Head thread (PNP head design)

4,796 views
Skip to first unread message

Peter Betz

unread,
Feb 14, 2016, 3:45:56 PM2/14/16
to OpenPnP
Hi Guys, 

Moved to a new thread to discuss the head I will implement for Anthony's PnP design. I don't want to buy the $300 robotdigg one.

Here are the two contenders:

1. A more compact and direct to carriage mounting robotdigg style head (tried and true):






2. A belt driven setup that will force down and force up (I will model this one to compare in the next day or so, should be really easy and use less parts.....), another variant would let gravity move it down:


Peter Betz

Anthony Webb

unread,
Feb 14, 2016, 8:48:27 PM2/14/16
to ope...@googlegroups.com
My vote would be for the tried and true. I don't think there is enough to be gained by the time you have to deal with attaching the belt, level before you turn on machine (or leveling via end stop and software) the gravity approach would be finicky unless you went with some hand picked and expensive rails. 

I need to look closer at the compact version. The arm seems to be much longer than it should need to be.  It also will limit the X travel if it is too wide.  Can you reach all 4 of those mounting holes without disassembly or do they need to be moved slightly up?

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/770ffe38-b033-42cb-aa62-cbd3172c44a3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Peter Betz

unread,
Feb 14, 2016, 9:42:38 PM2/14/16
to ope...@googlegroups.com
Yes I think you are right. I will focus on that design. 

Yes your comments are true, I haven't tweaked the design yet. I am going to start over there were some dependency issues as I started with the robotdigg schematic as a canvas to start. 

Did you buy your rails from Sven? He said he has the grease nipples for the MGN carriages. I saw them on the data sheet so I asked about them. 

Peter Betz
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/-XGpRTdaTrk/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.

Anthony Webb

unread,
Feb 14, 2016, 10:51:43 PM2/14/16
to ope...@googlegroups.com
The first 2 rails I got were already installed on the head from robotdigg. When the rail/slide stuck so bad that the spring would not pick up the head they sent me a replacement.  The replacement was not much better but I was able to make it work by cleaning and lubricating it. Now they sell replacements on their site but be advised they will be hit or miss. http://www.robotdigg.com/product/671/HIWIN+Origin+MGN9-1C-75+Linear+Rail

Sent from my iPhone

Anthony Webb

unread,
Feb 14, 2016, 11:35:01 PM2/14/16
to ope...@googlegroups.com
One other note. In looking at the side view there is not much room behind the nema17, this is because it is being mounted direct on the slide. On my machine (I need to make a proper closeup walk through video) the cablechain fits perfect on the X extrusion, even attaches using my existing bracket screws.  Be aware you may need to rethink cable management if there isn't an extrusion spacer in there.  Might not need the chain if you did some sort of stiff bundle wrap instead?

Sent from my iPhone

DAniel Dumitru

unread,
Feb 14, 2016, 11:44:18 PM2/14/16
to ope...@googlegroups.com
I consider that version 2 with belt driven up and down it's significant better.
For example using a stepper driver that has stall detection could be implemented very easy a control for component height
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+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/770ffe38-b033-42cb-aa62-cbd3172c44a3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


This email has been sent from a virus-free computer protected by Avast.
www.avast.com

betzt...@gmail.com

unread,
Feb 14, 2016, 11:53:37 PM2/14/16
to ope...@googlegroups.com
Do you mean the space between the bottom of the stepper and the top of the X extrusion?

Anthony Webb

unread,
Feb 14, 2016, 11:57:27 PM2/14/16
to ope...@googlegroups.com
Daniel, can you elaborate on why component height control would be useful?  In OpenPNP I set up the component height one time, not a big deal.  All my feeders deliver parts that are all picked at the same depth too. 

One useful thing long term may be to know exactly where the tip is at, especially if we ever get to automatic tip changing, but I imagine you would not need stall detection for that, you could use a simple pressure pad somewhere on the bed and touch off it to get an accurate enough idea of where the tip is at. 

Sent from my iPhone

Anthony Webb

unread,
Feb 15, 2016, 12:10:20 AM2/15/16
to ope...@googlegroups.com
Actually, you can disregard my comment. In looking closer at the photos of my machine I can see there is enough room.  The nema17 is pretty low profile. https://hackaday.io/project/9319/gallery#1da8149dcfea1bf57750630dc608c890

Sent from my iPhone

Cri S

unread,
Feb 16, 2016, 3:53:59 AM2/16/16
to OpenPnP
It help you if I publish plan for my 2/4 head. Just one ALU sheet to drill.
It uses nema8 motors.

Peter Betz

unread,
Feb 16, 2016, 6:48:38 PM2/16/16
to OpenPnP
Hi Cri,

 Sure, go ahead and share them, it would be nice to see other ideas.

On that note, I think I am going to go for the belt design, it seems better for a lot of reasons (to me anyway).

I would really like to do a stepper and belt for each nozzle, but I see that the options for 6 axis motion controllers are limited. I will give this more thought. The advantage is that the rails could be half the length and the other head doesn't have to move unless needed......

Stand by for some models.

Peter.

Cri S

unread,
Feb 16, 2016, 11:07:02 PM2/16/16
to OpenPnP


nema17 + nema8 steppers, 
20 pulley  and 126mm belt
9mm linear rails and 20mm round rails.
one 8mm holder holds camera,
43mm between rails center drills

Cri S

unread,
Feb 16, 2016, 11:13:33 PM2/16/16
to ope...@googlegroups.com
forget, movement is 22mm .

--
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/-XGpRTdaTrk/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/699fa763-5452-4208-8548-a8ddff8b6388%40googlegroups.com.

Anthony Webb

unread,
Feb 17, 2016, 12:09:35 AM2/17/16
to OpenPnP
What does the blue line represent?  Also curious what the round rails are for?  Also curious where the 8mm camera mount is?

Cri S

unread,
Feb 17, 2016, 6:07:44 AM2/17/16
to ope...@googlegroups.com
The blue represent the block used to mount it on y axis where 20mm round rail is used resulting in 50x50mm mounting holes and further using distance holders.
The camera is mounted using cots CNC round rail supports fitted from behind trought the quadratic hole and using the screw from rail too that need to be longer and fixed using nut from behind.  I know the mounting solution t y axis using distancials is probably not so good .
--
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/-XGpRTdaTrk/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.

Frank Herrmann

unread,
Feb 17, 2016, 10:41:13 AM2/17/16
to OpenPnP
Sorry for the offtopic question, which cad program you prefer or use for this design?

Anthony Webb

unread,
Feb 17, 2016, 10:43:35 AM2/17/16
to ope...@googlegroups.com
I think Peter like several of us use Fusion360

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.

Peter Betz

unread,
Feb 17, 2016, 10:58:10 AM2/17/16
to ope...@googlegroups.com
Correct. Ultimate version :)

Peter Betz
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/-XGpRTdaTrk/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

unread,
Feb 19, 2016, 1:45:21 AM2/19/16
to OpenPnP


Here is an update on the pulley head (currently about 6"  tall):








Thats not my final thought on the homing switch, but it is all I've got for the day!


Peter.


Jason von Nieda

unread,
Feb 19, 2016, 1:47:45 AM2/19/16
to OpenPnP
Hi Peter,

Looking good! I do have one thing you might want to keep in mind: Assuming you are using NEMA 8 hollow shaft steppers there for the nozzles, you will need some type of adapter on the top of the motor to pass the vacuum through. This may interfere with the top pulley, so I'd suggest modeling it up before you go too much further so you can make sure it won't be an issue.

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/e445083b-d7c3-46d4-8c6d-67d333e26973%40googlegroups.com.

Peter Betz

unread,
Feb 19, 2016, 1:58:17 AM2/19/16
to ope...@googlegroups.com
Yup. That totally didn't factor in. Thanks for the heads up. I have Anthony's adapter nozzle so I will throw that on and adjust as required. 

About the position on the rail, I have it mounted roughly half way, is there a preference to offset it up or down ?

Peter Betz
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/-XGpRTdaTrk/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.

Jason von Nieda

unread,
Feb 19, 2016, 2:04:11 AM2/19/16
to ope...@googlegroups.com
The offset would depend on the rest of the machine's design. You want to maximize Z travel while taking into consideration the height of objects on the bed. For instance, if you have a bed but intend to use magnetic board holders that are 10mm high, you need to take that into consideration as it will cut into your travel. So you'd move the whole head up 10mm.

I highly suggest at least mocking up the complete machine including fixtures, nozzles, PCB, feeders, etc. so you can accurately measure what you Z travel will be. 

Jason


To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/C574D6C8-ACD6-42BB-8CBF-874B030C3090%40gmail.com.

Michael Anton

unread,
Feb 19, 2016, 2:18:10 AM2/19/16
to OpenPnP
Looks really good, but I have a couple of suggestions.

I would use a 16T pulley if possible.  This will allow you to get the rails closer together, and increase the resolution.  On that note, it looks like the idler pulley is not the same diameter as the toothed pulley.  If this is the case, that is a not a good idea, as the belt won't be parallel to the axis of motion.  I usually try to use another toothed pulley as the idler, as the teeth running on a smooth pulley will cause some tension pulsing to occur which is less than ideal.  This is a bit more difficult to arrange though, unless you can find a pulley with internal bearings, but it can be done.

You might want to put in some sort of mechanical stop, so that the carriage blocks travel past the end of the rails.  We wouldn't want the carriages to hit these normally, but it may save someone a bunch of grief, if they could go off the rails (bearings can pop out etc.).

Mike

Rich Obermeyer

unread,
Feb 19, 2016, 2:24:52 AM2/19/16
to ope...@googlegroups.com
@peter, why the optical versus a mechanical switch?  Mechanical would simply bump up against the top of the motor mount.  I would be concerned the optical out in the wild will be subject to contamination.
Just my 2c.

Great job by the way!

Cri S

unread,
Feb 19, 2016, 7:05:28 AM2/19/16
to ope...@googlegroups.com

You surely don't want have limit switch when nozzle is maximum down. Travel seems to be a lot more then needed. What is distance motor nozzle tip including holder and overall travel.

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/-XGpRTdaTrk/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/4DC74D96-DAA4-4D19-B123-2E38F76E0CB7%40gmail.com.

Jason Parmenter

unread,
Feb 19, 2016, 10:32:52 AM2/19/16
to ope...@googlegroups.com
​Online with what Cri said, would it be better to make the optical end-stop in the middle of the motion when the two heads are at the same height.  Depending on homing operation and other things it could be dangerous to send the "home" command and forget that you have one head down.  Looking nice!

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAJGcfUgQsZawkAOu63%2Bz6EUG%3D8YeUDLFjEamPWAKyT4b_m-UJw%40mail.gmail.com.

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



--



JASON PARMENTER 

Electrical Engineering | 208.339.0822
ja...@tetontechnology.com
tetontechnology.com 





Anthony Webb

unread,
Feb 19, 2016, 10:38:13 AM2/19/16
to ope...@googlegroups.com
I dont think you will get one endstop precisely in the spot where both heads are at zero, I presume the idea of the endstop is to have a reference point that you can then apply and offset too to travel to that zero point.  In that case it doesnt matter where it is at because your driver will be only using it to use as a base location to level your heads with.  Basically in your driver the homing sequence would find optical home, then travel X mm back to true zero.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAP3w4qFN7wCaLMxkW-c5%2BaV9dT0z1uu266XqduYNHkw2w4m77Q%40mail.gmail.com.

HauntBots

unread,
Feb 19, 2016, 11:08:45 AM2/19/16
to ope...@googlegroups.com
The current thread on head design made me start thinking about my head.
I didn't want to hijack that thread so I started this one. My hardware
is completely different from what I've seen as the reference design and
I'm just wondering how much work will be involved in modifying code to
work with what I've got.

My head carries 3 nozzles,
The up down motion of the nozzles is operated pneumatically by 3
separate air valves.
However, all 3 nozzles are belt driven to rotate from one motor.
I haven't delved into the code but I was wondering if the software model
used to represent nozzles could be made to work with this.

Pete

Cri S

unread,
Feb 19, 2016, 11:09:37 AM2/19/16
to ope...@googlegroups.com
Usually refection optocopuler is used, or mechanical switch, because
it need know, if drive
the head up or down, safely to the position where both heads are approx leveled.
Otherwise if something is below nozzle, nozzle get damaged. Offsets
from Zero is then
defined inside OpenPnP or from driver if using G54 coordinates or elsewhere.
Example, there is power outage after when the nozzle that have placed
max height cap and is returned to safeZ.
Now assume that the nozzle that have placed the cap is the nozzle
without limit switch and
it go full down based on you'r design.
User need to do homing, otherwise openpnp don't move.
Hit home, result is, nozzle get damaged because of homing process, or
to name it better
because unsafe Homing switch design.


2016-02-19 15:38 GMT, Anthony Webb <anthon...@gmail.com>:
>>>> <https://lh3.googleusercontent.com/-ZbBDGzJDtgQ/Vsa5Z8trEjI/AAAAAAAAAgQ/QgqQrq0NnAo/s1600/Screen%2BShot%2B2016-02-18%2Bat%2B10.38.30%2BPM.jpeg>
>>>>
>>>>
>>>>
>>>> <https://lh3.googleusercontent.com/-KZ6V3Ay5H5A/Vsa5dIEZkyI/AAAAAAAAAgU/dmEvOQVBwsU/s1600/Screen%2BShot%2B2016-02-18%2Bat%2B10.38.44%2BPM.jpeg>
>>>>
>>>>
>>>>
>>>> <https://lh3.googleusercontent.com/-SgS1ir-lvd0/Vsa5gE9GyaI/AAAAAAAAAgY/KIKGibI4V0E/s1600/Screen%2BShot%2B2016-02-18%2Bat%2B10.39.00%2BPM.jpeg>
>>>>
>>>>
>>>>
>>>> <https://lh3.googleusercontent.com/-JowFx9MolF8/Vsa5i3P_0NI/AAAAAAAAAgc/T30I0j4v900/s1600/Screen%2BShot%2B2016-02-18%2Bat%2B10.39.14%2BPM.jpeg>
>>>>
>>>>
>>>>
>>>> <https://lh3.googleusercontent.com/-CC1AblgiGUA/Vsa5lZSVvMI/AAAAAAAAAgg/qT52cwYvtOI/s1600/Screen%2BShot%2B2016-02-18%2Bat%2B10.39.26%2BPM.jpeg>
>>>>
>>>>
>>>>
>>>> <https://lh3.googleusercontent.com/-RvuDjZK3FnM/Vsa5xWC_BMI/AAAAAAAAAgo/sI-tzVFvaTg/s1600/Screen%2BShot%2B2016-02-18%2Bat%2B10.43.58%2BPM.jpeg>
>>>>
>>>>
>>>> Thats not my final thought on the homing switch, but it is all I've got
>>>> for the day!
>>>>
>>>>
>>>> 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.
>>>> To view this discussion on the web visit
>>>> https://groups.google.com/d/msgid/openpnp/e445083b-d7c3-46d4-8c6d-67d333e26973%40googlegroups.com
>>>> <https://groups.google.com/d/msgid/openpnp/e445083b-d7c3-46d4-8c6d-67d333e26973%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>> .
>>>> 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/-XGpRTdaTrk/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/4DC74D96-DAA4-4D19-B123-2E38F76E0CB7%40gmail.com
>>>> <https://groups.google.com/d/msgid/openpnp/4DC74D96-DAA4-4D19-B123-2E38F76E0CB7%40gmail.com?utm_medium=email&utm_source=footer>
>>>> .
>>>> 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/CAJGcfUgQsZawkAOu63%2Bz6EUG%3D8YeUDLFjEamPWAKyT4b_m-UJw%40mail.gmail.com
>>> <https://groups.google.com/d/msgid/openpnp/CAJGcfUgQsZawkAOu63%2Bz6EUG%3D8YeUDLFjEamPWAKyT4b_m-UJw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>
>>
>> --
>>
>>
>>
>> JASON PARMENTER
>>
>>
>> Electrical Engineering | 208.339.0822
>>
>> ja...@tetontechnology.com
>>
>> tetontechnology.com
>>
>>
>>
>>
>>
>>
>> --
>> 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/CAP3w4qFN7wCaLMxkW-c5%2BaV9dT0z1uu266XqduYNHkw2w4m77Q%40mail.gmail.com
>> <https://groups.google.com/d/msgid/openpnp/CAP3w4qFN7wCaLMxkW-c5%2BaV9dT0z1uu266XqduYNHkw2w4m77Q%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .
>>
>> 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/-XGpRTdaTrk/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/CALsNZy35QjU-hErDC57%2BU_LLnwpFP94q1CqpKq87RgFQawwUSQ%40mail.gmail.com.

Cri S

unread,
Feb 19, 2016, 11:18:01 AM2/19/16
to ope...@googlegroups.com
you need just modify the Z code on the driver that you use to go down/up depending on Z value and Nozzle. Then it works. No need to modify anything else.

--
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/-XGpRTdaTrk/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/56C73DF2.7000906%40hauntbots.com.

Jason Parmenter

unread,
Feb 19, 2016, 11:27:49 AM2/19/16
to ope...@googlegroups.com
As long as your careful about setting the right homing order. 

Sent from my iPhone

Peter Betz

unread,
Feb 19, 2016, 11:32:02 AM2/19/16
to OpenPnP
@ Anthony: Exactly what I was thinking.

Ok, lots of great feedback, I will try and hit on all of it:

@ Michael: I understand the logic there but i am not prepared to buy into the difference in diameters between the pulleys, especially when you realize the optical difference you are seeing isn't even that much. The teeth of the belt will settle into the pulley and they will ride proud on the smooth idler. I already envisioned this difference and intend to attach the belt near the top of the carriages. Also, regarding diameter, I can't get too much smaller and move the rails together closer because one feature I am trying to keep is the ability to access the X axis carriage screws without disassembly. Also, regarding precision, I am more worried about speed than precision on spring loaded nozzles. If you or anyone else have more thoughts on all of this, please speak up and we can discuss further!

@ Maddogg: Yup, I agree with you, I will switch to mechanical. I have them on my mill and have been GREAT.

@ Cri and @ Jason: If the switch is not at the limit of one of the travels, how will it know which way to start moving to reach it? It has no idea where it is when it starts. I understand the logic here of not wanting it to slam into something if you chose to home it at the wrong spot, and you can't get it to move to a safe location forst in case it is already down.... Thoughts anyone??

@ Cri: you may have a point. Right now the rails are 100mm. 30mm of that is the carriage, so that leaves you 60mm total travel, or about 30mm on each tip. I suppose we could stick to 75mm rails, which would allow us 22.5mm travel on each tip if we run them to the end of the rails, which I assume is ok-ish. Or I could get 80mm rails made, to be safe. I don't think the price suffers from an odd cut. Based on my conversations with Sven. Thanks for pointing out your thoughts on this.

Thoughts on travel, is 22.5mm or slightly less totally fine??

Peter.

Jason Parmenter

unread,
Feb 19, 2016, 11:36:15 AM2/19/16
to ope...@googlegroups.com
Reply. Ha I was just thinking of that. You are exactly right.

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.

FredG

unread,
Feb 19, 2016, 11:40:29 AM2/19/16
to OpenPnP
What about lowering the nozzle to a kind of tool setter like on a CNC (Tormach ETS) on a defined location? 

Peter Betz

unread,
Feb 19, 2016, 11:55:15 AM2/19/16
to OpenPnP
Unfortunately having a completely different question thrown into this thread is making it rather confusing to follow. I assume you were not replying to my post? In the chance that you were, In order to safely move to a tool- setter position you would first have to know you were at safe-Z, and the only way to do that is to home.

Chicken vs. Egg.

Peter.

Cri S

unread,
Feb 19, 2016, 12:10:48 PM2/19/16
to ope...@googlegroups.com
use opto reflector , and check if carrige is over or not. Same with mechanical switch.
You need to know height of nozzle holder and nozzle.
You have solenoid that defines usable height. Assuming 20mm height of solenoid. If you said, the solenoid could be
used at max 12mm (based on force info from datasheet), the max height of Drag feeder components could be 11.5mm
Because this height is the height limit that limits the design overall.
 the board and specially tray heights are lower, in order that bigger components could be placed.
If the Nozzle holder + Nozzle have height of 35mm, nozzle is 12mm and you want places components with 50x50mm, but for moment,
limit is 25x25, assumed for to be placed on both nozzles 
Now motor is 20mm, rail = 10mm = 20mm distance to Alu from nozzle, good for max 15mm component diameter.
if you want place bigger components on both nozzles, both nozzles must be lower then alu /solenoid , always.
From Tray it must be possible to place 50x50mm component with just one nozzle.
If the design allow, it could be possible to place components bigger as 25x25 (it could be 22-25 need to check better)
Then you have limited movement that is useful.
If you go a lot up, the rail limits the usable with, reducing the 25x25 max size to 15x15mm or so, because otherwise
the rail interfere with components.
... 

Height of Drag feeder, height of pcb, height of tray feeders and nozzle height need to be considered in relation to what
max height/diameter need to be placed and height of nozzle on motor. 
And some leds needs additional nozzle soft rubber, same as melfs, ... 



 

--
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/-XGpRTdaTrk/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.

FredG

unread,
Feb 19, 2016, 12:37:58 PM2/19/16
to OpenPnP
Oh, sorry I don't want to confuse anybody! My idea was to home in X and Y, move to a desired toolsetter position and then move with the nozzle against a switch/toolsetter to home or index the Z axis...

Jason von Nieda

unread,
Feb 19, 2016, 12:46:24 PM2/19/16
to OpenPnP
If you home X and Y before Z you risk dragging the nozzle across the table, feeders, parts, etc. while homing.

In practice, I don't think this is too complex. You can setup your driver to always return to home on shutdown, or at least return Z to it's neutral position. If you want to be extra safe, put an opto at the neutral position (where both nozzles are half height) and refuse to home if it's not triggered. If the user attempts to home and the opto is not triggered pop a dialog asking them to manually move the nozzles to a safe position.

Jason


To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/57e4549f-074b-41eb-9826-d3b3be536e51%40googlegroups.com.

Peter Betz

unread,
Feb 19, 2016, 12:47:51 PM2/19/16
to OpenPnP
Yes but it is the same issue, if your nozzle is down, you might collide with something on your way over to the toolsetter right? A CNC mill for example will always home in Z before attempting X or Y homing.

Having said that, I have a VERY VERY cool idea for this issue. Going to PM Jason about it first before sharing it. Pretty excited!

Peter.

Rich Obermeyer

unread,
Feb 19, 2016, 4:40:08 PM2/19/16
to ope...@googlegroups.com
Peter, I would think that you would put that mechanical switch so it engages about half way.  Then you would be able to know which way is home.
If the switch is on then go the other way until it just turns off and you are ready to go.  Its not a limit switch or you would need 2 of them.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/606c9859-e890-4581-b2bd-62698144ccea%40googlegroups.com.

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



--
Rich

Jason von Nieda

unread,
Feb 19, 2016, 4:42:22 PM2/19/16
to ope...@googlegroups.com
Rich,

The problem with that is that you don't know which way to start traveling. If the switch is midway and the nozzle is left below the switch and you start traveling the nozzle down it will never hit the switch. Same as if it's above and you start traveling up. You need a point of reference to determine which way to start moving.

Jason


To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAOvy4mYjXYW_QLkfJB_JdEqk26w-_qxKkc-fju%3Dru3eL0caM8Q%40mail.gmail.com.

Rich Obermeyer

unread,
Feb 19, 2016, 4:47:09 PM2/19/16
to ope...@googlegroups.com
Why would you be traveling with the head down?
If the switch is ON one of the heads is lower than the other.
Start moving the head the other way until the switch turns OFF.
Now you are level and can start traveling.
If the switch is OFF start moving the head until the switch turns ON.
Now you are level and can start traveling.
Sounds obvious to me.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jyVjm1BASFV4a7oi%3Dou-9VzRe3cudFAg5knMCj3bSWVtQ%40mail.gmail.com.

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



--
Rich

Jason von Nieda

unread,
Feb 19, 2016, 4:50:41 PM2/19/16
to ope...@googlegroups.com
Rich,

By traveling I meant the head / Z, not XY. In this design there are two heads (I call these nozzles in OpenPnP, but we can stick with head). You don't know which is up and which is down. Unless the switch can be engaged by either one, there is no way to determine which one to move in which direction.

If the switch can can somehow be triggered by either one you are set but if not there is an unknown.

Jason


To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAOvy4mY2RosHUf0uqoAtEj9R0zh6YRW6J1fsRRmcWarJOsFt8A%40mail.gmail.com.

Rich Obermeyer

unread,
Feb 19, 2016, 4:58:48 PM2/19/16
to ope...@googlegroups.com
Maybe you are confused a little with what I think the switch does.
The switch is placed midway which is when the heads are level.
If the right nozzle goes below level with the left one the switch goes ON.  
No matter how far down it goes.
If you loose power or your mind, looking at the switch means the right nozzle is lower than the left.  Move the nozzle UP until the switch goes OFF and you are calibrated again.  
If you power UP and the switch is OFF then the left nozzle is probably lower than the right.  Move it up until the switch changes and again you are calibrated.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jwN0vU%3DM6YfEovfFgb2k6x6f8fxdKw8QbOr8Gm9thETJw%40mail.gmail.com.

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



--
Rich

Jason von Nieda

unread,
Feb 19, 2016, 5:03:08 PM2/19/16
to ope...@googlegroups.com
I see what you mean. If the nozzle can keep the switch depressed anytime it's below level than yes, that would work fine.

Jason


To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CAOvy4mYkQdEC9ysQmKweGyBRUt7q73_rhKcFHFSVT-UniDh7LQ%40mail.gmail.com.

Peter Betz

unread,
Feb 19, 2016, 6:20:42 PM2/19/16
to ope...@googlegroups.com
Hmmmm. I think I see what you mean. The switch would not contact the end, but the side. So you could travel past the switch in the other direction. Interesting. 

Well way to go, you might have ruined my super cool idea by making it obsolete. 

Peter Betz

Robert Walter

unread,
Feb 19, 2016, 8:50:35 PM2/19/16
to OpenPnP
What I learned from setting my machine up, is that the switch position doesn't really matter. As long as you put it to one side or the other. DO NOT put it so that it is made when both heads are home. It makes homing routines a bit more difficult to write.

What I did was mount my optical limit switch at about the 1/4 down travel position (most convenient for me) for the left head. This ensures that the switch is NOT made when the heads are in their neutral position. In the homing routine, I first disable the head stepper drive for about a half second, which causes the spring loaded heads to always return to a neutral / known position. Then the standard homing routine in most motion controllers can handle the rest. Travel in a known direction, until the switch is made. This will typically load a fixed value of user choice into the position buffer. Then command the axis to go to zero with a G command. This should put your heads at the balanced / neutral / zero position.

Essentially, what you want to prevent is the ability for the homing routine to start looking for home (ie, moving in a fixed direction), when it is already past the limit switch. Since it is past the limit switch, it wont find it and try to drive a head into the machine bottom, or worse, your board. Always starting from an absolute known position is critical. Of course, having limits on both ends of travel so you can avoid this is possible, but you can get the same effect with using the springs to push your cam back to a relatively close to neutral position, so that you are sure that heading in a fixed direction will get you to the switch.

Rob. 

Anthony Webb

unread,
Feb 19, 2016, 8:56:44 PM2/19/16
to OpenPnP
Hey Pete,

I'm very interested in your head design, do you have any images or drawings of it somewhere?  Pretty sure your design could be driven from the driver model that OpenPNP exposes.

HauntBots

unread,
Feb 19, 2016, 9:28:23 PM2/19/16
to ope...@googlegroups.com
Well, I'm sure I can find some photos, but I would like to point out that this isn't my design.  My interest is in retrofitting the control system on a 20 something year old machine made by Phillips.  So, if you still want to see it, I'll be happy to look for some photos

Here is a video of my machine running:
https://youtu.be/2k63xUrsnMQ

Only 1 of the 3 nozzles is in use on this test run.

Pete
--
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.

Anthony Webb

unread,
Feb 19, 2016, 9:34:23 PM2/19/16
to ope...@googlegroups.com
Oh wow! Thats pretty cool! Thanks for sharing.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/56C7CF46.2030301%40hauntbots.com.

Peter Betz

unread,
Feb 19, 2016, 10:40:08 PM2/19/16
to OpenPnP

Ok, Here was my idea:





This resistive pad is stuck to the main block:


Then this spring loaded wiper rides on it:


Then you use the voltage (through a divider) to determine the head position (all the time). I looked at the smoothie board and it would be easy to use one of the thermistor pins to read this sensor (it is set up for the divider already and you would see about 1.5v difference top to bottom. Then speaking to Jason it would be relatively easy to implement the homing routine in the driver using M code to request temperature from smoothie.

Data sheet says no hysteresis and infinitely repeatable depending on the wiper used. I have no doubts that this setup would home the head to within a reasonable tolerance, especially considering we are using spring loaded nozzles. $12 in parts.

Peter.






Peter Betz

unread,
Feb 19, 2016, 10:47:36 PM2/19/16
to ope...@googlegroups.com
Will model Rich's switch idea as well. 

Peter Betz
--
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/-XGpRTdaTrk/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.

Anthony Webb

unread,
Feb 19, 2016, 11:07:04 PM2/19/16
to OpenPnP
haha very clever, I like it!

Robert Walter

unread,
Feb 19, 2016, 11:41:03 PM2/19/16
to OpenPnP
Last time I looked into those pressure sensitive flexible circuit board position transducers, the lifespan is rather limited. I don't think you will get more than a week or two of placements before it wears out. If you are hell bent on absolute position, put a stepper with built in encoder on the assembly. It will be a lot more accurate, and last a lifetime.

Other than that, beautiful CAD drawings.

Robert Walter

unread,
Feb 19, 2016, 11:50:02 PM2/19/16
to OpenPnP
I double checked the lifespan on those sensors, and they rate at about 1million cycles. Considering every pick + place movement would be considered fours cycles (pick down + pick up + place down + place up), you are only looking at 250,000 operations before you can expect sensor death. That isn't much for a PNP.

On Sunday, February 14, 2016 at 12:45:56 PM UTC-8, Peter Betz wrote:
Hi Guys, 

Moved to a new thread to discuss the head I will implement for Anthony's PnP design. I don't want to buy the $300 robotdigg one.

Here are the two contenders:

1. A more compact and direct to carriage mounting robotdigg style head (tried and true):






2. A belt driven setup that will force down and force up (I will model this one to compare in the next day or so, should be really easy and use less parts.....), another variant would let gravity move it down:


Peter Betz

Peter Betz

unread,
Feb 20, 2016, 12:01:00 AM2/20/16
to ope...@googlegroups.com
Fair point, I guess .... But we are talking about an $8 part that you literally just stick on  ;) 

I am thinking about something different that might be better, and contactless (and encoder less). 

Peter Betz
--
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/-XGpRTdaTrk/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.

Rich Obermeyer

unread,
Feb 20, 2016, 12:09:14 AM2/20/16
to ope...@googlegroups.com
Although its cool, why would you need to know absolute position?
Your machine already has an opinion of where it is.  It just needs to know from where to start.
With a dirt cheap and reliable Microswitch you can tell if you are above the center point.
Anything else is just because you can.
You just need to home once when you power up and your set.
This solution is way overkill.  
$12 versus $1 plus when the sensor craps out you will see the micro switch was a way better idea :-)

--
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/fa628e6b-b6c0-44bd-9865-a02c9182b6c5%40googlegroups.com.

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



--
Rich

Peter Betz

unread,
Feb 20, 2016, 12:39:28 AM2/20/16
to ope...@googlegroups.com
The microswitch is also an idea I don't like because it would cycle 4 times with every part pick and place. I don't want to know absolute position, I want to know which arm is higher when it powers up so you know which way to go to home, so you don't collide with anything on the deck by going full travel. 

I have a contactless idea in mind. So I agree about the whole parts wearing out thing. 

Peter Betz
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/-XGpRTdaTrk/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.

Rich Obermeyer

unread,
Feb 20, 2016, 1:05:51 AM2/20/16
to ope...@googlegroups.com
I don't see why a micro switch would cycle 4 times for a pick and place.
You go down and back up. 1. Then other side goes down and up.  0
Total of 1 cycle for 2 pick and place.  You only need the switch to cycle when it goes more than half way.

Peter Betz

unread,
Feb 20, 2016, 1:28:12 AM2/20/16
to ope...@googlegroups.com
The heads don't move independently. When one goes down the other goes up. This is irrelevant. Point is, the switch will be constantly cycling. To put 2 parts on the boards , 4 clicks. Even if it was two, that's a lot. 

Like I said I think I have a solution. Will try and draw it up this weekend. I think everyone will be pleased with it. And I can finalize this design!

Peter Betz

Rich Obermeyer

unread,
Feb 20, 2016, 1:38:43 AM2/20/16
to ope...@googlegroups.com
You don't understand the discussion.  It would be 1 click for 2 parts.

Peter Betz

unread,
Feb 20, 2016, 2:33:54 AM2/20/16
to ope...@googlegroups.com
Why don't you draw what you are talking about, if you think it is actually a solution solving idea. And I understand the discussion just fine, I just don't know what you are saying ;) 

Peter Betz

Rich Obermeyer

unread,
Feb 20, 2016, 3:07:49 AM2/20/16
to ope...@googlegroups.com
That's ok, if you understand just fine, then I can't help.  You finish it.

Malte R.

unread,
Feb 20, 2016, 7:16:26 AM2/20/16
to OpenPnP
Hi,

why not use two endstops, one for each head?

Could be trigger switch, optical, reed + magnet or hall-sensor + magnet; all have their advantages and disadvantages.

But with one motor driving two heads and a potentially unknown start position, it seems to be the safest to measure both extreme positions - or am I missing something?

Regards
Malte

Robert Walter

unread,
Feb 20, 2016, 12:07:14 PM2/20/16
to OpenPnP
$8 part or not, what you want is reliability.You can be almost 100% confident in knowing that a limit switch is going to give you better repeatability than the linear contact sensor. As the sensor wears its precision will get progressively worse. Also, temperature / humidity is going to cause drift in that sensor. I had the sales rep in my office a while back, and we had long discussions about this. These were designed for low motion count / low resolution / compact applications. Personally, I would go optical photo-interrupter for simplicity / cost ($2). Will last a lifetime. It really not about the sensor cost though. A failed sensor can crash your machine, so figure the cost of fixing the machine when a life limited sensor fails.

If you want perfection (near), get a small encoder, or better yet, get a small servo drive / motor pair or closed loop stepper system. The servo / closed loop stepper will be far easier to implement, as you can feed it with step / direction signals, so it would appear no different than a normal stepper. The matching drive will have the input for the stepper. However, you will still need to home accurately with some form of switch, unless of course, your servo / stepper encoder has a Z / Index pulse, then you could theoretically use that by disabling the drive, allowing springs to center the heads, and then search for the first occurrence of the index pulse, and call that the reference point relative to home. But in the end positional accuracy will be excellent.

Peter Betz

unread,
Feb 20, 2016, 1:53:59 PM2/20/16
to ope...@googlegroups.com
For sure. Trust me I am listening to everyone's suggestions and I think I have a good all round solution. Just about done modeling.  

Peter Betz

Anthony Webb

unread,
Feb 20, 2016, 2:47:48 PM2/20/16
to OpenPnP
Lets not lose sight of the fact that the spring return version is cheap, accurate, requires less electronics, and pretty compact and thin.  It will probably be hard to beat as there are very few tradeoffs.

-Anthony

Peter Betz

unread,
Feb 20, 2016, 3:46:31 PM2/20/16
to OpenPnP
Ok!

Objective: Create non-spring return head, thus requiring some way to level heads.

Issue: Limit switch(es) placed at the Z rail extents creates a collision hazard when homing, since it is not known whats under the nozzles (Thanks Cri!). A microswitch simply placed at center does not allow for enough information for the driver to decide which way to travel to 0. It is also not possible to home X and Y first as you risk Z collision.

Solution: Create a simple switch system that allows the driver to know if the "homing head" (in this case the right hand head) is above or below 0. This is accomplished by ensuring a switch is continuously tripped while the head is on one side of centre (Thanks Rich!). This way when the driver initializes it will ask "Is the switch on, or off?", based on the result it will move towards centre. The driver will allow everyone to set their number of steps from the switch transition point to zero. 

Issue: My [super cool] resistor wiper setup and other peoples suggested microswitch setups are subject to wear and drift.

Solution: Optical, contactless sensor, and interrupter milled or printed (?) into the head.

Thanks to everyone for their input, here is where I am at:






Here is the switch:


Are the smoothie end stops 5v or 3.3v? 

Let me know what you guys think!

Peter.






FredG

unread,
Feb 20, 2016, 6:07:07 PM2/20/16
to OpenPnP
Hello Peter,
in my opinion this is the best solution! When you reach the index position you know (after configuration) how far you can go in both directions. In this way it is defined to lift the lower head for homing.
Maybe its easier to screw a stripe of metal or plastic on the side. It will also reach deeper into the optical sensor...

FredG

unread,
Feb 20, 2016, 6:20:52 PM2/20/16
to OpenPnP
...but it may be dangerous in other configurations as you are allowed to move beyond the "endstop"...

FredG

unread,
Feb 20, 2016, 6:56:00 PM2/20/16
to OpenPnP
To avoid crashes in other confogurations it may be an idea to define a "home to index" function and detect rising or falling edge instead of high or low.

Anthony Webb

unread,
Feb 20, 2016, 8:28:15 PM2/20/16
to OpenPnP
Definitely the best solution of the bunch.  I wonder if it couldnt be made more compact if you could get rid of that wing for the optical sensor and instead place a sensor above or below the one of the rails that attached from the back and could "see" if there was an obstruction optically (or magnetically?) Essentially the same idea as here, but without the excess width?  What is the current overall width?

Peter Betz

unread,
Feb 20, 2016, 9:04:13 PM2/20/16
to OpenPnP
I am not sure what you mean exactly, It can't really be under the carriage because it will interfere with the rail? It has to be triggered for around 50% of the total travel amount or it doesn't work. And now that I think about it, we will need to make sure the firmware doesn't treat it as an actual end stop every time it triggers. 

Are the smoothie endstop plugs 5v?

Here are the dims:



This feature adds around 20mm to the width. And if there isn't anything on top of the Y rail, it shouldn't even limit travel compared to normal?

Peter.

Jason Parmenter

unread,
Feb 20, 2016, 11:22:33 PM2/20/16
to ope...@googlegroups.com
Peter,

I know the MKS SBASE board that we uses 3.3V logic at the end-stops.  The MKS SBASE board is a knock off board of the Smoothie. I think I read somewhere that there is 5V on the actual Smoothie board or at least a jumper would set it to 5V?  Most of the optical end-stops are 3.3V to 5V tolerant.

​

Michael Anton

unread,
Feb 21, 2016, 2:02:01 AM2/21/16
to OpenPnP
At the very least, you can eliminate the wing to the right of the flag that trips the sensor, which would save some material, and weight.

How about putting a flag on the piece that fastens to the belt, and triggering along that line?  The sensor could mount the main body, though one ear might need to tuck under a rail in a pocket, but that would still be better than adding material, and requiring the sensor to move.

I'm also not sure why the idler pulley at the bottom needs to be so low, when you are planning to contact the belt higher up.  It looks like it could be well inside the main body.

Mike

Peter Betz

unread,
Feb 21, 2016, 4:06:11 AM2/21/16
to ope...@googlegroups.com
Hi Mike. 

You can't have a flag on the belt and have it do what this is designed to do. This design allows for a somewhat safe homing routine by always ensuring the nozzles never go down further than where they are when the machine powers on. If we were doing a simple extend homing switch then we wouldn't have to extend the body. 

About the extra part of the flag, I guess you are right about the weight, but you aren't saving any distance, you still have to clear the sensor! 

I will have another look at the pulley, not sure how tight I want to make it, it's easy to zoom in with CAD and see something fitting. It's entirely different fitting your sausage fingers in there to assemble it. Don't forget, the nozzle is 50mm below the stepper. 

Peter Betz
--
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/-XGpRTdaTrk/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.

Robert Walter

unread,
Feb 21, 2016, 12:35:36 PM2/21/16
to OpenPnP
The MKS Base has a jumper to supply either 3.3 or 5.0 V to the end switches. The inputs are 5V tolerant.

If you go limit switch, no problems at, just connect between the input and ground. The input has a built in pullup.

If you go photo-interrupter, then make sure you have a current limit resistor in place on the emitter side to properly drive the LED. This resistor would be in series between the source voltage and the LED.

Rob.

Rich Obermeyer

unread,
Feb 21, 2016, 2:01:19 PM2/21/16
to ope...@googlegroups.com
Peter, how about a side or top view to go with this front view to give a little perspective of how this goes together.
--
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.

Peter Betz

unread,
Feb 21, 2016, 3:43:07 PM2/21/16
to OpenPnP
Here are a couple more shots:




Is that what you mean?

Peter.

Peter Betz

unread,
Feb 21, 2016, 3:47:22 PM2/21/16
to OpenPnP
Trying one more thing out. Then I need to make a decision and start ordering parts.

Compact spur gear drive:


What do people think about this idea? I am thinking any kind of accuracy issues would be pretty small and easily absorbed by the nozzles. This has significantly fewer parts, smaller overall, and easier to assemble. 

Peter.

Anthony Webb

unread,
Feb 21, 2016, 3:55:18 PM2/21/16
to ope...@googlegroups.com
Smaller less parts and easier to build. Sounds like you answered your own question. I would prefer it given you support the racks a bit more and can figure out the leveling.  

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.

Peter Betz

unread,
Feb 21, 2016, 10:58:35 PM2/21/16
to OpenPnP
Yes for sure, I am not done the model, wanted to throw it out there for critique in case there was something that made it impractical. 

For leveling I would use the same as the last one with the optical sensor. I will try and use the same stop you are using on the rest of the machine like we were discussing.

Seems like this will be the one I build. I will also order the parts for the pulley head in case there are issues and I have to switch.

Getting closer every day!

Peter.

Michael Anton

unread,
Feb 22, 2016, 12:32:12 AM2/22/16
to OpenPnP
You will have backlash in this case, but I'm not sure if that would matter much..

Mike

Michael Anton

unread,
Feb 22, 2016, 12:32:19 AM2/22/16
to OpenPnP
I wasn't really thinking of a flag on the belt, but rather one where you fasten the belt to the carriage.  I don't understand why it can't be configured to work exactly the same way.  Make the flag long like you propose now, but it is moving now instead of static.  I must really be missing something, if this is not possible.

Mike

Peter Betz

unread,
Feb 22, 2016, 11:41:34 AM2/22/16
to OpenPnP
Hi Mike,

 Yes on the backlash, I am sure there will be, but like you said, I think it won't matter.

I think the only thing you are missing is room to have the sensor near the rail and carriage. Maybe you mean to turn the sensor on its side.... interesting. I am sure things could be tightened up a bit, I am going to try and only make the new head grow in height for homing. I am also trying to keep the number of machining operations to a minimum, for example, the wing sensor idea requires only milling and drilling in the Z plane, no holes to drill and tap on the sides, etc.

Stand by for more!

Peter.

Peter Betz

unread,
Feb 23, 2016, 1:21:35 AM2/23/16
to OpenPnP


Completely re modelling to make it clean.... Pretty much back to where you all saw it last:


Peter.

Anthony Webb

unread,
Feb 23, 2016, 5:47:46 PM2/23/16
to OpenPnP
Looking very clean Peter!  You could almost have that NEMA plate cover the teeth/gear and yeild an amazingly thin design.  Making thing compact is only a goal for me because I know people will want to build machines that are budget or space constrained and even an inch or two here and there can really make a difference in the overall usefulness of the machine..

Peter Betz

unread,
Feb 24, 2016, 12:29:07 AM2/24/16
to OpenPnP
Peer review time!!

Introducing...... the ICE PICK.. (Name TBD)






Will be refining, and some dimensions need tweaking, for example, if We are going to have the opto sensor interrupter on the back side of the Z carriage, I will need to move the NEMA 8's out a bit to make room for the screws, don't want to have to machine on both sides (pockets for the screw head). 


For a full view, click here. 


Let me know what you guys think. Ordering all the rails tonight!


Oh, Anthony, will I be able to attach the belt to the bottom of this thing? Or maybe I should just mount this to one of your carriage blocks.....


Peter.






Peter Betz

unread,
Feb 24, 2016, 12:33:28 AM2/24/16
to OpenPnP
Oh, and I made this version to incorporate the cheap opto endstops that Anthony spec's for the rest of the machine. Now they are all the same.

Peter.

Jason Parmenter

unread,
Feb 24, 2016, 12:47:37 AM2/24/16
to ope...@googlegroups.com
Nice job Peter, go build that thing.  I want to see it in action!

On Tue, Feb 23, 2016 at 10:33 PM, Peter Betz <betzt...@gmail.com> wrote:
Oh, and I made this version to incorporate the cheap opto endstops that Anthony spec's for the rest of the machine. Now they are all the same.

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.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/eced2693-4ff1-4202-b764-b017c9f02d8c%40googlegroups.com.

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



--



JASON PARMENTER 

Electrical Engineering | 208.339.0822
ja...@tetontechnology.com
tetontechnology.com 





Anthony Webb

unread,
Feb 24, 2016, 12:52:18 AM2/24/16
to ope...@googlegroups.com
Attaching the belt should not be an issue assuming you are using 1/4" stock for the backplate, will just need to tap a couple holes in the side wall.  Really superb job Peter, you obviously know your craft.  I cant wait to make one myself!

Peter Betz

unread,
Feb 24, 2016, 12:53:08 AM2/24/16
to OpenPnP
Thanks Jason!

Finally started ordering parts for the whole machine today. Extrusions from Misumi and HIWIN class C rails! Pretty pumped. Will have to see if I have end mills small enough to do those racks, they have tiny features, like 1/64" tiny...

Anthony Webb

unread,
Feb 24, 2016, 12:59:10 AM2/24/16
to ope...@googlegroups.com
I am thinking that rack/nema mount would lend itself well to being 3D printed?  If you had some side walls around the sides of the nema it would probably make a pretty bombproof print/mount.

--
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/28d87a00-64b9-474c-9a81-6342573515e7%40googlegroups.com.

Peter Betz

unread,
Feb 24, 2016, 1:02:40 AM2/24/16
to ope...@googlegroups.com
Well, it is certainly possible I suppose. I actually ordered a linear rail for my mendel max printer with these ones. I have been having trouble getting really nice resolution, the Y axis (bed) is really really sloppy. Do you have a 3D printer ? Anyone else want an STL to try printing the part with the rack? 

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/-XGpRTdaTrk/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.

Anthony Webb

unread,
Feb 24, 2016, 1:22:28 AM2/24/16
to ope...@googlegroups.com
I have a 3D printer, if you give me an STL I will try to print it.

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/EADE29DB-DB05-48FF-949F-E1113982876D%40gmail.com.

Peter Betz

unread,
Feb 24, 2016, 2:03:54 AM2/24/16
to OpenPnP
Here you go! Excited to see it, because I think I just realized I can't mill those racks......

Peter.
Z carriage STL.stl

Tim Kent

unread,
Feb 24, 2016, 2:09:49 AM2/24/16
to OpenPnP
That looks awesome, good job Peter!

Michael Anton

unread,
Feb 24, 2016, 4:00:20 AM2/24/16
to OpenPnP
Looks great!!!  I'll be interested to see if 3D printing the racks would work out ok.  Heck, if you are going to print them, then you could use a herringbone rack and gear.  Then there would be no backlash.

Would it make sense to counterbore the linear block mounting screws, so the stepper could be moved closer to the X axis?

Mike


On Tuesday, February 23, 2016 at 10:29:07 PM UTC-7, Peter Betz wrote:
Peer review time!!

Introducing...... the ICE PICK.. (Name TBD)










Joao Matos

unread,
Feb 24, 2016, 10:20:13 AM2/24/16
to OpenPnP
This is looking like a great design.

If someone else is also thinking of building this, I've found a supplier for the rails: http://www.aliexpress.com/item/HIWIN-MR9-9-mm-linear-guide-rail-stainless-steel-rail-2pcs-MGN9-L-100-mm-with/32442735113.html

Peter, what is the exact length dimension for each rail?

Thanks

Anthony Webb

unread,
Feb 24, 2016, 10:30:19 AM2/24/16
to ope...@googlegroups.com
@joao, peter is using 75mm rails.  I know he has a guy on ebay that is putting together a listing for these as well.  I think they do sell this size on robotdigg as well.

--
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/fdda6b24-7766-4844-9156-c8a3f959baa5%40googlegroups.com.
It is loading more messages.
0 new messages