neoden machine platform

1,107 views
Skip to first unread message

Hedley Davidson

unread,
Jun 27, 2013, 12:59:55 PM6/27/13
to ope...@googlegroups.com
I have capitulated and ordered the Neoden machine expected next week . I got the smaller unit ( TM220a - US$ 5000 with shipping ) , as the only difference is the number of feeders and PCB working area.
Our intention is to drive this machine with openpnp software and we are busy with a flying vision camera system for the head based on the design as seen in the videos of the versatronics 
unit with a mirror that moves out of the way before placement . 

I will keep group posted once there is demonstrable progress. 


Regards

Karl Lew

unread,
Jun 28, 2013, 11:23:46 AM6/28/13
to ope...@googlegroups.com
Please post your experience with openpnp and new machine!

Jason von Nieda

unread,
Jun 28, 2013, 3:25:12 PM6/28/13
to ope...@googlegroups.com
There's someone else working on an OpenPnP conversion of this machine but for the life of me I can't find the email thread. Perhaps if that person is listening, they could speak up? I know it's been a few months since we last spoke, so there might be some progress.

Jason



On Fri, Jun 28, 2013 at 8:23 AM, Karl Lew <ka...@firepick.org> wrote:
Please post your experience with openpnp and new machine!

--
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/983574b1-65b7-42ca-ad78-3eb041d5b049%40googlegroups.com.

Michael Grant

unread,
Jun 28, 2013, 3:27:04 PM6/28/13
to ope...@googlegroups.com
IIRC, that might have been mentioned in the dangerous prototypes forum...

Michael

Jason von Nieda

unread,
Jun 28, 2013, 3:27:51 PM6/28/13
to ope...@googlegroups.com
It was definitely an email thread between me and someone else, but it may have been direct instead of on this list. I'll keep digging.


Jason von Nieda

unread,
Jun 28, 2013, 3:52:53 PM6/28/13
to ope...@googlegroups.com

Oleg Mazurov

unread,
Jun 28, 2013, 7:30:02 PM6/28/13
to ope...@googlegroups.com
I haven't had much progress, unfortunately. Still developing a plan for conversion. I use the machine quite often so I need the ability to switch quickly from OpenPNP to the original control until the conversion is done - this adds extra work.

Ami

unread,
Jun 29, 2013, 7:56:22 AM6/29/13
to ope...@googlegroups.com
In a way it's interesting to know that two owners of TM220 decided to run openpnp.
On the other hand, it's alarming to know that two people decided to run from a working machine rather than building ones.

Hedley Davidson

unread,
Jun 29, 2013, 9:05:44 PM6/29/13
to ope...@googlegroups.com
At the price point I don't have high expectations re accuracy and am exposed re spares etc . On working out engineering costs per hour and sourcing all the components the buy vs make made sense . . No machine vision is a real negative and possibly can not be easily integrated. If not , we will use tm220 as is, for small production and prototyping and persue openpnp at a more leisurely pace for hobby interest . We still working on scara based design as we have a small epson scara robot at the office doing mechanical assembly of inserts into the plastic housings and it is fast and accurate so we have a reference target. As with all engineering endevours the end user is paying a portion of recovery of the non recurring engineering costs , factory spares holding and machine material costs . Our teardowns and analysis of commercially available units show use of high cost , high precision machine parts to get the speed , accuracy and reliability required for volume production . For small runs ie 100 boards a month the tm220 may be ok ? The processing electronics provides increasing bang for the buck almost monthly but the real difficulty and cost is the mechanics and software.

Hedley Davidson

unread,
Jul 2, 2013, 12:36:11 PM7/2/13
to ope...@googlegroups.com
Neoden are busy with a vision based offering which will probably be released in a few months . Wish I had waited but perhaps will get experience on 220a first .

Hedley Davidson

unread,
Jul 7, 2013, 1:10:08 PM7/7/13
to ope...@googlegroups.com
link to video of single step on my new toy -:

http://www.youtube.com/watch?v=CQGWTi4F40Y

there is another one at full speed on youtube 



On Saturday, June 29, 2013 1:56:22 PM UTC+2, Ami wrote:

Jason von Nieda

unread,
Jul 7, 2013, 1:23:40 PM7/7/13
to ope...@googlegroups.com
Thanks for posting these Hedley. Very cool to see the machine running. It's certainly fast! Does it use just a solenoid with a pin for the tape advance, or is it something more complex?




--
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.

Hedley Davidson

unread,
Jul 7, 2013, 1:54:32 PM7/7/13
to ope...@googlegroups.com
Small solonoid with pin. The whole design is very minimilist but works well . The laser targeting is accurate . On posted video the distance from pickup to placement point is x 100 mm and y 110 mm . At this distance the placement is 4500 placements per hour .

The air manifold has pump in for blowoff and vacum out . Four small solonoid s then select suck or blow for each of the two heads .

You set feed displacement for each part in the job which is a csv file .

Hedley Davidson

unread,
Jul 7, 2013, 2:03:10 PM7/7/13
to ope...@googlegroups.com
The nozzles have an integral spring to allow 1.5 mm of z movement . Each component has a z height setting but the integral spring in nozzle provides wide tolerance . I am very impressed with build quality and only change would be vision system for ic placement . The feeding and take up spools also work very well . All mc,s have design parameters so if you placing > 100 pin fine pitch ic,s then better resolution required .

Ami Sukarsa

unread,
Jul 8, 2013, 11:25:57 PM7/8/13
to ope...@googlegroups.com
My brain says:  way-t-go Neoden, I should go and get one! My job is to build boards not machines. And the TM220 is certainly giving a solid starting up point.
My heart says: Is this the end of openpnp hardware?
 After a while with openpnp I'd legitimately want to see a real openpnp machine coming out. My old Sylvestre is still working, but it's not reproducible, and slow. So c'mon, c'mon guys, nothing fancy please, just something that works! I still have hope for it.....


On Sun, Jul 7, 2013 at 8:03 PM, Hedley Davidson <hedley....@gmail.com> wrote:
The nozzles have an integral spring to allow 1.5 mm of z movement . Each component has a z height setting but the integral spring in nozzle provides wide tolerance . I am very impressed with build quality and only change would be vision system for ic placement . The feeding and take up spools also work very well . All mc,s have design parameters so if you placing > 100 pin fine pitch ic,s then better resolution required .

--
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/GWtDo0QOW2Q/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,
Jul 8, 2013, 11:51:52 PM7/8/13
to ope...@googlegroups.com
If OpenPnP dies you will hear it from me first, and I have no intention of quitting unless someone else does what I am trying to do first.

~$1000, open source pick and place. As soon as it exists I'll consider my mission complete :)

Many people wonder why it's taking so long. Simple answer: I'm one guy working on this in my spare time. My day job has nothing to do with electronics or pick and place. Every hour of work and every dollar of R&D put into OpenPnP has been in my spare time and out of my pocket. I have to balance that against my other interests. So much to do in each short 24 hours :)

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.

Hedley Davidson

unread,
Jul 10, 2013, 12:42:37 AM7/10/13
to ope...@googlegroups.com
Whilst a good platform it is closed so any improvement no matter how small needs factory to implement . For example I wanted jog and manual place for fine pitch - no go . I believe in the end openpnp will be an awesome front end pnp system with vision . The driver electronics , motor control , feeder and head design are not waiting on Jason or any specific developer in addition there seem to be many observers , of which I am one , who are waiting for Jason and others to do all the development on the software rather than being active participants .
Message has been deleted
Message has been deleted
Message has been deleted

David Armstrong

unread,
Jul 10, 2013, 6:24:48 AM7/10/13
to ope...@googlegroups.com
i'm wondering how is the pickup head rotated , is it a geared motor with slider etc 
i'm looking for ideas or pictures etc

Dave

Jason von Nieda

unread,
Jul 10, 2013, 12:05:36 PM7/10/13
to ope...@googlegroups.com
David,

On the Neoden machine, the nozzle is directly coupled to a small stepper (looks like a NEMA 8) for rotation and the whole assembly slides up and down for Z.

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.

Hedley Davidson

unread,
Jul 10, 2013, 3:41:36 PM7/10/13
to ope...@googlegroups.com
Here's the link of normal speed . The components on the right I scraped off for next trail so they not from machine placements

http://m.youtube.com/watch?v=qYqeUAix-34

chris

unread,
Aug 11, 2013, 9:24:54 PM8/11/13
to ope...@googlegroups.com
Hi,
 how do you proceed and was is planned? 
I have made such a conversion. It´s a partial implementation, the orginal controller places the components from the feeder and then components from the tray
are placed  using vision. If the chips comes in tape, then there is a job first that places the components to trays using the orginal controller.

Chris

Oleg Mazurov

unread,
Aug 12, 2013, 5:45:38 PM8/12/13
to ope...@googlegroups.com
Which controller are you using for Openpnp? And how do you switch between original controller and Openpnp? Can you post pictures of the camera mount?

chris

unread,
Aug 12, 2013, 9:23:36 PM8/12/13
to ope...@googlegroups.com
From the main board, there is a flat cable that goes to the driver board with step and dir signals. I have made a board with 7486 xor chips to interconnect between
these two boards in order to allow external overrides. All dir/step segnals have the overrides. In addition, the step signals are buffered having enable pins and pull-down
in order to add externally emergency button for implementing controlled shutdown.
 Additional there is a 4x changeover relais because i had it in the box that switches controller  for the solenoid signals used for vacuum and blow. Further i have added magnet to the head for checking if it´s at the park position or not using hall sensors.
As controller, i have used one usb enabled pic with 4 axis interface. Only 4 axes for x/y/z/c because head 1 is reserved for the XS Nozzle and so it don´t make sense
to use it for tray operation. Eventually there is no problem adding more axes if needed, just a recompilation of code.
The camera mount, i don´t have it. I have only uplooking camera, no downlooking. The usb camera pcb is attached to the base using some small wood pieces and hot glue.
For the setup of feeder i use barcode reader and positioning the head to indicate the lane to change.

chris

unread,
Aug 12, 2013, 11:09:11 PM8/12/13
to ope...@googlegroups.com
I cannot add images, there is a problem with the browser/antivirus whatever so i post a link.
The pinout is this, viewed from the underside, pin1 is +5V:
http://s24.postimg.org/mzz7fuq6t/pp_step.jpg
And the interface used for driving it is this:
http://s11.postimg.org/3rf3er74z/pp_usb.jpg

A-C is the head, F is is the feeder.

Hedley Davidson

unread,
Aug 13, 2013, 12:58:39 AM8/13/13
to ope...@googlegroups.com
As I understand your post you have set up a separate controller which can drive the machine via usb . Does this mean that you don't use their front controller board at all or their software but have just tapped into the driver boards for head manifold and motor movement . If not how do you synchronize movement eg the machine does a pickup and then you insert movement to go to up camera , do offset adjustment and then pass control back to machine ?

chris

unread,
Aug 13, 2013, 4:59:53 AM8/13/13
to ope...@googlegroups.com
Am Dienstag, 13. August 2013 06:58:39 UTC+2 schrieb Hedley Davidson:
As I understand your post you have set up a separate controller which can drive the machine via usb .
Yes, i use a PC, migrating it to a embedded board that support OPENCV is possible without problems and driving it directly.
 
Does this mean that you don't use their front controller board at all or their software 
No, their controller is used to populate things and to fill the tray from tape.
 
but have just tapped into the driver boards for head manifold and motor movement .
Additionally the solenoid for vacuum and blow is tapped into.
 
If not how do you synchronize movement eg the machine does a pickup and then you insert movement to go to up camera ,
I have inserted a magnet and sensor in order to sense the head if it stay into the park position, on the right back side.
The machine does its job, and then it park the head into the "park position" emitting a beep. I do not sense the beep, only the park position.
If it stay more then 200ms on the "park position" the pc takes control. That´s all really simple and kiss.
The machine is not allowed to travel to the park position during normal pick and place movement for the orginal controller (sd-card).
This limit the usable pcb size a bit, but it´s not a issue for me.
I have added emergency stop button that stops the steps signals and cut power to the overall machine, just in case.
 
do offset adjustment and then pass control back to machine ?
I don´t pass control back to the machine, for this a sort of dwell command must be present in there controller, i don´t know if such thing is present or not.
First the machine place the components, and then the pc place the components. Actually it´s the operator responsability to syncronize the things.
Maybe a Transcend Wi-Fi SD Card could resolve this but i doubt.
I´m interested of knowing a method of syncronizing the thing, eventually driving it to pick from a free feeder using vacuum control and at the same time doing the
nozzle change automatically of nozzle 2. This require a change of the override interface, so it´s a low priority for now. The override interface was made using breadboard pcb style and no docs.

chris

unread,
Aug 15, 2013, 10:10:28 AM8/15/13
to ope...@googlegroups.com
Hi Oleg,


Am Montag, 12. August 2013 23:45:38 UTC+2 schrieb Oleg Mazurov:
Which controller are you using for Openpnp?
It´s pic18f2550 board, if i need to replicate it, i would use the STARTUSB FOR PIC from Mikroelektronika, 14.316€ from TME.
 
Can you post pictures of the camera mount?
You intent for the downlooking camera. I don´t have mounted it, but have a solution for that if you need it.
To explain why i don´t use it.  I have mounted one cheap 16mPixel Camera above the machine with some additional illumination.
I use it to read in the pcb offset/orientation and to make a sort of AOI. Everything equal or above 12mpixel work.

For the solution, the camera has to mount on the laser using a sort of clamp above the mounting block.
Further the clamp need to stay a bit over the square laser holder in order to remain aligned because of the initial calibration.
Annother thing to remember, you need at least 35 degree optic or less and you need shielded usb cable.
If interested i can explain it further.

Michael LaFleur

unread,
Aug 15, 2013, 6:11:57 PM8/15/13
to ope...@googlegroups.com
I recently purchased a TM-220A. I have a problem where my 0402 parts are flying out of the tape because it is being advanced too fast. Is there a way to slow down the advance rate?

Mike

Michael LaFleur

unread,
Aug 16, 2013, 11:28:31 PM8/16/13
to ope...@googlegroups.com
I noticed that the unit is designed to pick up the third component showing after advancing the tape. This third one often gets displaced, les second much less, but the first one, the one right next to keeper never gets dislodged.

I fixed my issue by adjusting the pickup location in the script for all stacks by -4 mm. This makes the unit pick up the 1st exposed component instead.

bie...@gmail.com

unread,
Jan 11, 2015, 5:03:00 PM1/11/15
to ope...@googlegroups.com
Hi Chris,

Do you still have the TM240 working with openpnp ?

I would love to do the same with my TM240, but I have some questions about your setup...

Remco

m34...@gmail.com

unread,
Jan 12, 2015, 3:48:50 AM1/12/15
to ope...@googlegroups.com
Yes, sort of, custom sw.

bie...@gmail.com

unread,
Jan 12, 2015, 1:21:54 PM1/12/15
to ope...@googlegroups.com
I there a way I can make use of your SW ? I would like to try it out.. :)



On Monday, January 12, 2015 at 9:48:50 AM UTC+1, m34...@gmail.com wrote:
Yes, sort of, custom sw.

John W

unread,
Feb 18, 2015, 9:22:55 AM2/18/15
to ope...@googlegroups.com
If you are interested in the Neoden machines, check out http://www.neoden.eu.

mojalovaa1

unread,
Feb 10, 2016, 10:34:42 PM2/10/16
to OpenPnP, johnw...@gmail.com
Now I m see this post , and have  some question about that TM220A pnp machine.

Some one know how much is max feed rate on that machine , acceleration  , rotation speed ?
I m today test my machine , you know that Jason  , and have  on feed rate max speed for x axis 47500 mm/min and for y is37500 mm/min .
All so acceleration is for x 4750 mm/sec2 or how go that  , and on test work that is so fast  , and work with 2 head 1400-1600 cph .
How much than is speed on TM220 or 240  , when all so use stepper for motion , I have better driver , and I have 48 V driver supply ?
With that story  that machine have min. 4x that speed and accel.?

Zachery Wheat

unread,
Sep 21, 2017, 3:43:53 PM9/21/17
to OpenPnP
Does anyone have any experience using the new NeoDen 4? I recieved mine yesterday and have been getting it setup, but when i go to start it, it just boots up through the logo screen, then doesnt do anything. I lose my VGA signal and then the screen just goes into power save mode.
Reply all
Reply to author
Forward
0 new messages