you prolly dont give a shit but does someone know of a way to log back into an instagram account?? I was dumb lost the password. I appreciate any tips you can give me.
@Uriel Aden Thanks for your reply. I found the site on google and Im in the hacking process now. I see it takes a while so I will reply here later with my results.
Congratz dude, using the pcb like that something i wasn't expecting to work. I don't know jack about interactive systems but just seeing someone like you enjoying the pay-off is making my day.
Here's a thought, If you get the drive wheel PCB made with a matt black solder resist, then use a white silkscreen instead of slots. The way you should be able to get a much finer resolution on the indexing, which will be helpful for different sized SMDs.
If you use RS485 and put a DIP switch that sets the address of each feeder you can address each feeder over the same 3 wires. If you use CAN you get a few network type advantages. The only down side with CAN is a smaller selection of controllers
Great Design. I'm currently working on my own PNP. I am to the point where it's populating the board fairly accurately. Still need to set up the pipelines for up facing camera. I've recently ordered boards to give your feeder a try. Funny thing is, JLCPCB refused the indexing wheel board. Didn't give much reason except it was beyond their capability.
I don't think you need to use gold solder - just give it a very very very short sand on some very fine grit paper, sellotaped down to some float glass. That'll get her shining.
Wicked cool design! I am so excited that the design files will be released. One thing I thought of as a potential improvement to the rotating wheel is that instead of using those pads to reflect light, you could just use them like a switch. Just put a brush on a spring to make the electrical contact, The other part of the switch would be a circular pad where the shaft of the motor goes through the center of.
I m started to build a INDEX unit on your info. When looking to all the feeder design i think that this is rather a problem. After thinking I wonder why the control of the feeder must be done with your piggy board on yuour special protocol. ??? Now I think you are repace this by RS485 protocol. Now I have the idea , why not place a sensor in front of the feeder able to move a mechanisme to set the tape forward. A actuator can be mouted on the Z axle eg. IR or magnet (Hall sensor on feeder). In Openpnp you sent the head to feeder wanted, this response on the sensor activation and go!. I hope my idea is clear, sorry for perhaps the not correct English , regards, Cees from the Netherlands
I might have tried Reverse mounting an LED to the main board behind the indexing wheel. this way the light shines THROUGH the PCB to make a bright spot in the background, then use the spoke in the indexing wheel would break the light flow through. Then it becomes a light break beam... BUT you nailed it. WAY TO GO
Nello buddy, As an electrical engineer I was thinking about using the pcb itself as a rotary encoder, Not in an optical way, just in electrical conductivity way... I know the project is over by now but you can use this method for low pitch encoding. And great job dude u doing well keep going!
I am not really sure if a PID loop/controller is the best solution for this application. Don't get me wrong, I love a good PID loop but I think the way you are doing it now is the right way. A PID controller takes an analog input (sensor) value and an analog setpoint to decide some sort of output. If you were directly measuring the rotation of the wheel (by means of an encoder or something similar) then I think it would make sense. In your case, however, your output is binary (with noise of course). Either way, super sweet video and project. Can't wait to see even more progress.
I use PCBs for all kind of things. They are also great and good looking for front panels. Nowadays it's even not to expensive to get PCBs made out of aluminum.
Would it be viable to use a motor to just pull on the film? The film would pull the tape forward, and then a thru-beam interrupter could count the holes in the tape
Just part way through. Question. If you're using a slotted wheel with a light sensor, why did you not just put the light and sensor on opposite sides of the wheel so the sensor could see the light through the gaps instead of relying on a reflective surface alongside slots? I mean, isn't the either the reflective surface or the slot superfluous? Instead of a slot, couldn't a flat black paint have worked as well? Not saying it was wrong or trying to troll, just wanting to see your reasoning behind the decision.
Cool project, if you do get the little opto encoder boards redone in the future with gold as apposed to HASL, spray a little clear coat or conformal coating over the board to stop the gold from tarnishing over time. Photo etching is also a cool process for making inexpensive encoder wheels. 👍
Or the fifo used for the LEDs to send a stream of data to the faders and make each one remove the first entry of the dataset and pass it along? (the output can be passed back to the motherboard and contains the feedback on the feeders?)
AHHH good eye! If you don't have a cap on your programming arduino, it will reset when your computer tries to send the firmware over. The cap keeps it alive and holds the serial port open!
I'm currently working on a toolchanger with a PnP-Toolhead as well. I love your idea of this ''independent'' PnP-Feeder. If this project is opensource were I get the files? If not will you sell this Feeder in the near future?
I know it has been 4 months since you made the video, so not sure if it is relevant anymore. But if you take some solder wick and remove the solder from the pads, wouldn't that leave them with a thin, ever coating? (like you can not remove the last thin layer with wicking) Just a thought 🙂
Seeing the evolution of the feeder has been interesting. The tape peeler design is especially neat. For the drive motor, to get increased torque, Instead of having that encoder wheel, you could use a motor drive with built in encoder. Like this: www.dfrobot.com/product-1457.html Or this www.ebay.com/itm/DC3V-6V-5V-2-Phase-4-Wire-Micro-Worm-Gear-Stepper-Motor-Mini-Gearbox-Reducer-DIY-/322695331937 There are also ic's that can read the angle of a magnet nearby to a degree accuracy that could measure the wheel position. (like the as5600) You could also add a photo interrupter shining through the holes on the pcb tape. It would detect the advancement and also let you know when the tape is out. Like the EE-sx1321 Looking forward to see how you work this out. It may be useful to go with a completely 3d printed case without pcb to troubleshoot mechanics. Is there a reason you are using the pcb for a significant amount of the structure?
Im surprised you didn't just have the plastic cover peel back over a thin rod in the opposite direction with a gear set squeezing and rolling it off as it moves
Oooh, the joy of a blinking LED. hahahaha so relatable. Well done!!! Weird question, with the reflective wheel + sensor, does that mean this machine needs to operate with a light on?
Wahoo thanks!! And to answer your question, yes! The little sensor I'm soldering to the PCB is comprised of a suuuuper tiny LED and a photodetector in one package, but it's not putting out light in the visible spectrum (I think it's UV? not sure). So if you take a peek inside it, you won't see any light, but there's definitely some in there! I'm also adding some baffling so any stray light doesn't mess with the readings.
so i am also building my own pick and plce machine from a lazor cutter on thing i will say is i am going to hace the reals on the sides of the machine the front is a bad place in my opionon and from the looks of it thats where you are puting yours but also for the feeders you could try using i2c for the control since you can do 127 slave devices i also want to know if you have a discord for the pick and place project if so i would like to join as a fellow builder and i would even like to share some pics of my machines build progress on said discord
Yeah, I decided to put the feeders on the front because that way I can add in a conveyor belt from left to right so boards can automatically be fed through. I like the sides, but it was pretty important to me that I'm able to add the conveyor system! I hadn't thought of making a Discord for it. I'll look into making one and let you know if I do! Would love to see your build.
I would have used a sensor that shoots the light through the wheel not reflect off it. Much more binary situation aka solid on and off state. When I'm designing things that have to be hand assembled I avoid chips that have lead-less bodies. As you have shown they are very difficult to solder by hand.Where are the design files available?
I considered doing that! The reason I didn't was because then I'd have to put components on the other side of the wheel, which means another PCB, and a connector for that PCB. Plus, with analog feedback I can position the wheel exactly on the edge of each encoder window, instead of only having a binary signal which limits my resolution to the width of a window. And hopefully, I'll be building these things with a pick and place soon so lead-less bodies won't be an issue! 😁Design files are a bit of a mess at the moment, but once they're in a good state I'll post all the source on Github!
@@stephen_hawes The problem with lead-less components is soldering. They were designed for automated assembly with solder paste stencils, pick n place machine, and re-flow oven.
Wha? Why not just get a wheel from an old printer. Its LED and sensor shines through the clear spots on the wheel, and not through the black lines on the wheel. You are actually reinventing the wheel!
Okay don't use solder wick on the microcontroller pins like that. It's abrasive and when hot can damage the solder resist. Just get a flux pen and drag solder for the best results.
@@mattweger437 Hmm really? I've never had issues with that in the past. Interesting to know you've had problems with the soldermask though. Motor controller holds a constant speed for a while, then ramps down to a slower speed on the approach in a naive effort not to overshoot. When the PID is in there, motor speed will be completely based on feedback from the sensor.
![My New Pick & Place Machine [Liteplacer]](/img/1.gif)
Awesome work on this Stephen! Can't wait to see it running a job!
you prolly dont give a shit but does someone know of a way to log back into an instagram account??
I was dumb lost the password. I appreciate any tips you can give me.
@Uriel Aden Thanks for your reply. I found the site on google and Im in the hacking process now.
I see it takes a while so I will reply here later with my results.
@Uriel Aden It did the trick and I finally got access to my account again. I'm so happy:D
Thanks so much you really help me out!
@Reyansh Dario glad I could help =)
Congratz dude, using the pcb like that something i wasn't expecting to work. I don't know jack about interactive systems but just seeing someone like you enjoying the pay-off is making my day.
The idea to use the pcb as a mechanical part is absolutely genius! I would not have thought of that! Great video as always :D
Awesome video man. Cool to see this coming together. And it's an honor to be inscribed on your workbench. Bless up.
Thanks! It's an honor to have you there. Thanks for the support!
I wish you ordered the controller and wheel in different colors
Again, I just want to commend how clean your projects are. Nice video Stephen!
Is the PCB and 3D files available for download?
Great project, Stephen. Why not just turn the encoder wheel the other way.? Less signal strength though, but I guess there's not an issue.
This is awesome and pretty much the biggest stumbling block that most in house assemblers have to deal with
Congrats, that's pretty gangster for a DIY build and the excitement here is palpable. 👍
Looking forward to see the full DIY pick and place! Awesome job!
Here's a thought, If you get the drive wheel PCB made with a matt black solder resist, then use a white silkscreen instead of slots. The way you should be able to get a much finer resolution on the indexing, which will be helpful for different sized SMDs.
If you use RS485 and put a DIP switch that sets the address of each feeder you can address each feeder over the same 3 wires.
If you use CAN you get a few network type advantages. The only down side with CAN is a smaller selection of controllers
Great Design. I'm currently working on my own PNP. I am to the point where it's populating the board fairly accurately. Still need to set up the pipelines for up facing camera. I've recently ordered boards to give your feeder a try. Funny thing is, JLCPCB refused the indexing wheel board. Didn't give much reason except it was beyond their capability.
Huh.
Use solder braid to remove the excess solder from the reflecting surfaces and they will be much flatter.
That trick you pulled making them print the gear part is veery clever
And btw, the clouseup soldering parts were extremely satisfying to watch
I don't think you need to use gold solder - just give it a very very very short sand on some very fine grit paper, sellotaped down to some float glass. That'll get her shining.
Stephen, you rock , man. You are awesome showing your stuff. Two thumbs up, buddy.
Wicked cool design! I am so excited that the design files will be released. One thing I thought of as a potential improvement to the rotating wheel is that instead of using those pads to reflect light, you could just use them like a switch. Just put a brush on a spring to make the electrical contact, The other part of the switch would be a circular pad where the shaft of the motor goes through the center of.
I m started to build a INDEX unit on your info. When looking to all the feeder design i think that this is rather a problem. After thinking I wonder why the control of the feeder must be done with your piggy board on yuour special protocol. ??? Now I think you are repace this by RS485 protocol. Now I have the idea , why not place a sensor in front of the feeder able to move a mechanisme to set the tape forward. A actuator can be mouted on the Z axle eg. IR or magnet (Hall sensor on feeder). In Openpnp you sent the head to feeder wanted, this response on the sensor activation and go!.
I hope my idea is clear, sorry for perhaps the not correct English , regards, Cees from the Netherlands
why not use a optical through-beam sensor instead of optical reflective sensor?
I might have tried Reverse mounting an LED to the main board behind the indexing wheel. this way the light shines THROUGH the PCB to make a bright spot in the background, then use the spoke in the indexing wheel would break the light flow through. Then it becomes a light break beam... BUT you nailed it. WAY TO GO
Nello buddy,
As an electrical engineer I was thinking about using the pcb itself as a rotary encoder,
Not in an optical way, just in electrical conductivity way... I know the project is over by now but you can use this method for low pitch encoding.
And great job dude u doing well keep going!
I am not really sure if a PID loop/controller is the best solution for this application. Don't get me wrong, I love a good PID loop but I think the way you are doing it now is the right way. A PID controller takes an analog input (sensor) value and an analog setpoint to decide some sort of output. If you were directly measuring the rotation of the wheel (by means of an encoder or something similar) then I think it would make sense. In your case, however, your output is binary (with noise of course). Either way, super sweet video and project. Can't wait to see even more progress.
Hey Stephen is it open source ??
please help where I can find complete project..
😊
This was from over a year ago but did you add a tape empty/out detector?
I know it is too late now, but you could have some kind of analog hysteresis on it so you can use a digital signal
I use PCBs for all kind of things. They are also great and good looking for front panels. Nowadays it's even not to expensive to get PCBs made out of aluminum.
Would it be viable to use a motor to just pull on the film? The film would pull the tape forward, and then a thru-beam interrupter could count the holes in the tape
Which N20 motor are you using? How many volts?
Ayy its done, looks great! Keep up the good work dude!
Just part way through. Question. If you're using a slotted wheel with a light sensor, why did you not just put the light and sensor on opposite sides of the wheel so the sensor could see the light through the gaps instead of relying on a reflective surface alongside slots? I mean, isn't the either the reflective surface or the slot superfluous? Instead of a slot, couldn't a flat black paint have worked as well?
Not saying it was wrong or trying to troll, just wanting to see your reasoning behind the decision.
Cool project, if you do get the little opto encoder boards redone in the future with gold as apposed to HASL, spray a little clear coat or conformal coating over the board to stop the gold from tarnishing over time. Photo etching is also a cool process for making inexpensive encoder wheels. 👍
I really enjoy this series. Where did you get this heavy duty pcb holder? 1:49
Damn cool project, love your enthusiasm!
I was not at all prepared for the screen rotation, almost puked xD.
Hey . Where can I download feeder models, circuit boards, circuitry?
Congratulations. Are you going to publish the design files for this?
The guy is like me, I'm into electronic since the age of 9 years. I'm 42 now and still same exciting like him. Hahaha 😂 good job
Nice Project. Where do you get these intusiasm from?
What is the name of the optical sensor used in this project?
Why didn't you use Multiplexers and serial communication for those Fenders?
And you could use some dip switches or solder jumpers to encode the position on the fingjoint indexing bords?
Or the fifo used for the LEDs to send a stream of data to the faders and make each one remove the first entry of the dataset and pass it along? (the output can be passed back to the motherboard and contains the feedback on the feeders?)
Couldn't you just detect the rising edge instead of a threshold to fix the "lumpy" signal?
Great project! One question, what´s the deal with the cap between ground and reset on your programming arduino?
AHHH good eye! If you don't have a cap on your programming arduino, it will reset when your computer tries to send the firmware over. The cap keeps it alive and holds the serial port open!
@@stephen_hawes Thx buddy!
Hey Stephen, awesome work. Where can i buy this pcb holder for solder? Looks dope 🤤
Keep it up dude! This is so cool. Looking forward to more of your videos!
I'm currently working on a toolchanger with a PnP-Toolhead as well. I love your idea of this ''independent'' PnP-Feeder. If this project is opensource were I get the files? If not will you sell this Feeder in the near future?
I know it has been 4 months since you made the video, so not sure if it is relevant anymore. But if you take some solder wick and remove the solder from the pads, wouldn't that leave them with a thin, ever coating? (like you can not remove the last thin layer with wicking) Just a thought 🙂
Seeing the evolution of the feeder has been interesting. The tape peeler design is especially neat.
For the drive motor, to get increased torque, Instead of having that encoder wheel, you could use a motor drive with built in encoder. Like this: www.dfrobot.com/product-1457.html
Or this www.ebay.com/itm/DC3V-6V-5V-2-Phase-4-Wire-Micro-Worm-Gear-Stepper-Motor-Mini-Gearbox-Reducer-DIY-/322695331937
There are also ic's that can read the angle of a magnet nearby to a degree accuracy that could measure the wheel position. (like the as5600)
You could also add a photo interrupter shining through the holes on the pcb tape. It would detect the advancement and also let you know when the tape is out. Like the EE-sx1321
Looking forward to see how you work this out. It may be useful to go with a completely 3d printed case without pcb to troubleshoot mechanics. Is there a reason you are using the pcb for a significant amount of the structure?
I know it is late, but why didn't you get a stencil
Really awesome video! Can't wait to see more! You used the DRV8833 in the previous video, what motordriver are you using now?
Then I continued to watch, actually my second time through the video =D I saw 838 on the chip, so I guess its a TI DRV8838 chip.
Im surprised you didn't just have the plastic cover peel back over a thin rod in the opposite direction with a gear set squeezing and rolling it off as it moves
Oooh, the joy of a blinking LED. hahahaha so relatable.
Well done!!!
Weird question, with the reflective wheel + sensor, does that mean this machine needs to operate with a light on?
Wahoo thanks!! And to answer your question, yes! The little sensor I'm soldering to the PCB is comprised of a suuuuper tiny LED and a photodetector in one package, but it's not putting out light in the visible spectrum (I think it's UV? not sure). So if you take a peek inside it, you won't see any light, but there's definitely some in there! I'm also adding some baffling so any stray light doesn't mess with the readings.
@@stephen_hawes Likely InfraRed
Is the pcb and stl files available download?
Amasing work ! I'm very interrested by your project but where can I find designs code etc ? Thanks a lot keep going like that !
What is that thing used to hold the PCB. I want one
It's called an Omnivice! I've tried a lot of helping hands, and this is the best one I've ever used. A bit pricey but worth every penny!
Really cool, keep up the good work!
Subbed due to excitement... needed that
so i am also building my own pick and plce machine from a lazor cutter on thing i will say is i am going to hace the reals on the sides of the machine the front is a bad place in my opionon and from the looks of it thats where you are puting yours but also for the feeders you could try using i2c for the control since you can do 127 slave devices i also want to know if you have a discord for the pick and place project if so i would like to join as a fellow builder and i would even like to share some pics of my machines build progress on said discord
Yeah, I decided to put the feeders on the front because that way I can add in a conveyor belt from left to right so boards can automatically be fed through. I like the sides, but it was pretty important to me that I'm able to add the conveyor system! I hadn't thought of making a Discord for it. I'll look into making one and let you know if I do! Would love to see your build.
should have gold flashed the wheel
hi this is the first comment love ur passion never loose it love all ur videos
i am happy for you
So good job, I interesting your video.
I would have used a sensor that shoots the light through the wheel not reflect off it. Much more binary situation aka solid on and off state. When I'm designing things that have to be hand assembled I avoid chips that have lead-less bodies. As you have shown they are very difficult to solder by hand.Where are the design files available?
I considered doing that! The reason I didn't was because then I'd have to put components on the other side of the wheel, which means another PCB, and a connector for that PCB. Plus, with analog feedback I can position the wheel exactly on the edge of each encoder window, instead of only having a binary signal which limits my resolution to the width of a window. And hopefully, I'll be building these things with a pick and place soon so lead-less bodies won't be an issue! 😁Design files are a bit of a mess at the moment, but once they're in a good state I'll post all the source on Github!
@@stephen_hawes The problem with lead-less components is soldering. They were designed for automated assembly with solder paste stencils, pick n place machine, and re-flow oven.
Wha? Why not just get a wheel from an old printer. Its LED and sensor shines through the clear spots on the wheel, and not through the black lines on the wheel. You are actually reinventing the wheel!
Use brass not gold. Brass is shinier.
Okay don't use solder wick on the microcontroller pins like that. It's abrasive and when hot can damage the solder resist. Just get a flux pen and drag solder for the best results.
Is your motor controller holding speed?
@@mattweger437 Hmm really? I've never had issues with that in the past. Interesting to know you've had problems with the soldermask though. Motor controller holds a constant speed for a while, then ramps down to a slower speed on the approach in a naive effort not to overshoot. When the PID is in there, motor speed will be completely based on feedback from the sensor.
@@stephen_hawes yeah it's a thing. It gets worse with closer pitches.
Son of a b*tch I need more!
That solder paste is bad. Trow it away
5:30 IS THAT A SCREW IN YOUR EAR?? 😂😂😂😂
why screw in your ears ????
Cool project. Less talking please.