mini clones from AZDelivery that I have contains that voltage divider from 3v3 to 1v. That usb reconnecting issue maybe related to power delivery. esp booting itself shouldn't reset usb serial chip.
dude... i wanna teach you how to design and manufacture PCBs because that soldering hurts my eyes (no offense) also... stop using those green PCBs, i had tons of problems with them because they have a very poor isolation resistance sometimes. i recommend proper FR4 boards
Eee, looks like weak power on USB port. It turns on, takes too much current and resets. Please take out those boards from trash and maybe try to power it differently, like GMD, D-, D+ from USB and GMD and VCC from power bank or charging brick (but first make sure there is no voltage across grounds)
Instead of using any kind of microcontroller, can't you just wire that light wire to a relay or transistor that powers a slow dc motor to turn the drum? It really is just an over glorified limit switch anyway. Just need to invert it.
@@Mellow_labs If you want it smarter, it needs a charge minder on the ones it's already charged so they don't self discharge if you don't use them immediately. That would be something worth using intelligence for.
A better way to look at this is to consider the charger as an always-on background device. This way, you can have a surplus of batteries at home, ready to be swapped in and out of the charger as needed. Unlike traditional chargers where you can charge multiple batteries simultaneously, but need to constantly monitor and swap them once they’re charged, this system is more convenient and requires less supervision. At least that’s the idea, might need a couple more revisions 😂
Great idea well executed. Reminded me of downfall. Ive seen that with the D1 mini before and had to replace the diode on it and it was back to normal. Can't say what caused it. Also what happens when there are no batteries left? Those steppers have a specific number of steps for a full turn so you could set it so that it does a full turn and if no charging go to sleep until a button is pressed or something. Just to save power.
@@Mellow_labs Could also be a wrong flash mode used by the programming software. I had a similar issue with ESP8266. When changed the baudrate, I saw crashes and exeptions reporting from the esp. After changing the flash mode to "dout", all worked normally again. Google for "ESP8266 flash mode dout crash"
the frequency for the stepper motor is way to high, then they wont turn. or the wires are in the wrong order... yo dude, just hmu i f you need some help
Are you using genuine D1 mini's? Genuine ones have a voltage divider on analogue pin so can handle 3.3v. Are you trying to program it "in circuit"? I had issues some time ago where I'd actually used one of the pins required for programming, or one of the pins pulled high by default.
I found out recently, there was a kickstart back in 2016, that didn’t get funded, but this is basically what they made. So I’ll just stick to making it an open source project.
This is awesome! Its annoying when things don't work and after spending tons of time its hard to pivot but I'm glad that setbacks didn't slow you down! Looking forward to the next video!
flashing the ESP with micropython, that cable looks funky.. i would suspect that as a possible culprit, other than that, my experience with micropython on ESP is limited.
given that you can indefinately trickle-charge Nickle-Cadnum/Nickle-Alloy batteries, one can make an even simpler setup for them, provided you didn't mind it taking a very long time for them to charge. this setup is faster, tough. but an 18650 version.... MMMMM, that gives me ideas. if you're going to build an E-bike battery, you're talking 50+ cells, all needing to be charged. could get rediculously fancy and stick lables on them too, noting capacity or bank # for balancing. even though that'd be insanely complicated.
When flashing a microcontroller you need to have a properly powered USB hub otherwise it will not have adequate power to flash, and will give random read write errors.
Ypu actually don't need a microcontroller and stepper motor at all. This could all be done with a handful of discrete components and a plain dc motor. Cool design.
Hey, thanks for your comment! I totally wish I could share everything for free. It’s just that making these videos costs more than you’d think, and the ad money really doesn’t cover it. I’m doing this without any sponsors. My dream is to go completely open source with my projects, but I’m not quite there financially. Right now, the best I can do is offer some perks to my Patreon supporters or sell some of my work. But hey, I promise, I’m always working on improving, and anyone who buys now will get all future upgrades for free!
Neat! I wonder if it would be simpler to have the charger along a drop instead of in a barrel. Have a stopper at the top and bottom of a channel, and then charing pins that clamp in from the sides to charge 2-4 batteries at a time. When charging is complete, disengage the charging pins, open the bottom gate to clear the channel, close the bottom, open the top to load the next set, close the top, and start the next charge cycle. Also,absolutely agree about soldering. Sure, you aren't supposed to breathe the fumes... But it is nerd incence and takes me to my happy place of doing projects with my dad as a kid.
I would probably add a board that supports charging in either polarity, also i really like the fact that this whole thing could be the charging circuit, a relay and a geared motor, instead of the microcontroller.
v3 features:
1) add a sensor to detect a battery in the hopper - use that to trigger the motor to run - motor runs for a few loops and if no detection of the charger - stops.
- allows you to just always leave it there and on - doesn't do anything until it is needed - then when you drop a battery in, it runs and charges it.
2) add multi-polarity support / simplest way would be 2 relays - one which allows voltage measurement - so when the battery touches, voltage detected, and polarity as well. then second relay energized to change polarity if required / then first one to switch over to charger. - same proc you currently have
- you can then dump them any which way - wont matter. There is probs a better not relay way to do this.
3) bigger hopper space on bottom and top for the cells. - more cells
4) custom pcb (see if jlc/pcbway will sponsor you for this :D ) - and move the PCB to the bottom as additional weight when empty and out the way/neat. - add a small fan that switches on when charging as well just to help out the charger IC
5) change to DC motor to simplify circuitry. - move the motor to the back with a belt (or even inset it a bit so it is hidden) - minimize size
For the charger IC - they are all pretty common, so you can just get the IC without the circuit board and then you can tap into it directly - it will probs also give you more information like state etc...
So you could even add a little LED on the front to show it's states.
These are really good ideas and I might try to make my own "improved" version and make it available for free.
there is some nice ideas here, only thing I would add is you can potentially get away with just having 2 sets of contacts oriented in reverse to each other if reverse polarity protection is a thing on the charger module (or if you design your own)
i.e. the motor rotates the battery to 1st set of contact, if reverse polarity protection are triggered (i.e. not charging) than just move to the 2nd set of contacts
Even without sponsorship JLC is dirt cheap for small boards. I can get 5pcs 100x100mm delivered for less than 5 euro...
Re 1) You can add IR diode & receiver - that way you can (a) align the hole to the hopper so it waits in "open" position and (b) you can immediately resume operation when battery drops in. I'd say that the drum either should be a little more "funnel shaped" so the battery has time to fall in - or better yet you should wait a little bit when the opening is up because this way you can easily pinch the battery and possibly rip your motor off... Or shear the gears.
6) Add continuous loop so you can store the batteries inside & keep them charged 😁
NGL, the electronics part was a bit painful (poor D1 mini 😅) for a person that desings circuits, but love your design - nice one 😎
May I suggest to add:
5) A thermistor for the battery, if it get past a certain temperature stop charging to avoid possible fire hazard
6) Charged batteries can can back inside and be recharged automatically:)
love your point about uncounciously finding a solution when given time. theres a lot to be said for sleeping on a problem especially if you hit a mental block
Actually, an arduino is an overkill too. The LED on the charger, is enough to drive a FET to run a DC motor, but it would then keep turning the motor when the hopper is empty, but that could be meidated with photo transistor.
Would achive that with a timer and a self holding relay. Basicly when the motor runs to long the timer shut everything down.
@@jonasstahl9826 But you want it to start when adding a battery.
@@Xanthopteryx You can activate the selfholding relay with a switch, either you push it manual or put it in the hopper so the battery push it.
@@jonasstahl9826 More fun to use something like a hall sensor or similar.
@@Xanthopteryx I am not a fan of "more fun" if you want stuff to work reliable over a long time, keep it simple.
How do you even want to implement a Hall-Sensor as far as I know they can only detect magnetic field, what a battery doesnt have?
I use a bunch of these IKEA batteries for my Xbox Series X controllers, and I love the idea of modding this design into a miniature Xbox that charges batteries for the giant Xbox 🥰
One thought to avoid having to glue the stepper motor mounts on - could you just integrate them into a mounting plate with thru-holes lined up to the existing threaded inserts on the main body?
That's not terrible, i like it.
I also actually like that you went ahead and just handwired it because honestly too many vids are just PCB manufacturer ads.
I’m so glad you said that 😂 I don’t see the point of doing PCBs for some thing, that is one of one
@@Mellow_labs I think the point must be explained in an email from pcbway, because 'tubers always seem to realise how much better pcb's are after they get their sponsorship offer 😆
Honestly in the film industry we use a crazy amount of rechargables which all get left too long in their charges. Imagine being done with your shoot, instantly dumping the empty ones in this system and it just dumping them out when done, it truly is genius! I can see this being a product in the future.
Instead of having the V1 servo hold pressure on the contacts it could release a spring that keeps the contacts on and when powered can compress the spring. Kind of like truck air brakes.
Heh, my first thought on how to improve V1 would be (like you thought) a rotating drum, but I would've solved the connection issue with a normally-extended solenoid rather than a servo, although the solution with the compliant mechanism is fairly clever.
Of course, like other comments have stated, making it able to be left on, and have some sort of sensor in the input hopper would be ideal (as well as some bits to disallow sticking foreign objects in the drum), and for the output, I'd imagine making the bottom open so you can set it on top of variable size output hoppers would be a good idea, so if you don't need much space, you can choose a smaller one, or if you have a lot of batteries to store, a larger one.
What happens when someone loads the battery with the wrong polarity?
Current system doesn’t support it so don’t do it 😅
That's a major design flaw. Either design your magazine as to not allow the battery to be inserted incorrectly using a keyway. Or add a polarity switch.
@@Mellow_labs
Yeah I know I’ll get around to it.
Surely could add a circuit to just not charge if that happens, maybe activate an LED
@@fredpinczuk7352it is safe to assume, if you can build this on your own, you know what you are doing.
I hate how the motor is sticking out on the side ^^
Maybe you can fit the motor inside of the spinning barrel? Or use a belt? (Could be 3d printed)
I really like the compliant mechanism!
Maybe add some status leds on the front.
Once the "collector" is full, the device will have to stop charging.
Maybe you can add a button which you press when you take out a battery to tell the device to charge another battery (or two if you press twice)
You could also just measure the LED voltage with an OPAMP instead of an µC and switch the motor with that (and a transistor/fet).
Great work. Now make it fool proof by installing something like a volt meter, thats reads the current and 2 charging positions, so it does not matter, how you put it in. And also add a switch, so it turns of, when there is no battery in it
I have a nice 30W solar car battery tender, and I started using it in a project to charge my 18650s for free.
This is so genius! Something like this should be available to purchase for not so tech savvy people..
spring load the pins (like in regular battery chargers) and then just have a regular motor with a small plastic mechanism that compresses the spring and pushes the charged battery out in a timed motion, once the next battery falls in, release the spring and let it charge, much less need for complex electronics
A thought I had other than checking polarity is to also check whether the battery is a dud. If so then have it put into a bad battery hopper. This way you only have known good batteries coming out.
a dud?
Now add barcode or similar to the batteries so they have individual IDs, and then use charge statistics and analyze statistics to send to your computer via bluetooth so you can see stats for each battery, like charge time, health and more.
I’ve had that kind of issue with esp’s when they don’t get enough low impedance power. I have a separate 5.1V power supply when programming, and like 200uF of capacitance on the 3.3v rail and now I rarely have issues
That's a cool product, and it looks like an old-fashioned matchbox holder.
That's amazing. You consistently make these awesome projects. This is one I have to make
About that stepper motor just buzzing with the D1 mini - it looks like only 3 of the 4 LEDs on the driver board are lighting up, so that could indicate something's not working with the 4th GPIO pin chosen, which would result in only 1 out of 2 coils in the stepper being energized, which would definitely just make it buzz without progressing the rotation. Maybe that pin wasn't actually set as an output? Also, at 9:06 it looks like you're missing another wire to share the ground between the D1 mini and the driver board, which could also cause weird issues like this. Awesome project btw!
No, I did have the code run really slowly flashing each of the four LEDs
Crazy idea hat on: use a bimetal temperature switch as the charging contact that applies pressure and another that act as the barrier for the next battery. Use resistor to heat trigger them to open. So when the batteries is approaching its nominal charge a resistor near the bimetal contact engages heating it and releasing tension. Then when the battery falls the resistor losses its nominal voltage and if there is no load o. The charger engages the upper resistor to bend the bimetalic strip connected to the barrier. No motors all electrical. Crazy hard tolerances to hold batteries so don't do it horizontally rather do it at 10-15 degs off vertical so it can fall but its not fighting gravity. If you need to bounce ideas around on how to do it talk to the classic pinball community. That hardware if filled with stuff like that.
You could make it slightly more efficient by having multiple slots in the barrel, that way it'd only need 1/3rd of a turn for each step.
Interesting to say the least. And some times things not working as expected makes you learn so much more to fix the problem(S). Now that you have that working, make it automated to charge the batteries even if you put them in the hopper the wrong way. Check the polarity and if the battery is reversed, swap the charging polarities around. Than no need to watch how the hopper is loaded... Just an idea to further the automation...
Possible problems with the ESP:
- You may be using the same pins as the UART port.
- Your serial to USB converter is not well connected and it can't reset and put the ESP in boot mode correctly.
For the second one, short pin 0 and GND, reboot and it will be ready to flash.
So the ESP‘s do that, even if they’re brand-new and nothing is connected to them, and only when it comes to installing micropython, something like WLED and ESPhome work fine
@@Mellow_labs Oh I see. In that case I would have rewrote the thing in C since it's fairly simple.
@@DiThi Addition to this. If you don't want to use official tool chain you can also use arduino ide.
@@LimbaZero I use PlatformIO
@@Mellow_labs which wemos d1 mini (or clone) do you use? I had somewhat similar issues with a lot of the newer d1 minis which use the newer ESP8266 EX chip. It claims to be 100% hardware and software compatible, but I tried to use them with multiple bigger projects (GBS-C scaler and wled) and they did not work at all. Some minor projects with arduino worked fine though. I then ordered some d1 minis which use another ESP8266 versions and they worked fine. Maybe you're using the ex version as well? Wouldn't be surprised if they don't work with micropython as well.
Slick.
I'll just keep manually loading my 12-gang battery charger in 30 seconds and enjoy 12 times faster charges
The first idea for the new version would have worked if you rotated the terminal (of the connector) by what ever degrees would be required for it to slide onto it instead of directly 90° into the strip.
I wonder if it would be an improvement to add what i think is called a Geneva mechanism, like they have in movie projects that use film, to add a mechanical way to put the battery in the same place every time easily. that (or at least something mechanical) was my first thought for how to control the batteries' movement through the charging section
you need a very specific driver to get the d1 mini working, and you need to reinstall it everytime you restart your pc
Brillant. Love how you keep finding ways that it could have been done more simply, too. BTW, I think the D1mini has a divider on the analog input already, so I think it's set to divide down from 3v3 (but check me on that, it might be 5v). You could have used esphome instead of micropython, but then you might be tempted to make it much more complicated....
It does, and it’s not clearly stated in the documentation I found out the hard way after making my own divider and trying to figure out why I’m not getting the correct values.
@@Mellow_labs haha yeah, plus if one happens to be buying clones (as I may have done from time to time) then who knows what they might change from the original!
Pretty cool
Thank you 🙏
Just found your channel and I have a few minor suggestions:
1.
I noticed your room lights flickering in a couple of your closeup shots. I recommend that you film in 25 or 50fps. to avoid that. Or even better, make your own lighting system and use constant current drivers for the LED’s.
2.
When you film over-the-shoulder shots, always lock the focus on what you are working on. That way, the autofocus won’t hunt between your shoulder and the thing you are working on.
(You could also benefit from choosing a higher F stop on the camera. This will allow more “wiggle room” in the ‘in-focus’ distance.)
I look forward to seeing more videos from you!
Cheers from Denmark
Bo
Thanks, I’ll look into better lighting. But not much I can do about the camera at the moment. It’s either my phone or my WebCam.
Instead of having the springy plastic, just orient the terminals so the battery slides on the spring (a bit like most AA battery holders, but two sides?)
1:35 my brain works the same
I think i found Zack Freedman's long lost cousin
this comment has 2 functions: first it drives engagement 🙂
second, you said you like the soldering of the prototype boards, can you provide some hints or a dedicated video on how to do that properly?
That is something that always hinders my projects, I have no idea how to do them properly, BUT i have to confess that I did not look that far into it.
So maybe that is an idea, or not.
for the standoffs, a way better approach would be to just add thread inserts to the plate and screw the stepper in with a not glued standoff and screw long enough to reach the threaded insert
But the side would have to be a lot thicker.
Slick. Photographers might be interested in this. You could use the drum assembly, rotating one way then the other, to distribute a hopper of cells to two of these to increase throughout.
great video and awesome idea but as it only charges 1 battery at a time its pretty limited to the usefulness,. what if you adapted itso that the next version could charge multiple batteries at a time as most things require 2+ batteries these days so it would be a lot more useful. maybe add a tiny screen to show the amount of charge in the batteries.
Looks pretty interesting. Subscribed
Thank you 🙏
You need a detector for the hopper, so the motor dont spin indefinately when its empty.
also you should watch some videos of micro soldering. It will help alot soldering like they do with flux, solder joints will be much prettier and you save alot of time when de-soldering and cleaning off PCB's. Its quite easy really.
And get solder with lead in it, its much better suited for home soldering as the lead free stuff takes to high temperatures, hardens to fast and is much more fragile. You are not mass producing stuff so no need to respect RoHS.
another V3 feature could also be to see if a battery is dead? not sure if you can detect that
Hey Tomasz! Nice to have found your channel❤ Fajne filmy!
Hey, thanks!
if this was a product at a good price I would buy it
D1-Mini clones from AZDelivery that I have contains that voltage divider from 3v3 to 1v.
That usb reconnecting issue maybe related to power delivery. esp booting itself shouldn't reset usb serial chip.
This is designed to work only if the battery polarity is correct. How about making a new design that detects the reverse battery and flips it?
WIP 🚧
What in the world is that dorky thing on your glasses? (I like it! 😂)
th-cam.com/users/shorts1q7CwN8jOoY?feature=share
Well why, didn’t you just do a drop shot in stead of the drum.?
You should make one for lithium ion 18650s also what electric screwdriver do you use?
Look up makers muse (I think) he goes over how to design with sacrificial support. Or maybe it's teaching tech. Not sure but both are fun to watch
dude... i wanna teach you how to design and manufacture PCBs because that soldering hurts my eyes (no offense) also... stop using those green PCBs, i had tons of problems with them because they have a very poor isolation resistance sometimes. i recommend proper FR4 boards
8:30 looks like the baud rate is set wrong to me
Not baud rate, i tried them all same problem. But thx for the suggestion
Eee, looks like weak power on USB port. It turns on, takes too much current and resets. Please take out those boards from trash and maybe try to power it differently, like GMD, D-, D+ from USB and GMD and VCC from power bank or charging brick (but first make sure there is no voltage across grounds)
Instead of using any kind of microcontroller, can't you just wire that light wire to a relay or transistor that powers a slow dc motor to turn the drum? It really is just an over glorified limit switch anyway. Just need to invert it.
It seems most people want it to be smarter not dumber.
@@Mellow_labs If you want it smarter, it needs a charge minder on the ones it's already charged so they don't self discharge if you don't use them immediately. That would be something worth using intelligence for.
So what happens if you drop the battery in backwards?
For now, I don’t dare to try, but I will have something in the future to fix it
Did you desolder a socketed IC with the IC still socketed? Cool project though!
Yep 😎
@@Mellow_labs absolute mad lad
Hmm, compared to a 4 cell charger, it's not exactly fast serving them up one by one
A better way to look at this is to consider the charger as an always-on background device. This way, you can have a surplus of batteries at home, ready to be swapped in and out of the charger as needed. Unlike traditional chargers where you can charge multiple batteries simultaneously, but need to constantly monitor and swap them once they’re charged, this system is more convenient and requires less supervision. At least that’s the idea, might need a couple more revisions 😂
Great idea well executed. Reminded me of downfall. Ive seen that with the D1 mini before and had to replace the diode on it and it was back to normal. Can't say what caused it.
Also what happens when there are no batteries left? Those steppers have a specific number of steps for a full turn so you could set it so that it does a full turn and if no charging go to sleep until a button is pressed or something. Just to save power.
Yeah for now, it just kind of spins, it doesn’t really need to step a motor. I’ll make a list of improvements for the next version.
@@Mellow_labs Could also be a wrong flash mode used by the programming software. I had a similar issue with ESP8266. When changed the baudrate, I saw crashes and exeptions reporting from the esp. After changing the flash mode to "dout", all worked normally again. Google for "ESP8266 flash mode dout crash"
the frequency for the stepper motor is way to high, then they wont turn. or the wires are in the wrong order... yo dude, just hmu i f you need some help
What is that on your glasses?
th-cam.com/users/shorts1q7CwN8jOoY?feature=share
Are you using genuine D1 mini's? Genuine ones have a voltage divider on analogue pin so can handle 3.3v. Are you trying to program it "in circuit"? I had issues some time ago where I'd actually used one of the pins required for programming, or one of the pins pulled high by default.
No, I’m not using the genuine ones which is probably most of the problem. The ones I get use a ESP-12F which doesn’t have such luxuries.
Silly question, without the reed switch when the last battery has been done, won't it just keep rotating the drum until you turn it off?
Yes, but the reed switch wouldn’t have affected that
♥pride
> PCB board
(－‸ლ)
Instant "do-not-recommend-channel"
I only say it to make people mad 😂
@@Mellow_labs U've successfully made me mad, grats
I think I know what's wrong with your D1 minis.
They know you pronounce it djif and they refuse to work with someone who does that!
I guess it’s time to change my ways. From now on, I will be using raspberry Pi pico/s exclusively 😂😂
I see a potential future commercial product in this.
You can find similar products in Japanese market.
I found out recently, there was a kickstart back in 2016, that didn’t get funded, but this is basically what they made. So I’ll just stick to making it an open source project.
remove the dumb device from your glasses it's distracting
Then it’s gonna be a silent video
This is awesome! Its annoying when things don't work and after spending tons of time its hard to pivot but I'm glad that setbacks didn't slow you down! Looking forward to the next video!
i really love the Idea of a automatical battery charger.
Omg, ive been asking the world why this doesn't exist for YEARS
flashing the ESP with micropython, that cable looks funky.. i would suspect that as a possible culprit, other than that, my experience with micropython on ESP is limited.
could add a timeout in case battery is dead, and sort them into a faulty tray
given that you can indefinately trickle-charge Nickle-Cadnum/Nickle-Alloy batteries, one can make an even simpler setup for them, provided you didn't mind it taking a very long time for them to charge. this setup is faster, tough.
but an 18650 version.... MMMMM, that gives me ideas. if you're going to build an E-bike battery, you're talking 50+ cells, all needing to be charged. could get rediculously fancy and stick lables on them too, noting capacity or bank # for balancing. even though that'd be insanely complicated.
I solved similar esp issues by removing (bridging) the diode between usb5v and vreg.
When flashing a microcontroller you need to have a properly powered USB hub otherwise it will not have adequate power to flash, and will give random read write errors.
Ypu actually don't need a microcontroller and stepper motor at all. This could all be done with a handful of discrete components and a plain dc motor. Cool design.
You could have just used a relay instead of the raspberry
ikea charger does not overcharge and detects dead batteries.
This is a great design. Mechanical is more durable most of the time
Hab Nu schon Lust das nach mein eigenen Vorstellungen zu disingen xD
can you add the stl files in your printables plezz ?
Sorry, you have to purchase them from Patreon.
Would be great if you made it open source at some point
Hey, thanks for your comment! I totally wish I could share everything for free. It’s just that making these videos costs more than you’d think, and the ad money really doesn’t cover it. I’m doing this without any sponsors. My dream is to go completely open source with my projects, but I’m not quite there financially. Right now, the best I can do is offer some perks to my Patreon supporters or sell some of my work. But hey, I promise, I’m always working on improving, and anyone who buys now will get all future upgrades for free!
Neat!
I wonder if it would be simpler to have the charger along a drop instead of in a barrel. Have a stopper at the top and bottom of a channel, and then charing pins that clamp in from the sides to charge 2-4 batteries at a time.
When charging is complete, disengage the charging pins, open the bottom gate to clear the channel, close the bottom, open the top to load the next set, close the top, and start the next charge cycle.
Also,absolutely agree about soldering. Sure, you aren't supposed to breathe the fumes... But it is nerd incence and takes me to my happy place of doing projects with my dad as a kid.
I would probably add a board that supports charging in either polarity, also i really like the fact that this whole thing could be the charging circuit, a relay and a geared motor, instead of the microcontroller.
This is genius! Thanks for your work. Hope this channel rockets because your content is amazing!
Thank you 🙏 I’m glad you enjoyed it 😊
Bardzo dobrze się Cię ogląda, ciekawy projekt :)
heh heh heh, the nice small package skit needs to appear in every video :D
I have better ideas 😂 but I’m glad you liked it 🙏
but what if you put the battery wrong way in
Current system doesn’t support it so don’t do it 😅