Thanks a bunch, Sanjay! I'll be at MRRF with this contraption if you'd like to see it in-person. (I'm also absolutely stoked to finally get a closer peek at your folks' tool changer too!)
Can the motor have two more cables attached to the other side using a second pulley that when its posts spins it locks on the tool while a second locking mechanism releases the tool from its rest? Reason I ask is when you make tools that are electrical and need to plug into the head they may not come off the arm easily since the sockets will be very snug and will require a bit of force to pull them apart so a locking mechanism to hold them in the rest while the head pulls away will be needed. So essentially while the head's lock unlocks to release it to the tool rest, another locking mechanism built into the tool rest locks it at the same time so the tools will need a locking slot printed for both sides. What is there to trip the stall detection in the drivers when you lock it? There's no hard-stop there on your printed part like there is when it unlocks by the three printed parts colliding with each other. I do hear a click when you printed part on the motor's pulley faces the back of the motor where its wires are ran out.
Nice work. It is very compact. Regarding the flexing issue, this part seems more rigid: do you plan to put a metal shim insert in the tool head to do a metal/metal contact with the lock ? Is it a bad idea to support on both sides the 6 screws of the coupler to prevent any movement on the XY plane of such screws ?
Aye; this variant is much more rigid. So far, Im locking the tool into a wedged shaped printed part, so I wont need a metal shim like E3D's. These are intended to be consumables that can get replaced without having to realign the tool.
The idea is nice, but i think that no matter what cord you will use (nylon, steel) it will wear our pretty fast with so many, short distant movements inside spring guide. Correct me if I am wrong. Maybe kevlar will do.
did you put a bearing on the back side of the tool head? In the process of printing out my second extruder head right now! I like the new fatter design. Maybe I'll update mine too because there are significant ringing artifacts in the one print i did. However, I also just want to try it with two colors before changing too much
I added a piece of shim on the back side, but no bearing. Yeah, the fatter (and shorter) design has made the entire thing far more rigid on my machine.
Hi. Would you explain please how to configure the firmware to use this feature? I try to make it on Duet 2 WiFi board, but motor doesn't stop by stall detection.
Thanks a bunch for the kind words. I machined it starting from a rod of 0.25-in aluminum. In this entire system, it is the only machined part. I added a drawing to the thingiverse docs for reference. cdn.thingiverse.com/assets/3a/db/cb/6e/52/pin_lock_shaft.pdf I suspect you could print this solid with a tough filament, or perhaps grind down a screw to get a similar piece.
Dear Joshua, I ran into some problems while building my own toolchanger. I wonder of you have experienced the same problem. Kinematic couplings have a problem when placed vertically. The head has to move sideways when it is out of position. Therefore it has to overcome the friction of the balls in the V-grooves. When placed horizontally this is no problem. Vertically it is because you need a additional device which pulls the balls en V- grooves together. This systems has to pull relatively hard an therefore has large friction with the printhead. Combine these frictions and add a bit of gravity (heavy printhead) and you have a system which is unable to return to its desired position. Sadly enough this what is happening with my system. My system is different to yours because I don't rotate the T axle. I move it in and out. The T axle holds the head by spring force. The movement of the T axle is done by an external pull motor (only 1 cable needed instead of two with your system). Increasing the spring force doesn't help because it is the spring force which creates the friction in the first place. My ball sit in a V-groove with a 90 degree angle. Changing the angle doesn't help much because the main friction is created by the T axle pulling on the printhead. Did you try this solution? I get the feeling that this is why people are using a rotating system instead of my translating system. A rotating system overcomes the main friction problem because it is already rotating. However, when the head is locked it should be locked without a spring system on the axel. Or else you end up with the same problem as my solution. What do you think? Kind regards, Frank
Hi Frank, I suspect that your kinematic coupling grooves have too much stiction. In the case here and all future versions, the grooves are made from smooth steel dowel pins, which habe low friction and prevent the balls from sticking when they are pulled into position. That's not to say that this coupling is without it's problems. I did have issues where the wedge plate was catching with the twist lock, which would have the effect of rotating the tool instead of strictly pulling it into position, but this 8s largely solved by switching materials. More info on that issue here: github.com/machineagency/jubilee/issues/43
@@POOFjunior Thank you for your response. I appreciate it. It seems you had the exact same problem. I did polish the V-grooves. This helped a little bit. I ran some calculation. My design fails when the coefficient of friction is above 0.15. To be safe a coefficient of friction of 0.05 would be advisable. Sadly enough this is impossible. 0.15 is already rather low. This proves my design is fundamentally flawed. For my specific printer I came up with two solutions. 1) Run the head into de tool so the tool perfectly aligns and than activate the locking mechanism. 2) Pick op the tool and bring it to a position where there are some spring loaded pins and run the the tool into those pins. The pins might push the tool to its perfect position. Solution 1 is preferable but also very difficult with my specific printer. My printer uses a lift mechanism to park the tool. The lift mechanism might be to unstable for solution 1. We will see.
That's a really neat solution to move the motor's mass to the frame! Following with great interest
Brilliant work Joshua! It looks promising! Keep on going. Really good work
Thanks for the kind word, Lorenzo! I'm hacking-away in software-land now...
I love this thing! Great work dude, is certainly interesting enough to consider for our toolchanger!
Thanks a bunch, Sanjay! I'll be at MRRF with this contraption if you'd like to see it in-person. (I'm also absolutely stoked to finally get a closer peek at your folks' tool changer too!)
I'll be there for sure! Looking forward to meeting you and seeing your implementation!
@@SanjayMortimer Likewise!
Can the motor have two more cables attached to the other side using a second pulley that when its posts spins it locks on the tool while a second locking mechanism releases the tool from its rest? Reason I ask is when you make tools that are electrical and need to plug into the head they may not come off the arm easily since the sockets will be very snug and will require a bit of force to pull them apart so a locking mechanism to hold them in the rest while the head pulls away will be needed. So essentially while the head's lock unlocks to release it to the tool rest, another locking mechanism built into the tool rest locks it at the same time so the tools will need a locking slot printed for both sides.
What is there to trip the stall detection in the drivers when you lock it? There's no hard-stop there on your printed part like there is when it unlocks by the three printed parts colliding with each other. I do hear a click when you printed part on the motor's pulley faces the back of the motor where its wires are ran out.
I am very excited to watch this develop and make one of these myself!
How did you stop the pin from move back and forward through the bearing?
Nice work. It is very compact. Regarding the flexing issue, this part seems more rigid: do you plan to put a metal shim insert in the tool head to do a metal/metal contact with the lock ? Is it a bad idea to support on both sides the 6 screws of the coupler to prevent any movement on the XY plane of such screws ?
Aye; this variant is much more rigid. So far, Im locking the tool into a wedged shaped printed part, so I wont need a metal shim like E3D's. These are intended to be consumables that can get replaced without having to realign the tool.
Would it work for pen plotter Z axis movement? I don't need much, ~3-4 cm
Where did you find the cable drive?
The idea is nice, but i think that no matter what cord you will use (nylon, steel) it will wear our pretty fast with so many, short distant movements inside spring guide. Correct me if I am wrong.
Maybe kevlar will do.
Damn! This is cool ! Great work
did you put a bearing on the back side of the tool head? In the process of printing out my second extruder head right now! I like the new fatter design. Maybe I'll update mine too because there are significant ringing artifacts in the one print i did. However, I also just want to try it with two colors before changing too much
I added a piece of shim on the back side, but no bearing. Yeah, the fatter (and shorter) design has made the entire thing far more rigid on my machine.
Great work!
Hi.
Would you explain please how to configure the firmware to use this feature? I try to make it on Duet 2 WiFi board, but motor doesn't stop by stall detection.
Stunning. How you made the rotating shaft with pin?
Thanks a bunch for the kind words. I machined it starting from a rod of 0.25-in aluminum. In this entire system, it is the only machined part.
I added a drawing to the thingiverse docs for reference.
cdn.thingiverse.com/assets/3a/db/cb/6e/52/pin_lock_shaft.pdf
I suspect you could print this solid with a tough filament, or perhaps grind down a screw to get a similar piece.
great work
Dear Joshua,
I ran into some problems while building my own toolchanger. I wonder of you have experienced the same problem.
Kinematic couplings have a problem when placed vertically. The head has to move sideways when it is out of position. Therefore it has to overcome the friction of the balls in the V-grooves. When placed horizontally this is no problem. Vertically it is because you need a additional device which pulls the balls en V- grooves together. This systems has to pull relatively hard an therefore has large friction with the printhead. Combine these frictions and add a bit of gravity (heavy printhead) and you have a system which is unable to return to its desired position.
Sadly enough this what is happening with my system. My system is different to yours because I don't rotate the T axle. I move it in and out. The T axle holds the head by spring force. The movement of the T axle is done by an external pull motor (only 1 cable needed instead of two with your system). Increasing the spring force doesn't help because it is the spring force which creates the friction in the first place. My ball sit in a V-groove with a 90 degree angle. Changing the angle doesn't help much because the main friction is created by the T axle pulling on the printhead.
Did you try this solution? I get the feeling that this is why people are using a rotating system instead of my translating system. A rotating system overcomes the main friction problem because it is already rotating. However, when the head is locked it should be locked without a spring system on the axel. Or else you end up with the same problem as my solution.
What do you think?
Kind regards,
Frank
Hi Frank,
I suspect that your kinematic coupling grooves have too much stiction. In the case here and all future versions, the grooves are made from smooth steel dowel pins, which habe low friction and prevent the balls from sticking when they are pulled into position.
That's not to say that this coupling is without it's problems. I did have issues where the wedge plate was catching with the twist lock, which would have the effect of rotating the tool instead of strictly pulling it into position, but this 8s largely solved by switching materials.
More info on that issue here:
github.com/machineagency/jubilee/issues/43
@@POOFjunior
Thank you for your response. I appreciate it.
It seems you had the exact same problem.
I did polish the V-grooves. This helped a little bit.
I ran some calculation. My design fails when the coefficient of friction is above 0.15. To be safe a coefficient of friction of 0.05 would be advisable. Sadly enough this is impossible. 0.15 is already rather low. This proves my design is fundamentally flawed.
For my specific printer I came up with two solutions.
1) Run the head into de tool so the tool perfectly aligns and than activate the locking mechanism.
2) Pick op the tool and bring it to a position where there are some spring loaded pins and run the the tool into those pins. The pins might push the tool to its perfect position.
Solution 1 is preferable but also very difficult with my specific printer. My printer uses a lift mechanism to park the tool. The lift mechanism might be to unstable for solution 1. We will see.
great work!!, do u gonna upload the files when its done? cant wait!
Just did--and many thanks!
www.thingiverse.com/thing:3365456
Hello. I am building my printer, I want it to have 4 heads, I really like this solution but I'm not that good. Will you provide the documentation.
For sure! I just popped some documentation up here:
www.thingiverse.com/thing:3365456
Feel free to comment there or PM me with questions.