Peter Betz's Head- Revision 1 and 2

[Peter Betz]

Hi Everyone,

 Sorry I have been pretty quiet lately. Things are still chugging away in the background. First of all, this is a continuation of the original head design found here:

https://groups.google.com/forum/#!topic/openpnp/-XGpRTdaTrk ("Head Thread")

The design is perfectly fine and lends itself well to laser cut acrylic (or delrin if you are super fancy). My machine is currently running this head.

I decided to make a couple of improvements to the design (and change the default material to aluminum), enter REVISION 1:



1. New designed top plate with cable chain attach points:

The top plate has been designed as a stronger method of attaching the cable chain. The chain exerts more force on the head than I expected! This design also allows mounting holes for components on top of the head.



2. New vacuum gaskets:

In an attempt to improve Anthony Webb's gasket design, I have changed it to allow for screwing in a NPT push-to-connect fitting (for 4mm or similar vacuum line).



3: Integrated belt tensioners:



With integrated nut traps and everything being cut out of stock 1/4" material (waterjet or laser) this simplifies the design by removing the need to drill and tap the side faces to attach brackets. Special thanks to Mark Harris for helping zero in on the final design.

4. Away with the opto switch and fragile thin feature.

In order to make the design more robust, I decided to try using a high quality snap action limit switch instead. This makes the design slightly more compact and also does away with that very thin interrupter feature. Using contact sets in Fusion I was able to simulate the tripping of the switch and find the ideal mounting angle for it! These switches are rated for millions of cycles MTBF, so I consider this to me a long term, reliable solution.





5: I had a strong desire to try and integrate Z probing.

First, there was to be two plates at the bottom of the NEMA8 steppers. The upper plate is fixed to the stepper motor mounting screws and the bottom is fixed to the toothed plate attached to the linear rail. Motion is supported between the two plates along 4 precision ground shoulder screws:



Then, a new designed vacuum gasket would allow for contact with a microswitch:



Here is a GIF showing how it was designed to function:

Mark went ahead and had it cut on the waterjet to prototype. Long story short is that the required tolerances on the shoulder bolt holes is too great and would require a more precise manufacturing method, like CNC milling to maintain part placement accuracy. Unfortunately I don't think that it will be worth pursuing this method any further :(

Revision 2 is only modifying the Z probe style.

My new thought is to have the Z stepper completely unrestrained in its rotational axis (no mounting bolts securing it). The stepper shaft would pass though a flanged bearing in the main plate, and the stepper face would press up on a caged bearing. There would be a printed washer that went under the gear to allow the stepper to be drawn in to the head plate using the set screw on the gear (if that makes sense). Here is a simple mock-up:

The stepper motor would only be restrained in rotation by the contact force of two microswitches mounted on the back of the main plate:

In the normal course of operation, the only force that would be exerted on the switches is from acceleration, because the head is balanced. When the nozzle contacts something that provides resistance, it overcomes the contact pressure of the switch and it triggers.

Where we are at right now is proof of concept. I have a full set of metal parts here, and I have some switches with about the lowest contact pressure available. I am going to take my head apart and build it up with the new metal parts and also make a small 3D printed shoulder bushing to restrain the Z stepper similar to how the bearings would. I can then actually see the result on my machine and see if this method is viable. I actually have pretty high hopes for this. Any of the play in the Z system is not really that important with the spring loaded nozzles, at least that's how I feel about it!!

Now, thanks to Mikael, I really need to step up my rendering game.... sigh.

That's all for now, hope you guys like the changes.

Peter.

[Michael Anton]

How about putting a hollow tube down the center of the stepper that rides on the plunger in the nozzle, and then pick up the movement of the tube on the top of the motor?  This would directly measure the spring compression of the nozzle itself.  You might be able to use a clear hollow tube to maintain the vacuum at the top, but be able to sense the movement of the tube using an opto sensor.  The tube may need to be slightly spring loaded to keep it pressed on the nozzle.  You might need to check how consistent the plunger depth is inside the nozzle, but it looks pretty good by eye to me on the 9 different nozzles that I have.

Mike

<snip>

 

5: I had a strong desire to try and integrate Z probing.

First, there was to be two plates at the bottom of the NEMA8 steppers. The upper plate is fixed to the stepper motor mounting screws and the bottom is fixed to the toothed plate attached to the linear rail. Motion is supported between the two plates along 4 precision ground shoulder screws:

Then, a new designed vacuum gasket would allow for contact with a microswitch:

<snip>

Peter.

[bobgee...@gmail.com]

Peter,

You have been quiet, but not idle!

Great progress, many neat additions. Loosk like the Z-probing still needs some testing / thought and Micheal's idea sounds interesting to me.

Where I am right now with my machine, I think I should try and get the basic V1 parts cut and look at upgrading to something with Z-probing at a later date. I have plenty to be getting on with anyway!

With that said, is it possible for me to get the files for the V1, without any Z-probing so I can get a set manufactured? I am also intererested in your camera + LED mount, are the designs for that available somewhere too?

I guess it is criticial to use 1/4" stock without modifying the design (i.e. the finger joint for the cable chain plate, and the belt tensioner bar are cut for 1/4" slots). If I did want to go for a thicker acrylic, are there any other places in the design I would have to adjust?

Well, thanks for all the progress, and good luck with the Z-probing!

Regards,
Bobby

[Daniel Dumitru]

Thank you PEter !  Those revisions can be made still from acrylic ?

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/b01b160d-3806-4672-9b40-20bf2102540e%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Gregory Davill]

Really liking the design. Nice work!

Some random thoughts for z-probe:

I've been thinking it may be possible to utilize "stall detection" from Trinimatic TMC2130 Drivers. But a hardware solution will probably be more reliable.
Any reason why a straight vacuum sense wouldn't work for the z-probe though? Feed the value through a simple window comparator for logic level output?


Would love to see how your axially supported stepper turns out.

On Wednesday, November 16, 2016 at 6:57:02 PM UTC+10:30, Daniel wrote:

Thank you PEter !  Those revisions can be made still from acrylic ?

On Wed, Nov 16, 2016 at 10:06 AM, <bobgee...@gmail.com> wrote:

Peter,

You have been quiet, but not idle!

Great progress, many neat additions. Loosk like the Z-probing still needs some testing / thought and Micheal's idea sounds interesting to me.

Where I am right now with my machine, I think I should try and get the basic V1 parts cut and look at upgrading to something with Z-probing at a later date. I have plenty to be getting on with anyway!

With that said, is it possible for me to get the files for the V1, without any Z-probing so I can get a set manufactured? I am also intererested in your camera + LED mount, are the designs for that available somewhere too?

I guess it is criticial to use 1/4" stock without modifying the design (i.e. the finger joint for the cable chain plate, and the belt tensioner bar are cut for 1/4" slots). If I did want to go for a thicker acrylic, are there any other places in the design I would have to adjust?

Well, thanks for all the progress, and good luck with the Z-probing!

Regards,
Bobby

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

[Matt Brocklehurst]

Whats the drive in getting z-probe working? Unless I've missed something i certainly don't need it in my case, and I don't recall the iPulse i used to play with having it either... is it just a cool thing to have, or is there an actual need for it?

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0f0222ca-a498-4656-84e1-ebb1119d0ca0%40googlegroups.com.

[Trampas Stern]

I have the nano zero stepper, http://misfittech.net/product/nano-zero-stepper/, which can measure the motor's angular position and control the torque (current) to the motor.  What would be easy to do is on power up have it set the torque low and then run steppers down and up to find the min and max, from there it can keep track of the position.  Then the step and direction pins can be used to control the heads, ie both low heads parked mid position, then each pin controls which head is down.  It might even be possible to use the variable torque to control the part placement.. 

I will see if I can get a prototype in the next week. 

[Mikael Bohman]

Interesting. 

I would like to add some "active" current feedback / Force-feedback to my Z-stepper.

[Peter Betz]

@ Mike: That is an interesting idea for sure, not sure I am sold on the probable reliability though, especially reading through a tube.

@ bobby: I guess you mean revision 1 files. I will see about amending my Github with them. R1 does have Z probe features so you would have to work around them, like bolting the two plates together under the stepper. I believe you have to use 1/4 if you want everything to line up smoothly.

@ daniel: yes, 1/4" thick.

@ Gregory: It will be interesting to see where all of that ends up, i have been trying to pay attention to it. Vacuum sense will suck small parts up out of the feeder before the nozzle gets there.

@ Matt: For setup, all of the part feeders, like using strip feeders stuck to the deck, all must have their Z heights set. Its a bit of a pain. I can't see how this doesn't apply to you. The Z height of all part locations (any kind of feeder, but more variable for tape stuck to the bed) and the board must be set. Do you have a PnP that is running OpenPnp? I'm sure commercial machines just know the feeder heights because their location is repeatable and accurate, never seen one in person though.

@ Trampas: Having the head go to extents to find midpoint is risky, it can collide with something underneath the nozzles. You want to be able to home the dead before homing in X and Y. Using the feedback for probing is certainly promising, no one has gotten that far, as far as I know.

Peter.





 

[bobgee...@gmail.com]

Peter,

Thanks for the info. About amending your github, are you saying you need to put the files there, or you need to edit the files that are there?

I am wondering how much effort it would be for me to edit the file to work with metric stock sizes available in europe. I guess I only have to modify the finger joint for the cable chain plate, and the cut out in the belt tensioner bar. Are there any other places in the design I would have to adjust?

Also, about bolting the Z-probe plates together, will this be easy for the parts as they are (i.e. is there room for bolting them together with Nema 8 installed, or again, would it be easy for me to make some modifications?

Cheers,
Bobby

[Malte R.]

Hi Peter,

I really like the head design and I am considering to add this to my (already modified) LitePlacer.

I don't want to sound disgraceful, but could you (or someone else who has the software and skills) modifiy the design for metric?

1/4" (6,35 mm) stock is hard or impossible to get in most parts of the world.

6mm is common in Europe, 8mm might also be available. 7mm is difficult already. 6,35mm usually has to be custom milled and thus is unaffordable.

Thanks and best regards
Malte

[Peter Betz]

Ok people. I have linked a 6mm version of REVISION 1:

https://github.com/BETZtechnik/ICEpick--Direct-drive-pick-and-place-head/wiki/CAD-files

I did this quickly, there are likely errors. Let me know if you find any. I changed the NEMA8 plates to allow M3 tapped threads and screws to hold the plates together,

Peter.

[bobgee...@gmail.com]

Peter,

I (and I am sure others) am very grateful you have made the necessary changes for 6mm stock.

I am sorry to keep bothering you but I still have some questions, hopefully very simple ones for you:


Do you have a specific part number for the swich you have in mind? This is the swithc with the wheel on the arm. Just want make sure I order the correct part as I know they have different arm lengths and contact forces etc.

Will you by any chance be selling your custom LED pcb for the camera system? Would be nice if I could get 4 of these, or even just the pcb when you order a few for yourself (assuming you haven't).

Is the larger Motor a NEMA 17?

Is this a MGN12H block? If not, I suppose the more generic question is what are the pitches in X and Y of the mounting holes? Sorry if it sounds like a lazy question, but I don't know how to measure anything in the autodesk viewer (web browser).

I was also wondering how close are you to doing another run of quick change nozzles? I didn't get in on the last bacth becuase I couldn't afford at the time, but have just about go it in my budget now. Any idea how soon I might get a pair if I order?

The NEMA8 I currently have, has the wiring connector at the other end, i.e. at the front of the motor. It is a mirror image of this. It is not totally clear from the model if there would be clearance or not for the connector in the cut out hole or not. I might order a different NEMA 8 to be safe, but thought I would ask for your thoughts on whether it might fit or not if you have a physical copy in your hand.

Regarding the M3 tapped holes. Looking at this view, it looks like there might just be room for a small headed m3 bolt at the rear mounting holes negating the need to tap holes. What do you think? If so, I wonder if it would be a 1 minute job to move the holes toward the front out a mm or two so there is no tapping reqired? Is that possible do you think?

On the last point, I noted that Mark Harris is now offering cheap waterjetting in Aluminium, and if this is the case would be great to remove the need for tapping before people starting spending big money on Aluminium parts. This brings another complication in that Mark is offering to cut 1/4" stock and you have just modded it all to 6mm so you might end up with two versions. Sorry if this gives you a headache, it wasn't my intention!

Anyways, really appreciate your help, gonna start talking to people about getting some parts cut now, very excited. I am hoping to have all my gear before Christmas as I have a little break then when I want to get the basic machine built and motors turning at least.

Cheers,
Bobby

[Peter Betz]

Hi Bobby. Answers are inline:

Do you have a specific part number for the swich you have in mind? This is the swithc with the wheel on the arm. Just want make sure I order the correct part as I know they have different arm lengths and contact forces etc.

The part number is in the model.

Will you by any chance be selling your custom LED pcb for the camera system? Would be nice if I could get 4 of these, or even just the pcb when you order a few for yourself (assuming you haven't).

Yes I can certainly do that. I had lots of the PCB's made, so I can put them on the website. Do you have a printer or would you want the printed parts too?
 

Is the larger Motor a NEMA 17?

Yes.

 Is this a MGN12H block? If not, I suppose the more generic question is what are the pitches in X and Y of the mounting holes? Sorry if it sounds like a lazy question, but I don't know how to measure anything in the autodesk viewer (web browser).

You need to download fusion (its free). Yes it is a "long" block.

I was also wondering how close are you to doing another run of quick change nozzles? I didn't get in on the last bacth becuase I couldn't afford at the time, but have just about go it in my budget now. Any idea how soon I might get a pair if I order?

We are close, I believe we are booked in for the 5th of December or so (about 2 weeks late). I think Mark is going to be making an update shortly. You can purchase the nozzle holders here to secure your spot. We have a lot of preorders already and the run will only be so many.

The NEMA8 I currently have, has the wiring connector at the other end, i.e. at the front of the motor. It is a mirror image of this. It is not totally clear from the model if there would be clearance or not for the connector in the cut out hole or not. I might order a different NEMA 8 to be safe, but thought I would ask for your thoughts on whether it might fit or not if you have a physical copy in your hand.

Are you just saying the motor is rotated 180? I'm pretty sure there is only one hollow shaft NEMA8 motor...

Regarding the M3 tapped holes. Looking at this view, it looks like there might just be room for a small headed m3 bolt at the rear mounting holes negating the need to tap holes. What do you think? If so, I wonder if it would be a 1 minute job to move the holes toward the front out a mm or two so there is no tapping reqired? Is that possible do you think?

Those are the threaded holes, so the bolts go up from the bottom. This is a one of change for you.

On the last point, I noted that Mark Harris is now offering cheap waterjetting in Aluminium, and if this is the case would be great to remove the need for tapping before people starting spending big money on Aluminium parts. This brings another complication in that Mark is offering to cut 1/4" stock and you have just modded it all to 6mm so you might end up with two versions. Sorry if this gives you a headache, it wasn't my intention!

Bobby, I quickly modified the design so you could make one out of plastic. This has nothing to do with the Aluminum version that will no longer have those plates. I will be modifying the ALU design shortly to remove those plates.

Peter,

[Peter Betz]

An example of the switch (whose root part number is in the model) with 8gf and quick connectors is SS-01GL2-ET

Peter.

[Daniel Dumitru]

I have extracted the laser cutting files.

I will give them a try tomorrow. I am concerned that on the cut it's not exactly at 90degrees compared with surface.

BR

Daniel

​

[Peter Betz]

Yes, that “hourglass” cut you are referring to is exactly why I wanted to move to 5 axis water jet. 

Also, maybe don’t cut those files yet, I can get rid of one of those plates. Just might take me a day or two.

Peter.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAPU3koOvvtw4U2SzS6jbfGALYBg_P0uc2NtFoneixF4OCOXVrg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

<Peter Head R1.svg>

[Peter Betz]

FYI here is a set of waterjet parts :)

[Daniel Dumitru]

you made me jealous !  I will ask for a waterjet quotation locally ...

By the way , have you cut the pulley as well ? or is for testing purposes only ?

Kind regards,

Daniel

​

[Peter Betz]

I changed the 1/4" aluminum design. This design enables the new Z probe idea, or just allows you to bolt the NEMA17 in place. Aside from verifying the homing switch function this design should be ready for use.

Peter.

[Peter Betz]

On Saturday, November 19, 2016 at 9:51:03 PM UTC-8, Daniel wrote:

you made me jealous !  I will ask for a waterjet quotation locally ...

By the way , have you cut the pulley as well ? or is for testing purposes only ?

It was a test of sorts, we have some concerns about wear with ALU on ALU.

Peter.

[Daniel Dumitru]

VEry clever ideea on this. One single concern I have : what force will bring nema17 motor on "aligned" position ? 

I would add some lamelar springs to keep it aligned...

​

[Peter Betz]

On Saturday, November 19, 2016 at 10:15:39 PM UTC-8, Daniel wrote:

VEry clever ideea on this. One single concern I have : what force will bring nema17 motor on "aligned" position ? 

I would add some lamelar springs to keep it aligned...

It is the contact force of the switch that will keep it restrained (they are snap action switches). If additional force is required to hold the stepper stationary during normal movement, I will just keep ordering the next higher contact pressure switched (there are several in the same series).

One last image before bed:

[Bob G]

Peter,

Do you ever sleep?

Assuming you don't, thanks for all the responses / work, and let me respond to a few of your points:

Thanks for the switch part number, I will download Fusion and have a play with that, maybe then I won't have to ask you so many questions about your model :-)

About purchasing the PCB, I would like 4x PCB and 2x 3d printed bracket. Matt Baker has kindly offered to print me a few small parts, but he doesn't think these brackets will print well on his printer, so if you could sell me some, that would be fantastic. Please let me know the price and also, does the LED boards run on 12V? Would you also be offering the 3d printed vacuum mounts in your store? Any idea fo the price?

I will pre-order your Nozzle changer shortly, thanks. I am really hoping they will arrive before December 20th, is that likely? That is when I will start my build and have a solid week to work on it.

My Nema8 is the supplied with the FPD. It has the wires coming out of the front end. It is dual shaft and the shaft at the rear is very short. I will take the nozzle off tonight and see if the shaft is the same length at both ends, in which case, I just have it the other way up to you.

And finally, when you said:

"Yes, that “hourglass” cut you are referring to is exactly why I wanted to move to 5 axis water jet. 

Also, maybe don’t cut those files yet, I can get rid of one of those plates. Just might take me a day or two."

Are you saying you plan to modify the 6mm version to have the same arrangement as the 1/4" version for the motor mounts, i.e. remove the Z-probe Nema8 plate? If so, fantastic, as that gets rid of tapping holes. My earlier comments about moving the holes outwards, was I was think if there was room for the bolt head to sit at the top, you could fasten the nut underneath and you don't need to tap a hole. In the last version I looked at there probably isn't clearance for the bolt head, unless you move the holes towards the edge a bit more (even then there might not be room). If you are making it a single plate model, this point it defunct anyway.

Thanks,

Bobby

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.

To unsubscribe from this group and all its topics, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5666ab3d-6b65-4300-884b-670fc59fff39%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Regards,

Bob G

[Daniel Dumitru]

​I like a lot ideea having nema17 motor holding on his ax.

I would like some details : have you considered an axial bearing and a "bush collar" ?

BR

Daniel

[Trampas Stern]

I assembled Peter's head design and connected a Nano Zero Stepper board to see if I could detect force applied to the heads.  For the test I set the hold current to around 200mA on the motor and then used a spring to press on each head. 

Here is a video showing that it works.

https://www.youtube.com/watch?v=9BcHgHxcJtc

[Peter Betz]

On Tuesday, November 22, 2016 at 9:28:35 AM UTC-8, Trampas Stern wrote:

I assembled Peter's head design and connected a Nano Zero Stepper board to see if I could detect force applied to the heads.  For the test I set the hold current to around 200mA on the motor and then used a spring to press on each head. 

Here is a video showing that it works.

https://www.youtube.com/watch?v=9BcHgHxcJtc

Very interesting video. So you are tracking the deviation from the intended position as opposed to the current required to hold the position?

Peter.

[Trampas Stern]

Currently the hardware does not sense current but does sense the position. It basically can control position by changing current, but does not measure current directly.

What I did was limit current and measure position error. The best mental picture is that the motor has two springs holding motor shaft at desired angle/position, the K constant on the springs is the current. So in normal operation you want stiff springs, but when homing you can use weak springs, lower current, so as not to damage nozzles.

[betzt...@gmail.com]

Interesting, can you change motor current on the fly (theoretically being
sent commands by openpnp)?

Peter.

[Daniel Dumitru]

or better to store some profiles in driver and apply them on beggining of movement ?

On Wed, Nov 23, 2016 at 4:12 AM, <betzt...@gmail.com> wrote:

Interesting, can you change motor current on the fly (theoretically being sent commands by openpnp)?

Peter.

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/444EE1068D9548BE97543D94FD68560C%40BETZtechnik.

[Trampas Stern]

Peter-  yes the current can change dynamically.  The way the stepper driver works for microstepping is by changing the current in a sine/cos waveform. So for example if the motor is running 300RPM and is a 1.8 degree per step motor with 16 micro steps it means the current has to be changed. 330/(60*200*16)=0.0015 seconds, ie ~640 times a second.  

The controller's firmware is really neat in that with open loop operation and my vise grip load I can only get a 100-200RPM or so with 16 microsteps, with the controller in closed loop I get 300RPM. The firmware does this by adapting the step size to the error during closed loop operation, so at high RPMs in order to minimize error will move the motor in full steps. Hence the controller actually does dynamic step size, for example if the error is 1/256 of a step it will move 1/256 of step to minimize the error.  The microstepping setting of the controller is really just to tell it how much it should move when you toggle step pin, under the covers it always uses best step up to 256 microsteps.   

The peak amplitude of the sine/cos waveform dictates the maximum current into the motor. This maximum current is a parameter in the firmware. 

The firmware currently has two settings for the current one for the maximum current during movement, and one for the hold current. The idea is that the current to over come the inertia of the machine load is high when you start moving from a dead stop. However once you stop moving the current required to hold position is less.  Most stepper drivers (ie most step sticks) just have one current setting so it is set high all the time, which causes noise and heat. I did some experiments with various loads to figure out what would work out of the box.  For example on a Z axis you need larger hold currents than on an X or Y for most machines. The worse is a dynamic load like vise grips on motor, when horizontal the load is high when vise grips are vertical load is low.  The vise grip example is very hard for PID loops as the load changes which can lead to oscillations. 

My long term goal is to have the host controller (smoothie, Marlin, etc) communicate with the stepper controller.  For example when machine is powered on and before homing axis the host would set the currents (move, and hold) low and the run the axis to each end while monitoring the error line. When the error line is asserted it is at the limits.  Then the host would set the current back to desired levels for operation.  This is great for leveling the bed on machine, as that the sensor for the bed level is the head so all distances are relative to the head, which is what you want.  Note that it is possible in the controller via communication with host to have a double check on the limits, that is controller could prevent host from running machine into rails. However it theory since the controller is fixing any missed steps the host and motor will never be out of sync, assuming no bugs in the host firmware. 

With the error line (think alternative functionality of the enable pin on step stick) the host can constantly monitor the error pins and if one axis has an error, think about something stopping head, or several missed steps, then the host can stop the other axis until that one self corrects.  This feature might not be useful for PnP but on a printer or CNC mill what it would do is stop all action until that axis catches up.  An example for printer is that speed of one axis is too high and you are doing a diagonal move but the axis can not keep up, then the error line will be set and hold off the host until that axis catches up.  On a CNC mill think about trying to cut too fast, here the motors are trying to drive bit into material so fast that it will break bit, If your current limits are set correct then it will not put enough force to break bit, but if you go faster than bit can cut it will flag and error. 

The development and testing of the stepper controller has taken me awhile, you can see the progression on my blog at www.misfittech.net. I am not done yet but it is getting real close,  my next steps are to assembly my PnP using the controller and implement communications with host, I have a Marlin RAMPS board at moment for this but may have to get a smoothie here soon.  I also want to put on my CNC for detecting tool height. 

On Tuesday, November 22, 2016 at 11:27:01 PM UTC-5, Daniel wrote:

or better to store some profiles in driver and apply them on beggining of movement ?

On Wed, Nov 23, 2016 at 4:12 AM, <betzt...@gmail.com> wrote:

Interesting, can you change motor current on the fly (theoretically being sent commands by openpnp)?

Peter.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

[Daniel Dumitru]

Theoretically this "driver implemented" Z - down limiter can be implemented as well using stall detection​. I will give it a try this weekend.

Daniel

[Peter Betz]

Significant update.....

The model is now fully PARAMETRIC in relation to stock thickness. I have updated the link on github,

This means that in the menu [modify > change parameters] you will be able to adjust the new "stk" expression which is default 6.35mm. If you change this value to 6mm, it will not only adjust the extrude command distances, but also MANY sketch features and joints so the model is properly rebuilt for your desired stock size. Things like bolt hole edge distances and offsets are all adjusted where required. Changing the stock thickness changed significantly more features than I thought.

This was hours of work but hopefully will make the model more user friendly for everyone.

Peter.

[Erik Grönvall]

wow! fantastic work!

[bobgee...@gmail.com]

Peter,

Excellent, well done and thank you!

I will get in touch with Daniel about cutting me one to try out, really excited now. Looks like lots of interesting discussion on the Z-probing too, really encouraging progress!

I am so close to pushing the button on my purchases now, just want to know a couple of quick points so I make sure I get it right:

Will you be able to sell me 4x LED PCB, 2x new vacuum mount, 2x large bracket for camera and LED (I want two so I can use the same for up and down vision)? Any idea of the costs?

What voltage will the LED boards require? (I am currently planning to go for 12V everything rather than 24V but need to check that won't give me a problem with the LED).

Cheers,
Bobby

[Daniel Dumitru]

PEter, I am trying to generate DXF's in order to cut them.

I have an issue with NEma 8 mount (both 1 and 2).

Square hole that allows assembly with rack-z-axis has a small bug and I don't know how to make it to be correct.

Thank you for all your work.

Could you please tell us as well the code of needle bearing that you have planned to use ?

BR,

Daniel

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/dae7284a-ca2f-4e46-9089-ef434d079f3e%40googlegroups.com.

[Peter Betz]

Hi Daniel,

According to my system it appears as though you downloaded the wrong model.

Use the GitHub link in the post I made. Scroll down to the A360 link under "revision 1".

Peter.

[Daniel Dumitru]

I have deleted it and import it again. I am getting a lot of warnings and one error. 

Warning: ImplicitJointOrigin5

<b>1 Reference Failures</b><br/>The body reference is lost, try editing this feature to reselect the lost body.

<b>2 Reference Failures</b><br/>The model is using cached geometry to solve. Please reselect reference geometry for failed features in the timeline.

Warning: ImplicitJointOrigin8

<b>1 Reference Failures</b><br/>The body reference is lost, try editing this feature to reselect the lost body.

Warning: ImplicitJointOrigin9

<b>1 Reference Failures</b><br/>The body reference is lost, try editing this feature to reselect the lost body.

<b>2 Reference Failures</b><br/>The model is using cached geometry to solve. Please reselect reference geometry for failed features in the timeline.

Warning: Position1

Compute Failed

Warning: ImplicitJointOrigin11(ImplicitJointOrigin3)

<b>1 Reference Failures</b><br/>The body reference is lost, try editing this feature to reselect the lost body.

<b>2 Reference Failures</b><br/>The model is using cached geometry to solve. Please reselect reference geometry for failed features in the timeline.

Error: RigidGroup85

Compute Failed

Warning: RigidGroup89

Compute Failed

Warning: Sketch3

Project2

Project4

Project5

Project6

Project7

Project8

Project9

Warning: Sketch10

Project2

Project3

Project4

Project5

Warning: Sketch1

Compute Failed

Warning: CopyPaste M3 nut (x4):3

Compute Failed

Warning: CopyPaste M3 nut (x4):4

Compute Failed

Warning: Position356

Compute Failed

Warning: ImplicitJointOrigin4(Rigid145)

<b>2 Reference Failures</b><br/>The model is using cached geometry to solve. Please reselect reference geometry for failed features in the timeline.

Warning: Sketch16

Project2

Project3

Project4

Warning: ImplicitJointOrigin21(Rigid154)

<b>2 Reference Failures</b><br/>The model is using cached geometry to solve. Please reselect reference geometry for failed features in the timeline.

Warning: Position384

Compute Failed

Warning: Sketch2

Project3

​

[Daniel Dumitru]

Hello All,

Please disregard previous message and sorry for posting all warnings. 

Model it's loading and performing right.

I will laser-cut them soon.

Thank you again PEter !

DAniel

[Trampas Stern]

Peter,

I was connecting my belts to the Revision 1 head design and realized in the revision 2 you added belt connections. I would recommend that rather than screwing belts to plate instead just add a slot for belt to loop through.  

It looks like the the revision 2 the screws are used to tension the belt and/or clamp belts. The tension of the belts can be done by moving idler and/or stepper mounting plates. The clamping of the belts is not needed instead just loop belts and wire tie (zip tie) the belt to it's self.  Therefore the motor plate just needs a slot to loop the belt through. 

Also I 3D printed a toothed idler for the belt rather than using smooth idlers, not sure this makes a difference or not, but figured it would reduce belt wear.  

Trampas 

[Peter Betz]

Hi Trampas,

I think you may be misunderstanding how both the V1 and V2 belt attachments work. In neither do the belts get held with screws. In the original head, you use the brackets from anthony's frame design which have slots for the belt to go through. Those are attached to the head with bolts, which allow it to be tensioned. In both designs you do not use tie wraps. You use small pins that the belt wraps around and goes back through the slot.

Here is a rough image of how it works. The teeth lock as they pass through the slot:

Then you screw the bolts in to tension the system. This allows for very precise tensioning compared to just sliding the idler.

Peter.

[oliver jackson]

Where do you suggest I get the rails from?  Noulei have the right size but C class.  Robotdigg have their own "high quality" replacements and supposedly genuine Hiwin.  Ali express have nothing better than the robotdigg "high quality" ie fake Hiwin.  Another option is to go somewhere I can easily configure like Misumi.

I'm leaning towards Noulei as they are likely to be the best quality other than Misumi.  Do you think that being C will matter? 

I'm going to do an order at robotdigg after chinese new year for the hollow steppers and the 0.8M 16 tooth pinion gear.  Anything else I should get from there for the head?

0.8M 16 Tooth Pinion Gear 4 OpenPnP

0.8M 16 Tooth Pinion Gear 4 OpenPnP

[Michael Anton]

Aliexpress has lots of linear guides that claim to be genuine Hiwin.  You just have to search for what you are looking for, with Hiwin in the search, and make sure they are from Taiwan.  One of the suppliers for genuine Hiwin on Aliexpress is Noulei.  I would ask them if they have anything better than C class, if that is what you want.  If you leave enough adjustability in how you mount the blocks and rails, you can probably tune out most of the problems with C class guides.  Look at the Hiwin specifications to see exactly how the precision classes affect the dimensions.  The running precision which you do care about, is really good for C class as well.

Mike

[Peter Betz]

Make sure the listing says "HIWIN" or it is a replica. Overn Noulei will sell you both. C has running parallelism of 0.0007" over 500mm etc. Noulei will get you ANYTHING from Hiwin, even custom.

Email Sven at Noulei and tell him Peter Betz sent you, he will look after you :) 

[Anthony Webb]

I will echo Peter in say Sven was very great to work with, would use him again without hesitation. 

Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/e1e8fa41-f48b-4c2f-b85e-e95d55fabcd6%40googlegroups.com.

[Mark Harris]

Noulei are good :) they will get you whatever class and preload you desire. C class are fine.

On Jan 17, 2017 5:32 PM, "Anthony Webb" <anthon...@gmail.com> wrote:

I will echo Peter in say Sven was very great to work with, would use him again without hesitation. 

Sent from my iPhone

On Jan 17, 2017, at 5:11 PM, Peter Betz <betzt...@gmail.com> wrote:

Make sure the listing says "HIWIN" or it is a replica. Overn Noulei will sell you both. C has running parallelism of 0.0007" over 500mm etc. Noulei will get you ANYTHING from Hiwin, even custom.

Email Sven at Noulei and tell him Peter Betz sent you, he will look after you :) 

sv...@noulei.com

On Tuesday, January 17, 2017 at 3:59:38 PM UTC-8, Michael Anton wrote:

Aliexpress has lots of linear guides that claim to be genuine Hiwin.  You just have to search for what you are looking for, with Hiwin in the search, and make sure they are from Taiwan.  One of the suppliers for genuine Hiwin on Aliexpress is Noulei.  I would ask them if they have anything better than C class, if that is what you want.  If you leave enough adjustability in how you mount the blocks and rails, you can probably tune out most of the problems with C class guides.  Look at the Hiwin specifications to see exactly how the precision classes affect the dimensions.  The running precision which you do care about, is really good for C class as well.

Mike

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/e1e8fa41-f48b-4c2f-b85e-e95d55fabcd6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/9C0FBB14-FB87-4FEC-BCA3-56DCF6828633%40gmail.com.

[oliver jackson]

Waiting on my first order from them at the moment for my delta printer. Good to know they can get hold of other options. I will contact Sven when they're back from their holidays as I know the order won't go through now. Glad you mentioned preload actually-does it make sense to get a light preload?

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAJMBTyRydT1J_WuiFwmvLek7sLO9hRRVsE5EGz5FEhMgeEhG%2Bg%40mail.gmail.com.

[Peter Betz]

Yes go with "very light preload" I think that's Z0?

Peter.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAJMBTyRydT1J_WuiFwmvLek7sLO9hRRVsE5EGz5FEhMgeEhG%2Bg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/08648963-55D8-48FD-9A8E-40DEC8C9DB99%40gmail.com.

[Mike Menci]

Hello, 

This is old post but here enclosed a solution for identical parts - twins

See picture. 

How have you solved the leak of vaccum trugh the bearings of Nema 8 ? anybody ??

Thnaks

Mike

On Wednesday, 18 January 2017 03:24:49 UTC+1, Peter Betz wrote:

Yes go with "very light preload" I think that's Z0?

Peter.

On Tue, Jan 17, 2017 at 5:54 PM, oliver jackson <analogmon...@gmail.com> wrote:

Waiting on my first order from them at the moment for my delta printer. Good to know they can get hold of other options. I will contact Sven when they're back from their holidays as I know the order won't go through now. Glad you mentioned preload actually-does it make sense to get a light preload?

On 18 Jan 2017, at 00:50, Mark Harris <ma...@rris.com.au> wrote:

Noulei are good :) they will get you whatever class and preload you desire. C class are fine.

On Jan 17, 2017 5:32 PM, "Anthony Webb" <anthon...@gmail.com> wrote:

I will echo Peter in say Sven was very great to work with, would use him again without hesitation. 

Sent from my iPhone

On Jan 17, 2017, at 5:11 PM, Peter Betz <betzt...@gmail.com> wrote:

Make sure the listing says "HIWIN" or it is a replica. Overn Noulei will sell you both. C has running parallelism of 0.0007" over 500mm etc. Noulei will get you ANYTHING from Hiwin, even custom.

Email Sven at Noulei and tell him Peter Betz sent you, he will look after you :) 

sv...@noulei.com

On Tuesday, January 17, 2017 at 3:59:38 PM UTC-8, Michael Anton wrote:

Aliexpress has lots of linear guides that claim to be genuine Hiwin.  You just have to search for what you are looking for, with Hiwin in the search, and make sure they are from Taiwan.  One of the suppliers for genuine Hiwin on Aliexpress is Noulei.  I would ask them if they have anything better than C class, if that is what you want.  If you leave enough adjustability in how you mount the blocks and rails, you can probably tune out most of the problems with C class guides.  Look at the Hiwin specifications to see exactly how the precision classes affect the dimensions.  The running precision which you do care about, is really good for C class as well.

Mike

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/e1e8fa41-f48b-4c2f-b85e-e95d55fabcd6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/9C0FBB14-FB87-4FEC-BCA3-56DCF6828633%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAJMBTyRydT1J_WuiFwmvLek7sLO9hRRVsE5EGz5FEhMgeEhG%2Bg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.

To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

[Peter Betz]

Hi Mike,

We don't have any issues with leakage through the stepper bearings. I think the only people who do are using pumps below their rated voltage etc. 

Peter Betz

BETZtechnik.ca

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/ff8e0530-1b22-45f9-96bb-3b2f98aa47e8%40googlegroups.com.

[Mike Menci]

Hello Peter, 

Thanks for fast reply 

Well I am using this:

MEDO VP0140 Linear Vacuum Pump

This MEDO VP0140 Linear Vacuum Pump is used and in excellent condition.

Configuration:
• Attainable Vacuum: -15.7 in.Hg
• Free Air Displacement: 0.106 cfm
• Voltage: 115 VAC/230V

Part Number: VP0140-V1006-D2-0511 

Which one do you recommend?

Thanks

Mike 

[Trampas Stern]

I use this vacuum pump, so far no problems:

http://www.ebay.com/itm/DC-12V-Micro-Air-Pump-Electric-Pumps-Pumpping-Vacuum-Pump-Booster-For-Medical-/252826439054?hash=item3adda14d8e:g:SpEAAOSwzgBY02yl

For the NEMA8 I have made a 3D printed plate system which uses an 0-ring such that the NEMA 8 bearings are not exposed to vacuum.  This way I have no leaks and do not need large vacuum pump. 

The cap system I made is so that I can put a magnetic encode on the NEMA 8 (which I have not completed yet). 

[Mike Menci]

Make sure your friction is very low on gasket to Nema8 shaft or stepper will not turn...

[SMdude]

This is how I adapted air to the nema 8.
Make sure your oring is not a tight fit on the shaft and make sure it is a little bit taller than the oring spacer plate so the oring squashes a little bit, pushing a bit harder on the shaft.
And use a little grease.
The way I have done it, prevents any strain (that causes your motor mount to flex a little bit) on the motor from the air hoses and motor wires.

[Peter Betz]

On Friday, June 2, 2017 at 10:42:27 AM UTC-7, Mike Menci wrote:

Hello Peter, 

Thanks for fast reply 

Well I am using this:

MEDO VP0140 Linear Vacuum Pump

This MEDO VP0140 Linear Vacuum Pump is used and in excellent condition.

Configuration:
• Attainable Vacuum: -15.7 in.Hg
• Free Air Displacement: 0.106 cfm
• Voltage: 115 VAC/230V

Part Number: VP0140-V1006-D2-0511 

Which one do you recommend?

Thanks

Mike

Hey that is a neat pump. Is it a diaphragm pump?  I just use the 24v one robotdigg uses and the 3d printed vacuum gasket that Anthony came up with (I modded it to accept a push-to-connect fitting though), works great.

Peter.

[Mike Menci]

Hi yes Diaphragm pump, but the volume it sucks seams to be to small.... 

So I found this on eBay - 

http://www.ebay.com/itm/232333391361?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

and this 

http://www.ebay.com/itm/181602535662?_trksid=p2060353.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Wiring:

• Red - Positive supply (+24VDC)
  
• Black - Ground (0VDC)
  
• Green - Tachometer output (0-5VDC)
  
• White - PWM or analog control (0-5VDC)
• NOTE: PWM frequency is set by the manufacturer. Please see their documentation. Incorrect PWM frequency drive may damage the unit. Analog control means you can drive a 0-5V signal into the device and control its speed (no need to alter the supply voltage).

I tested the smaller one today works great - speed can be regulated to control  the volume... 

BR

[Jason von Nieda]

Those look very interesting. How's the noise level?

Jason

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/1cbc41ac-bb88-4612-aa0e-72f0c624439c%40googlegroups.com.

[Mike Menci]

Hi 

Not to noisy - motor runs very quiet (brushless motor) but from strokes you can hear noise - coming with two rubber support bumpers to reduce vibration

[Jason von Nieda]

Nice find. I think I'll try one. Thanks!

Jason

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/781f8101-3b84-4b6d-ab1e-51a6a7148070%40googlegroups.com.

[Mike Menci]

Schwarzer Precision Pumpe - Google

[Friedrich Mäckle]

I want to build a head like this, thank you for the great design!

I want to have it water jetted from 6 or 8mm Al, so I am looking for a manufacturer. I know that there are some people around doing stuff like this or maybe have recommendations about where to manufacture.

The location is Germany, so Europe would be great. Also if anyone else is interested in a set get in touch!

Am Mittwoch, 16. November 2016 06:16:06 UTC+1 schrieb Peter Betz:

Hi Everyone,

 Sorry I have been pretty quiet lately. Things are still chugging away in the background. First of all, this is a continuation of the original head design found here:

https://groups.google.com/forum/#!topic/openpnp/-XGpRTdaTrk ("Head Thread")

The design is perfectly fine and lends itself well to laser cut acrylic (or delrin if you are super fancy). My machine is currently running this head.

I decided to make a couple of improvements to the design (and change the default material to aluminum), enter REVISION 1:



1. New designed top plate with cable chain attach points:

The top plate has been designed as a stronger method of attaching the cable chain. The chain exerts more force on the head than I expected! This design also allows mounting holes for components on top of the head.



2. New vacuum gaskets:

In an attempt to improve Anthony Webb's gasket design, I have changed it to allow for screwing in a NPT push-to-connect fitting (for 4mm or similar vacuum line).



3: Integrated belt tensioners:



With integrated nut traps and everything being cut out of stock 1/4" material (waterjet or laser) this simplifies the design by removing the need to drill and tap the side faces to attach brackets. Special thanks to Mark Harris for helping zero in on the final design.

4. Away with the opto switch and fragile thin feature.

In order to make the design more robust, I decided to try using a high quality snap action limit switch instead. This makes the design slightly more compact and also does away with that very thin interrupter feature. Using contact sets in Fusion I was able to simulate the tripping of the switch and find the ideal mounting angle for it! These switches are rated for millions of cycles MTBF, so I consider this to me a long term, reliable solution.





5: I had a strong desire to try and integrate Z probing.

First, there was to be two plates at the bottom of the NEMA8 steppers. The upper plate is fixed to the stepper motor mounting screws and the bottom is fixed to the toothed plate attached to the linear rail. Motion is supported between the two plates along 4 precision ground shoulder screws:



Then, a new designed vacuum gasket would allow for contact with a microswitch:



Here is a GIF showing how it was designed to function:

Mark went ahead and had it cut on the waterjet to prototype. Long story short is that the required tolerances on the shoulder bolt holes is too great and would require a more precise manufacturing method, like CNC milling to maintain part placement accuracy. Unfortunately I don't think that it will be worth pursuing this method any further :(

Revision 2 is only modifying the Z probe style.

My new thought is to have the Z stepper completely unrestrained in its rotational axis (no mounting bolts securing it). The stepper shaft would pass though a flanged bearing in the main plate, and the stepper face would press up on a caged bearing. There would be a printed washer that went under the gear to allow the stepper to be drawn in to the head plate using the set screw on the gear (if that makes sense). Here is a simple mock-up:

The stepper motor would only be restrained in rotation by the contact force of two microswitches mounted on the back of the main plate:

In the normal course of operation, the only force that would be exerted on the switches is from acceleration, because the head is balanced. When the nozzle contacts something that provides resistance, it overcomes the contact pressure of the switch and it triggers.

Where we are at right now is proof of concept. I have a full set of metal parts here, and I have some switches with about the lowest contact pressure available. I am going to take my head apart and build it up with the new metal parts and also make a small 3D printed shoulder bushing to restrain the Z stepper similar to how the bearings would. I can then actually see the result on my machine and see if this method is viable. I actually have pretty high hopes for this. Any of the play in the Z system is not really that important with the spring loaded nozzles, at least that's how I feel about it!!

Now, thanks to Mikael, I really need to step up my rendering game.... sigh.

That's all for now, hope you guys like the changes.

Peter.

[Peter Betz]

Hi Friedrich,

I was actually thinking about asking the group if there was interest in a group buy. The one issue is that you really need a 5 axis waterjet for the parts to sit perpendicular. My plan is to waterjet on a standard machine, then machine the critical surfaces on my CNC mill. 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/89dce844-4c0f-45f1-857b-0e24a938a6fb%40googlegroups.com.

[oliver jackson]

I may be interested in a group buy, depends on cost. It's something I have in mind. Please update if you get anywhere with this

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/41C117EE-266B-4B68-A010-956355B3E9CF%40gmail.com.

[Peter Betz]

Ok. I will look into this. It's something I wanted to offer in the web store (of course it is completely open source as well). The price should be similar to the robotdigg one and maybe $60 more for a set of the quick change juki nozzle holders, clocked and tested on the NEMA 8's. 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

On Jul 31, 2017, at 10:22 PM, 'Friedrich Mäckle' via OpenPnP <ope...@googlegroups.com> wrote:

--

[Daniel]

Hello Friedrich ,

As I have offered before if you like to test it with a simple lasercut version from acrilic or hdf in order to test it, I offer this for free.

You have to pay only for shipping inside Eu.

REgards,

DAniel

[Peter Betz]

Just be aware there are some additional complications with having the design cut on a laser or 3 axis waterjet. It affects the perpendicularity of the nozzle face. I ran a laser cut version for a while when I first made it. It would probably not be an issue for people not looking for extreme accuracy out of their machine. I suppose you would even reduce the error with a file on the mating surfaces....

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/4bac2b75-4155-4782-8983-94626b8e90c8%40googlegroups.com.

[Friedrich Mäckle]

Hey Daniel,

this sounds great, I would like to come back to that offer, so I can get started with everything, also getting the machine running in the first place and then maybe later changing to a metal cut version.

Could you cut a set from the models Peter provided? What thickness of acrillic can you do? I am not sure if its better to do acrillic or hdf, I would go with your best guess.

I would be happy to pay the shipment and materials. How do we proceed here?

Best, Friedrich

[Friedrich Mäckle]

I would like to have one too. I already bought a pair of nuki holders and I like them.

For me the only issue is time. I would hope to get them in ~8 weeks, because otherwise I would probably try to source them in Germany.

[Peter Betz]

Ok, let me try and get some pricing together for everyone! 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/dddb29ba-c9f0-44ee-b3e5-2b96c2eb9a44%40googlegroups.com.

[Daniel Dumitru]

Hello Friedrich ,

Just send me on private (hoping that you could see my email address ) your physical mailing address. I can cut out of 6 mm thickness.

Bye,

Daniel

[Trampas Stern]

I have done my own head design and had a few comments.

1. Parts should be aluminum - I ran into repeat ability problems with plastics. 

2. Juki nozzles have springs  - This means the nozzles have built in tolerance for Z height being off. 

3. Z height detection - Autotronix uses a vacuum sensor for Z height detection, with spring in Nozzle this maybe best method

4. Minimize moving parts - all moving parts have slop which can cause errors. 

So before I would not do complex mechanical system for Z height detection,  before just adding a vacuum sensor.  

Even more to the point Z height errors are not the main problem I have ran into.  My main problem with head designs is the mechanical X/Y repeatably and nozzle run out.  I would be interested in hearing if anyone really has problems with Z height detection after their machine is up and running?  That is I don't think this is a real problem that can not be solved with a vacuum sensor. 

If I were designing a head today I would make it all out of aluminum and add some Raspberry pi zero cameras under each nozzle like Autotronix does such that each nozzle has up looking camera and can do rotation and alignment while moving head. By using Raspberry Pi Zero the rotation and alignment can be done concurrently (on the raspberry pi zeros).  Doing the camera mechanics under nozzle is a bit complex but is easily doable. This would address the run out problems and make machine much more robust and faster.  I would still have an up looking camera on machine base for large parts, but for caps and resistors the simple on the head camera would be great. 

If enough people are interested I could do the design and setup a group buy on the parts.  

 

 

[Peter Betz]

Hi Trampas,

Vacuum Z probe has been discussed but the issue is setting part heights. On small parts the vacuum will suck up the part before establishes the height. I assume Autotronix does not use this method for part setup? On commercial machines I assume the part locations are automatic as it knows the exact location of each feeder (and therefore the location of the part within that feeder). 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

--
You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/75e2c9ef-80a2-483d-8504-d6fce5696418%40googlegroups.com.

[Trampas Stern]

For the Autotronix the video I saw suggested doing the height adjustment on the tape holder not the part.  The indication in the video was that for tape and reel parts, the top of the part should be the same as the top of the tape holder.   This seems to work very well for the ~40 feeders I have installed.

[Michael Anton]

The machine I run, has a spring on the mechanism that lowers the nozzle.  Z height is set by lowering the nozzle manually, until you see it touch the part.  This is actually pretty easy to do, and quite fast, once you have done it a few times.  I find actually targeting the part locations in X, Y, and Z, to be the fastest part of the machine setup.  It takes far more time to actually load the feeders.

With the Juki nozzles, you could just lower the nozzle, until you see the spring start to compress.  That should be close enough.  Though, I am trying to work out a way to sense this for my design, so it can be automated, though in reality this is not all that necessary.

[Trampas Stern]

I agree that Z height testing is really not needed, and where it is vacuum sense is good enough. 

Personally after trying to use my machine without vacuum sense and plugging nozzles with solder paste when pick failed, I feel that vacuum sense is a requirement  (ie sensing when part is on nozzle) . 

[Friedrich Mäckle]

Hello Peter,

I saw that you changed the Fusion360 drawing with a bunch of sets. Is there any update?

Best, Friedrich

[Peter Betz]

Hi Friedrich,

Very keen observation! Yes I did that to have a waterjet panel quoted locally. I hope to hear back today. I will keep you guys posted. 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/aa44633a-b6c0-4bba-8443-d3d578d7ea5b%40googlegroups.com.

[Peter Betz]

Just a quick update, I wanted to verify my quote from my genuine HIWIN supplier, and the price of HIWIN has increased twice in the last 12 months!! My old cost was $19 per MGN9x75 and now it is nearly $35 per unit!! It really makes you consider the knock off's, but having zero play on the rails is pretty critical....

Still waiting for the waterjet quotes but that is the last piece of the puzzle.

Peter.

[Friedrich Mäckle]

Yes the prices went up quite a bit. I paid about the same mid July as you were quoted now.

[Peter Betz]

Hi Everyone,

Finally I have all the pricing together!! I'm actually excited about this as I have wanted to make a batch for some time now. I have tried to make the pricing as good as possible as this is something I would like to continue doing (and I hope it is cheap enough that people who have laser cut versions would want to upgrade to metal). As I always seem to do, I have likely priced this too low, and way to tight, and next run the price will realistically be a little more, so I'd suggest taking advantage of jumping on this run of 10 heads! This will also spur me into continuing my research on Z probing which these heads will have the ability to upgrade to (just by bolting on a few parts that will fit with no mods) if it proves successful. 

*****************

Price without our Juki Nozzle holders: $219 USD

*****************

Price with a set of Juki Nozzle holders: $319 USD

*****************

I will discuss a discount for the nozzle holders with Mark when purchased with a head but those are super expensive parts to have made, due to the tolerances, but the results are well worth it!

What you will get:

- One fully assembled V2 head (without the extra hardware for Z probing) complete with NEMA 17 Z axis motor, 2 NEMA 8 hollow shaft part motors, 2 linear rails, genuine micro switch for homing etc. I will even include mounting screws for the X bearing block. 

- waterjet from 1/4" ALU to 0.005" tolerance. 

- I will be fixturing the waterjet parts and CNC machining the required mating surfaces to have all the parts fit as perpendicular as possible. 

- I will also experiment with CNC machining a shallow pocket for the MG9 linear rails. This would create a very accurate reference edge for each rail to ensure the rails are completely parallel  to each other and perpendicular to the X axis rail. 

* The importance of the above features really become apparent when you start dialing the machine in for smaller parts. 

- Fully tested ! I will be performing a variety of tests to ensure that the assembled head meets the quality I would expect if I were purchasing it. This will be especially important with the rails, more on that later. 

If you order them with our Juki Nozzle holders, I will also fully test them and set the nozzle holders to the lowest possible runout configuration and ensure runout is within our specified range. 

Shipping would be extra but I will have economy and express options. I do not price shipping to make money, so it will be the real cost plus a little for packaging materials etc. 

I could also probably do custom backplates for people looking to put this on a machine that doesn't have a 12mm linear rail on the X axis. The kit will not have the 3D printed parts included by default (NEMA 8 vacuum gaskets and the camera mount) but I can look at making those available for those who are interested. 

Like I was saying, I'm not sure I will be doing this again for that exact price (assuming there is interest) (I'm a perpetual optimist when it comes to pricing a job :) )but I'm excited about the thought of getting this design out there made of aluminum !!! I'm hoping to get at least 4 people committed before I pull the trigger on spending a couple thousand dollars to make the 10 units.

Pictures for effect:

I will also create a dedicated thread for this for those that aren't subscribed. 

Please indicate if you are interested. Thanks everyone! 

[Thomas Langås]

I'm interested in this  (I have juki nozzle holders from you, already), but I'm not sure if I need to modify the top a bit.  I have 80x80mm extrusion as X-axis, with a linear rail on one side, cable guides on top, an ball screw on the backside.

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0FE975A9-D97F-449E-82EE-04813B4B1DBF%40gmail.com.

[Peter Betz]

Hi Thomas,

I'm sure we can make something work. Do you use google hangouts ? 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

On Aug 21, 2017, at 12:21 PM, Thomas Langås <thomas....@gmail.com> wrote:

I'm interested in this  (I have juki nozzle holders from you, already), but I'm not sure if I need to modify the top a bit.  I have 80x80mm extrusion as X-axis, with a linear rail on one side, cable guides on top, an ball screw on the backside.

On Aug 21, 2017 21:13, "Peter Betz" <betzt...@gmail.com> wrote:

Hi Everyone,

Finally I have all the pricing together!! I'm actually excited about this as I have wanted to make a batch for some time now. I have tried to make the pricing as good as possible as this is something I would like to continue doing (and I hope it is cheap enough that people who have laser cut versions would want to upgrade to metal). As I always seem to do, I have likely priced this too low, and way to tight, and next run the price will realistically be a little more, so I'd suggest taking advantage of jumping on this run of 10 heads! This will also spur me into continuing my research on Z probing which these heads will have the ability to upgrade to (just by bolting on a few parts that will fit with no mods) if it proves successful. 

*****************

Price without our Juki Nozzle holders: $219 USD

*****************

Price with a set of Juki Nozzle holders: $319 USD

*****************

I will discuss a discount for the nozzle holders with Mark when purchased with a head but those are super expensive parts to have made, due to the tolerances, but the results are well worth it!

What you will get:

- One fully assembled V2 head (without the extra hardware for Z probing) complete with NEMA 17 Z axis motor, 2 NEMA 8 hollow shaft part motors, 2 linear rails, genuine micro switch for homing etc. I will even include mounting screws for the X bearing block. 

- waterjet from 1/4" ALU to 0.005" tolerance. 

- I will be fixturing the waterjet parts and CNC machining the required mating surfaces to have all the parts fit as perpendicular as possible. 

- I will also experiment with CNC machining a shallow pocket for the MG9 linear rails. This would create a very accurate reference edge for each rail to ensure the rails are completely parallel  to each other and perpendicular to the X axis rail. 

* The importance of the above features really become apparent when you start dialing the machine in for smaller parts. 

- Fully tested ! I will be performing a variety of tests to ensure that the assembled head meets the quality I would expect if I were purchasing it. This will be especially important with the rails, more on that later. 

If you order them with our Juki Nozzle holders, I will also fully test them and set the nozzle holders to the lowest possible runout configuration and ensure runout is within our specified range. 

Shipping would be extra but I will have economy and express options. I do not price shipping to make money, so it will be the real cost plus a little for packaging materials etc. 

I could also probably do custom backplates for people looking to put this on a machine that doesn't have a 12mm linear rail on the X axis. The kit will not have the 3D printed parts included by default (NEMA 8 vacuum gaskets and the camera mount) but I can look at making those available for those who are interested. 

Like I was saying, I'm not sure I will be doing this again for that exact price (assuming there is interest) (I'm a perpetual optimist when it comes to pricing a job :) )but I'm excited about the thought of getting this design out there made of aluminum !!! I'm hoping to get at least 4 people committed before I pull the trigger on spending a couple thousand dollars to make the 10 units.

Pictures for effect:

<image1.JPG>

I will also create a dedicated thread for this for those that aren't subscribed. 

Please indicate if you are interested. Thanks everyone! 

Peter Betz

Betz Technik Industries Ltd. 

BETZtechnik.ca

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+unsubscribe@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0FE975A9-D97F-449E-82EE-04813B4B1DBF%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to a topic in the Google Groups "OpenPnP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openpnp/qdB5XmdrgOA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CANN7TQmzW--BmHJ7tjPhBYxpaa9CkGBivPV3UhNXoogGNRxyXA%40mail.gmail.com.

[Friedrich Mäckle]

Thank you Peter for setting this up, I placed my order for the plates!

To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0FE975A9-D97F-449E-82EE-04813B4B1DBF%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

[Friedrich Mäckle]

Dear Peter, as Daniel was so kind to cut a HDF version, the only part missing now is the spur gear. So far I found no matching one, could you send me a link to a gear that fits in?

[Peter Betz]

Hi Friedrich,

This is the one: https://www.robotdigg.com/m/product/828/0.8M-16-Tooth-Pinion-Gear-4-OpenPnP#