This is really cool, makes me want to get a lathe even more. For holding the round stock on the drill pres you might want to look in to V-blocks. A single block will hold a large range of diameters, and you are not limited to which ever drill size you got.
Nice. Slick mechanical setup. When I saw the stepper at the beginning, I was thinking 'no way that'll make speed and torque'. Then you swapped to the DC motor. I'll give you a warning about the controller you built: get it off the protoboard. The contacts are fairly high resistance, and at high currents, like with that motor, they will drop a lot of potential, reducing motor drive, and will heat up, making things worse as they lose their temper. The end result is burnt tiepoints on the board and/or damaged electronics.
+john john I figured that the breadboard was not ideal, I have some of the solderable boards that I'll probably transfer to this week (as well as adding heatsinks).
I am too late to comment on this video but would like to ask some questions 1. How did u do gear calculations according to your need? 2. For the timing belt pulley where did u get that mounting and pulley. Thanks alot
Hi scott, just found your channel and immediately subbed. I see you are using a stepper motor and timing belt. Why do you need encoder? Wouldn't it be possible to figure out the position by counting the number of steps and calculating steps/mm like they do in CNC systems? Correct me if I am missing something.
Ok, gotcha. No probs cause I will use a DC motor with a H bridges. Btw, I am trying to make my own inverted pendulum, but I am having some troubles with the pinion and the spur gears. In your case you are using a self made "gearbox", but in my case I do not have the tools for doing that. Do you know any possible solution (for buying) to link the pinion and the spur gears? Like a prefabricated gearbox or similar. Thanks in advance and congratulations for your work!
Were you using a constant current stepper driver like the A4988 or something similar? Or did you just hook up the stepper to a double H-bridge? If you didn't use a constant current driver, that might be the reason why your stepper lost torque and stalled at higher speeds. Try using one of those drivers and use the highest voltage the driver can take, even though the motor rated voltage is lower, using a higher voltage allows you to keep a good torque at higher speeds, and avoid stalling. If you can get it to work with a stepper driver, you can avoid all the problems involved in using a DC motor, like it humming but not spinning at low duty cycles. It's much better. I saw this 1 video where the guy used a stepper in a project just like yours and it worked amazingly, and could reach high speeds without stalling. Here is the link: ua-cam.com/video/OBcsSwayEsc/v-deo.html
Good to know, I experimented with the microstepping options but didn't get into the current. Probably good to do a project focused strictly on stepper performance before relying on it in such a sensitive application like this. Oh well, now I appreciate the importance.
@@ScottRumschlag I'm about to get my degree in control and automation engineering, and I built this exact project for my final paper. I'm planning to test some different control algorithms, but first I wanted to get all the instrumentation in order and working properly. I originally used a DC motor and I faced all exact same problems as you did, the main one was the "dead band" when using PWM where the motor would just humm but not spin up to about 30-40% duty cycle. I tried fiddling with different frequencies and got some different behaviours but nothing solved the problem. It wasn't the case to just add an offset to my control signal, because the dead band would change from time to time as the tightness of the rubber belt changed, and if the car was already moving, it would keep moving at low duty cycles, it just wouldn't get started from zero, not to mention that the motor wouldn't spin up to a certain duty cycle, and then it just skipped to a higher speed, and I couldn't get it to start at low speeds. So that messed with the balance point of the pendulum. I could get it to balance as it jittered around non stop, but it could never stand still and steady like I saw other people doing on videos, and I got very frustrated. I even got that phenomenon where the car would insist in idling to one side as it balanced, even though the plane was perfectly level. Anyways, for that reason, I decided to see if it was possible to use a stepper motor instead, as that would solve most of my instrumentation problems, so I set off to watch a bunch of videos, and that's how I found you. When I saw that you used a stepper in the beginning I got very excited, but very disappointed when you had the problem with the stepper stalling at higher speeds. But I kept searching and found some other videos about stepper drivers, and found this one guy that has a guitar workshop, and he has a CNC machine. He talks about this exact problem of the loss of torque and stalling at higher speeds, and that he uses a constant current driver, and when he increases the voltage on the driver, the motor doesn't lose too much torque at higher speeds, and he is able to go much faster without stalling, even with a load. That got me more hopeful, so I started reading more about constant current drivers, and I eventually found the video that I linked to you on the previous comment, and that made me very confident that it is indeed possible. I will be getting my stepper driver soon and start running some tests, but I would love to see you try it too on your project and see how it works out for you. The reason why increasing the voltage on the driver helps with the torque at higher speeds is because the driver uses a feedback sense resistor to measure the current on the motor coils, and manipulates the voltage it applies to the coils to get the current to the desired constant level in each state of the step sequence. So even if your motor is rated for 12 volts for example, you can use something like 35 volts on your driver, and since the current on the motor coils can't change instantly since it's an inductive load, the driver will first apply as high a voltage as it can (35V supply voltage) to get the current to the desired constant level faster, and then lower the voltage to maintain that current during that step. So the higher the voltage you feed your driver with, the faster it will be able to get your stepper's coil current to the desired level in each step. So for that reason you are able to work with much higher frequencies and still get the coils to reach the appropriate current level in each step before the next step comes.
This is really cool, makes me want to get a lathe even more. For holding the round stock on the drill pres you might want to look in to V-blocks. A single block will hold a large range of diameters, and you are not limited to which ever drill size you got.
Some pretty cool progress. Seems like it took a lot of work to make the motor-controller and gearing set up.
+akovaski Yeah, I'm sure I could order something nicer but I wanted to get this going immediately . . . and this is more fun.
+Scott Rumschlag It's not about the destination, it's about the journey.
Scott--it's nice to see you smile buddy!!!
cảm ơn bạn, nhờ bạn mà tôi dễ dàng hơn trong việc làm nó, bởi vì tôi đang muốn làm đồ án tốt nghiệpvề con lắc ngược này
Chào bạn, không biết bạn còn tài liệu về mô hình con lắc ngược này không
Nice. Slick mechanical setup. When I saw the stepper at the beginning, I was thinking 'no way that'll make speed and torque'. Then you swapped to the DC motor. I'll give you a warning about the controller you built: get it off the protoboard. The contacts are fairly high resistance, and at high currents, like with that motor, they will drop a lot of potential, reducing motor drive, and will heat up, making things worse as they lose their temper. The end result is burnt tiepoints on the board and/or damaged electronics.
+john john I figured that the breadboard was not ideal, I have some of the solderable boards that I'll probably transfer to this week (as well as adding heatsinks).
you should try making a gimbal for your camera
Hi! I´m very interesting with this proyect, i want to make an inverted pendulum for the school, but i don´t know how begin it... can you help me?
Awesome! I am about to make a project about kapitza's pendulum. Thank you!
Great video! Rip Radioshack...
please how did you make your speed control
hi Scott..this is really interesting..Can you please tell me the parts and their dimensions...I'm making school project...thanx
Hey maninder, have you got your school project working. I'm doing one and i have a few questions if you can help.
Thank you
I am too late to comment on this video but would like to ask some questions
1. How did u do gear calculations according to your need?
2. For the timing belt pulley where did u get that mounting and pulley.
Thanks alot
1. Ball park estimation
2. The belts and pulleys were from CNC machine parts available on "maker" sites, the shaft and mounting I made myself.
Hi scott, just found your channel and immediately subbed. I see you are using a stepper motor and timing belt. Why do you need encoder? Wouldn't it be possible to figure out the position by counting the number of steps and calculating steps/mm like they do in CNC systems? Correct me if I am missing something.
The motor can miss steps, particularly with higher speeds. The torque, at least on this particular motor, was much less at higher speeds.
Hi friend, what character is the motor and
because you added a larger pinion for the motor
It's an RC car motor, I forget the turns but it's in the video.
may i know spesifics of your motordc used?
It's an RC car motor.
how much torque can be accommodated?and may i know the maximum rpm of these motor?
thanks before
Very nice project
Hi Scott! What kind of electronic board you are using in minute 3:26?
The green one is a stepper driver, the one on plywood is just a breadboard I fastened down.
Ok, gotcha. No probs cause I will use a DC motor with a H bridges.
Btw, I am trying to make my own inverted pendulum, but I am having some troubles with the pinion and the spur gears. In your case you are using a self made "gearbox", but in my case I do not have the tools for doing that. Do you know any possible solution (for buying) to link the pinion and the spur gears? Like a prefabricated gearbox or similar.
Thanks in advance and congratulations for your work!
hello Scott please could you send the triple pendulum video that motivated you?
Link is in the video description.
oh yes thank you very much
Keep up the good work, Its brain food =)
Hi friend, what character is the encoder
Just a generic 2400 PPR quadrature encoder, they're about $16.
@Silviu Vijiala Yes, advertised as 600 but if you read each edge of the quadrature signal you get 4x that.
Were you using a constant current stepper driver like the A4988 or something similar? Or did you just hook up the stepper to a double H-bridge?
If you didn't use a constant current driver, that might be the reason why your stepper lost torque and stalled at higher speeds. Try using one of those drivers and use the highest voltage the driver can take, even though the motor rated voltage is lower, using a higher voltage allows you to keep a good torque at higher speeds, and avoid stalling. If you can get it to work with a stepper driver, you can avoid all the problems involved in using a DC motor, like it humming but not spinning at low duty cycles. It's much better.
I saw this 1 video where the guy used a stepper in a project just like yours and it worked amazingly, and could reach high speeds without stalling. Here is the link:
ua-cam.com/video/OBcsSwayEsc/v-deo.html
Good to know, I experimented with the microstepping options but didn't get into the current. Probably good to do a project focused strictly on stepper performance before relying on it in such a sensitive application like this. Oh well, now I appreciate the importance.
@@ScottRumschlag I'm about to get my degree in control and automation engineering, and I built this exact project for my final paper. I'm planning to test some different control algorithms, but first I wanted to get all the instrumentation in order and working properly. I originally used a DC motor and I faced all exact same problems as you did, the main one was the "dead band" when using PWM where the motor would just humm but not spin up to about 30-40% duty cycle. I tried fiddling with different frequencies and got some different behaviours but nothing solved the problem. It wasn't the case to just add an offset to my control signal, because the dead band would change from time to time as the tightness of the rubber belt changed, and if the car was already moving, it would keep moving at low duty cycles, it just wouldn't get started from zero, not to mention that the motor wouldn't spin up to a certain duty cycle, and then it just skipped to a higher speed, and I couldn't get it to start at low speeds. So that messed with the balance point of the pendulum. I could get it to balance as it jittered around non stop, but it could never stand still and steady like I saw other people doing on videos, and I got very frustrated. I even got that phenomenon where the car would insist in idling to one side as it balanced, even though the plane was perfectly level. Anyways, for that reason, I decided to see if it was possible to use a stepper motor instead, as that would solve most of my instrumentation problems, so I set off to watch a bunch of videos, and that's how I found you. When I saw that you used a stepper in the beginning I got very excited, but very disappointed when you had the problem with the stepper stalling at higher speeds. But I kept searching and found some other videos about stepper drivers, and found this one guy that has a guitar workshop, and he has a CNC machine. He talks about this exact problem of the loss of torque and stalling at higher speeds, and that he uses a constant current driver, and when he increases the voltage on the driver, the motor doesn't lose too much torque at higher speeds, and he is able to go much faster without stalling, even with a load.
That got me more hopeful, so I started reading more about constant current drivers, and I eventually found the video that I linked to you on the previous comment, and that made me very confident that it is indeed possible. I will be getting my stepper driver soon and start running some tests, but I would love to see you try it too on your project and see how it works out for you.
The reason why increasing the voltage on the driver helps with the torque at higher speeds is because the driver uses a feedback sense resistor to measure the current on the motor coils, and manipulates the voltage it applies to the coils to get the current to the desired constant level in each state of the step sequence. So even if your motor is rated for 12 volts for example, you can use something like 35 volts on your driver, and since the current on the motor coils can't change instantly since it's an inductive load, the driver will first apply as high a voltage as it can (35V supply voltage) to get the current to the desired constant level faster, and then lower the voltage to maintain that current during that step. So the higher the voltage you feed your driver with, the faster it will be able to get your stepper's coil current to the desired level in each step. So for that reason you are able to work with much higher frequencies and still get the coils to reach the appropriate current level in each step before the next step comes.
hey Scott
Me and my team is willing to make an inverted pendulum. May i get your email id to properly communicate with you.
Please..
My e-mail is in several places on my site.