Zenbot 1216 CNC

From LVL1
Jump to: navigation, search

Back to list of Equipment


Lvl1 qr header.png
Zenbot_1216_CNC&.png

Contents

General

  • Owner/Loaner: LVL1 (donation from chorgy)
  • Serial Number: none found on device so far
  • Make/Model: 1216
  • Arrival Date: ???
  • Usability: somewhat - in process of setup
  • Contact: Mike Lutz, Chris Pettus or Gerrit
  • Where: In the basement next to the big CNC system

Current status

As of Nov 29 2012
System usable but has no safeties, it is still trivial to cause the machine to destroy itself.

Operating instructions

The logical tasks/steps needed to go from an idea to a file that can drive the zenbot

As-of-Dec-2012 the equipment can be operated with extreme caution.

Given the lack of
  • limit switches
  • emergency power off
  • cleaned up software
There is a high risk that the machine will be damaged if used. If you really want to work with the machine now its STRONGLY recommended that you have Chris Pettus or Mike Lutz sit with you the first time you try and use the machine.


More instruction will be coming in the future, but the best thing you can do now is pick and learn a 3d CAD package. They are shockingly hard to use and you should expect spend 8+ hours going through the tutorials in order to drive the system such that all the faces have the surfaces you want and also learning how to use the snap systems to align things in 3d space (all of which will be critical when the file is handed over to the CAM software)

If you happen to be working on OS X, the popular commercial hobbyist-targeted 3d CAD Rhinoceros is freely available while it continues to be in development (see: http://mac.rhino3d.com). I think the primary appeal of Rhino is that it has a very good tutorial (... well, and its free for OS X ... for the moment)

Materials tested with zenbot

  • Machinable wax
    • Cuts easily - though not cleanly. Still learning
  • RenShape 460
    • Cuts easily and cleanly. Only had good experiences with it so far
  • Green-blue rigid foam (unsure what it is, was found in scrap box)
    • Cuts cleanly and easily, but kicks up static-attracting bits of foam, likely to fowl the mill if not cleaned while cutting
  • Some kind of fiber board (found as scrap)
    • Sloppy cuts, details tend to break off
  • Some kind of clear plexiglass (found as scrap)
    • BAD: within a few minutes of cutting, the plastic formed on the blade, completely isolating the cutting surfaces


Equipment Description

Parts that make up this system:

Wiring diagram for the zenbot
  • Zenbot itself (Includes the body, armatures, and the x, y, and z motors and actuation hardware (belts))
  • X, Y, Z motor controller
  • Power brick for X, Y, Z controller
    • Labeled "PA3290U-2ACA SEB 100P2-15.0". Input 100-240v 2.5A 50/60Hz. Output 19V at 6.3A. Center of connector is plus, outer shield is minus.
  • Spindle (the routing part) - Believed to be a Wolfgang Engineering ... um ... spindle? (they don't seem to have any model info on the spindle or historical information on the website)
    • Wolfgang engineering website
    • The belts needed to couple the motor to the spindle were missing but have been replaced by the smaller belts from faucet repair kit.
    • Gerrit has talked with Wolfgang about this and he may have enough info to order the belts as parts (They may effectively be big O-rings)
      • The info Gerrit received (with regards to the o-rings) didn't match our spindle. Chris P found a match by trial and error
    • The smaller o-rings from a DL-15 Single Handle Delta Kitchen Spout Repair Kit seem to be a "good enough" match to the belts that are needed for the spindle. As of Nov 6, 2012, these are the belts on the spindle.
  • Spindle motor controller
    • Cytron Technologies MD30B
  • Spindle power supply
    • An HP/Compact server powersupply "ESP115" from the boneyard.
    • Documentation for the MD30B controller says that it wants between 10 and 30Vdc with with a peak (output) of 30A (at 12Vdc?). Current product from wolfgangengineering has 200W 12Vdc motor, so 30A peek seems possible.
  • Computer running EMC2 CNC control software and that has a parallel port
    • Thin dell desktop sitting with the zenbot.
  • Bits
    • There is one bit with the Zenbot, but to be of use more will have to be acquired. There seems to be agreement that the best way is to buy of Ebay, more digging will be needed to understand what to buy and how to re-cert/re-measure it in order to make it usable with the machine.

Pictures

Picture of the heap of parts and pieces that make up the Zenbot
Front of the X,Y,Z controller
Back of the X,Y,Z controller
The power supply that has a plug matching the X,Y,Z controller
The spindle motor controller board
Spindle as it was mounted when Zenbot was first worked on by Mike Lutz
Spindle assembly with motor model visable
Powersupply used for spindle motor
Faucet repair kit. The smaller o-rings (not shown) fit the spindle.
Size of the bits that will fit in the collet that we have
measurement between the holes for the motor mount on the spindle (should be common with DC remote control car motors)

Notes from initial equipment build

Log of research (newest to oldest)


The way the collet must be held when unscrewing it switch out the bit (if not held the bit will slam into the base of the CNC and may break)
Picture of where a hex wrench that matches the collet/chuck is tapped to the zenbot
If the collet is complete unscrewed when changing bits (I suspect this is the safest way to do it
When the collet is loose the bit just slides out of it
A picture of the collet and the bit seperately

Nov 12 Mike L

- Copy of update email -

Virtually all progress in this update (and most from last) is thanks to the continued heroics from Chris Pettus. He is the unsung hero of this project and will be the reason the machine gets to whatever level of operation that it gets to.

Detail for this week:

Progress made:
  • Chris P found some < $4 faucet o-rings that are the right size to link the spindle motor to the spindle arm
    • They are installed on the zenbot, have about 30 min of time on them, haven't shown any issue with wear or slippage.
  • Spindle power supply interface case built with integrated fuses
    • currently fussing the spindle at 10 amps, though the power supply really should be able to drive 30 amps safely, we are using the low fuse to better learn what the spindle will really need -- In other words, if anyone tries to cut something really hard (or jams the cutter) the spindle fuse will likely blow.
    • I expect we will increase the fuse as we better understand the spindle (unless we find it never blows)
  • Last meeting, with the help of many other Lvl1'er, the Zenbot was moved down stairs next to the big CNC.
  • Running the spindle at full speed (likely too fast for the wax) Chris P and Mike L did a cut of the first few letters of "LinuxCNC" in machining wax. See the wiki page for full desc or jump here for picture: http://wiki.lvl1.org/File:Zenbot_Clockwise_cut_(first_cut).jpg
  • The spindle motor speed controller has been repaired and tested with a separate motor, not yet integrated back into the zenbot.
Next actions:
  • Get the speed controller installed between the spindle and its supply (also, perhaps trying to connect the PWM interface to LinuxCNC) - Chris P
  • Measuring spindle speed (at full and different settings of the speed controller - Chris P and Mike L
  • Writing up the software complexities of working with the CNC (Need CAD & CAM before can use LinuxCNC) - Mike L
  • Learning various CAD and CAM packages so-as to create g-code that we can mill with the Zenbot - Mike L, Chris P, and anyone else who is thinking they want to use this system
Future/dream/stuff-it-would-be-nice-if-other-people-would-jump-in-and-tackle:
  • More bits for the Zenbot
    • We now know that the collet in the zenbot takes 1/8" bits (... and given the collet is high-spec it will likely ONLY take 1/8" bits - it might be able to take 3mm bits, but I wouldn't be surprised if they slip given the collet design
    • The bits for the spindle are both expendable (very breakable and will dull if cutting hard stuf) and kinda pricy (new < 1mm blades seem to be around $35 new).
    • As talked about in the last update, multiple parties have said "get used blade off ebay", and while I'm starting to understand what they are talking about, I'm still not ready to plunge into the used blade world, nor has anyone committed to this work.
    • So I can keep working, I'm going to personally order a few new blades for my personal use and use the experience to better learn about the used market. For anyone else thinking about using the zenbot in the near term, I guess what I'm saying is "for the time being, the zenbot is a 'bring your own bits' device". (I'll add a bit ordering guide to the wiki as I better understand the process)
  • X, Y, Z limit switches
    • Chris P has some switches which may be quite good for this, we are not sure when or if we will get to this though (neither of us are mechanicals, so making readjustable and robust mounting brackets for switches is a bit baffling to us)
  • X, Y homing and measuring the cutter in the Z
    • X & Y homing might be doable with the limit switches depending on what the controller can do. Measuring Z will likely require something new. Chris P has this in mind when he is working with the X,Y,Z controller but no new work has been done on it since last update
  • Integrated vacuum system
    • Gerrit said at last meeting that he might have a vacuum that could be used for this. I have a 15A ac power relay (http://adafruit.com/products/268) that might also be able to be used, but no planning other then ideas of parts have been done.
  • Sacrificial milling surface overlays and/or jigs
    • Chris P, Gerrit and I have all agreed that it would be very good idea to build some type of sacrificial overlay over the Zenbot milling surface given how easy it is to drive the blade right through the mills surface (the -z limit won't help for this, the spindle is such that the cutting blade would always go through the mills surface before the limit hit)

With that said, no one has picked up the ball on this, and I doubt Chris P or I would any time soon. Help would be welcome, especially if you know what your doing with the big CNC and/or know how useful jigs would be built for specific tasks, like circuit board milling

  • All-in-one enclosure
    • Gerrit has been dreaming of an all-in-one enclosure for the zenbot, with its controls, and the computer system that LinuxCNC runs on packed underneath the zenbot. (something like http://buildyourcnc.com/redSproutCNCComputerSystem.aspx under the zenbot) Chris P has continued to refine/repack the control electronics and this may be relatively straightforward in the furture but no one has taken the ball for this task.

The CNC is nearly at the point where different tracks of work will need to be done for different projects (molds vs. circuit boards).

I'm interested in 3d mold making and to get the ball rolling for that sub-project on the zenbot, I've started a new wiki page about that use: http://wiki.lvl1.org/Copying_Guerrilla_guide_to_cnc-mold-casting (I've also ordered a board of Renshape 460 - millable plastic for mold making - that I would be up for selling parts off of for anyone else wanting to try 3d part mold making)

I know that a number of other people in the group are interested in cutting circuit boards, thats not really my cup-of-tea (at least not until I have my brain wrapped around the mold making stuff) so I'm hoping some of the other interested parties will jump in, start researching the software/router-bit/jigs that are required to do that work, and building documentation to teach others.

-Mike




Nov 8 Mike L

The included bit has been measured and is 1/8", pressumibly this is really the only size the collet can hold (Might be able to hold a 3mm bit but the teeth on the collet were so close, a 3mm might slip)

Numerous pictures were taken of the collet system, later to be use for instruction.


First (Clockwise, if the spindle is viewed from the top down) cut with the Zenbot since its repair.
Second, counter-clockwise cut (from our examination we feel that clockwise is the correct way to use the bit that came with the machine)

Nov 6 Mike L

Through heroic effort of Chris P, all the major drive systems of the CNC are now online (Power found for spindle, drive belts found for spindle, software for X,Y,Z all calibrated)

With the help of Chris C, Nathan and a few others the Zenbot was moved downstairs next to the big CNC.

An initial cut was done in machine wax of the "LinuxCNC" default on LinuxCNC (only the "L", "i", and 1/4 of the "n" fit on the wax)


Nov 5 Mike L

Copy of status update email:

Progress made:
  • The x and y stepper control wires were switched and the Axis config rewritten for X,Y,Z such that motion is true to Zenbot docs and is also much closer to real-world measurement (Was > 730% off, now estimated at < 1% of real world) - Chris P
  • Belts for the spindle are on order - delivery unclear, worst case expected before Nov 16th - Mike L
  • Motor controller blown - replacement motor control IC expected yet this week - Chris P
Next actions:
  • Directly test spindle motor make sure its isn't the cause of the trouble with the controller - Mike L & Chris P
  • (as parts come in)
    • Install/test belts connecting the spindle motor to the spindle rotor. - Mike L
    • Repair/troubleshoot spindle motor controller. - Chris P
  • Run a test milling using soft material (machinable wax) - Mike L
  • Run test milling to better lock-in the real-world vs. software measurement accuracy. - Chris P & Mike L
Unaddressed needs:
  • More spindle bits - have been told to "carefully" order off of ebay
    • Gerrit may pick this up, but if anyone else out there wants to take charge of getting and calibrating rough and fine milling bits off of ebay let me know.
Future/dream stuff:
  • Add limit switches to X,Y,Z -- important to prolong the life of the machine
    • Chris P has switches on order - still lots of planning to figure out how to install and use
  • Add homing switches for X,Y,Z -- required to allow high resolution milling and/or double sided alignment for circuit board cutting
    • Chris P has switches on order - still lots of planning to figure out how to install and use
  • Sacrificial milling surface overlays and/or jigs - surfaces that would screw in the zenbot surface to increasing the distance the mill would have to improperly travel before it started destroying the Zenbots primary surface. Best made on-mass, likely using the big CNC in the basement. A subset should also be created for specialized use of the mill (alignment pegs for circuit board milling, etc)
    • No work done, no one offering to do the work
  • All-in-one enclosure -- Desire of Gerrit
    • One of Gerrit's ideas - no significant planning or sourcing has been done to make this happen
  • Vacuum system -- Desire of Gerrit
    • Another desired feature Gerrit would like to see - no significant planning or sourcing has been done to make this happen
An aside:
For the last two weeks I've been working-on, writing-up and PM'ing anything I've found on the Zenbot because I'm highly amused by it and I've had a small block of extra time. I've been taking the communication approach you've seen (all info on wiki plus occasional email updates) because I expect I may suddenly run out of time and/or others may choose to work on the machine, and I both want to enable the project to continue with or without me and avoid people repeating others work.
With that said, I'm not under any illusion that I have sole (or even any) exclusivity to the machine and I apologize if anyone has taken my communication to mean I am claiming dominion over it.
I've had one person walk up to me, wanting me to take a different approach to the way I was proceeding with the machine, when I explained that I wanted to continue in the course I had planned (after all this is hobby work for me, I want to do it as I see fit) they were quite unhappy me. It didn't occur to me at the time, but what I needed to say to them was, "while I'm not going to do what you want, if you want to do that, go ahead, but it would be cool if you would add to the notes on the wiki so we all know what the current state is", instead I think my past communications may have lead them to think had dominion over the machine and I was forbidding them from trying what they wanted. That was definitely not what I meant to do.

Nov 4 Mike L

Moved the page now that the model is clear - its a 1216

The reason is that 12" by 16" is the x/y travel of the head, but the measurement of the baseboard of the CNC is roughly 16 by 24 (causing many of us to think it is zenbot 1624). If you look on the zenbot product page, the specs for the 1216 is:
  • Travels: X-12”, Y-16”, Z-4”
  • Table size: 23” x 16.5”
  • Footprint: 28” x 24”
  • Motors: 4 wire bipolar 285 ounce/in Nema-23
Which matches the machine the space has.

Nov 3 Chris P

The manual for the Zenbot 1216 states that "the drive pulley diameter on this machine is .458 inches. The circumference of the drive pulley is 1.439. This means that the motor rotates .695 turns per inch (1/1.439). The stepper motors on this machine are .9 degree per step motor, which means that there are 400 steps per revolution. Now you just need to know the micro-step settings on your drivers. The formula for calculating the steps per inch is .695 x 400 x micro-steps/step. For a 1/8 micro-step driver, the formula would be .695 x 400 x 8 = 2224 steps per inch."

The stepper motors on the Zenbot 1216 at the LVL1 space are labeled as 1.8 degree per step. The steppers have 200 steps per revolution. The DIP switches on the driver are configured for 1/8 micro-steps. Using the same formula 0.695 x 200 x 8 = 1112 steps per inch.

Settings for LinuxCNC:
Driver Type = Other
Step Time = 10000 (unsure, using the recommended value)
Step Space = 10000 (unsure, using the recommended value)
Direction Hold = 200000 (unsure, using the recommended value)
Direction Setup = 200000 (unsure, using the recommended value)
First Parport = 0x378
Base Period Maximum Jitter = 10816 (Measured with latency test, could be higher in which case
  a dedicated video card should be installed.  Current PC has a low profile PCI-E slot).

Parallel Port
Pin 1 = unused (inverted)
Pin 2 = X Step
Pin 3 = X Direction
Pin 4 = Y Step
Pin 5 = Y Direction
Pin 6 = Z Step
Pin 7 = Z Direction
Pin 8 = Spindle CW (future)
Pin 9 = Spindle CCW (future)
Pin 14 = Spindle PWM (future, inverted)
Pin 16 = unused
Pin 17 = unused (inverted)
Pin 10 = unused
Pin 11 = unused
Pin 12 = unused
Pin 13 = unused
Pin 15 = unused

X Axis
Motor Steps Per Revolution = 200
Driver Microstepping = 8
Pulley Ratio = 1:1
Leadscrew Pitch = -0.694927 (negative causes axis to move in opposite direction) 
Maximum Velocity = 3.33 (using the recommended 200 inches per minute from Zenbot manual)
Maximum Acceleration = 40 (using the recommended 40 inches/sec/sec from Zenbot manual)
Home Location = 0
Table Travel = 0 to 12

Y Axis
Motor Steps Per Revolution = 200
Driver Microstepping = 8
Pulley Ratio = 1:1
Leadscrew Pitch = -0.694927 (negative causes axis to move in opposite direction) 
Maximum Velocity = 3.33 (using the recommended 200 inches per minute from Zenbot manual)
Maximum Acceleration = 40 (using the recommended 40 inches/sec/sec from Zenbot manual)
Home Location = 0
Table Travel = 0 to 16

Z Axis
Motor Steps Per Revolution = 200
Driver Microstepping = 8
Pulley Ratio = 1:1
Leadscrew Pitch = -0.694927 (negative causes axis to move in opposite direction) 
Maximum Velocity = 3.33 (using the recommended 200 inches per minute from Zenbot manual)
Maximum Acceleration = 40 (using the recommended 40 inches/sec/sec from Zenbot manual)
Home Location = 0
Table Travel = -4 to 0

Spindle (future, using guessed numbers now)
PWM Rate = 5000 Hz (MD30B can do up to 10000 Hz)
Speed 1 = 100
Speed 2 = 800
PWM 1 = 0.2
PWM 2 = 0.8

The MD30B spindle motor driver was damaged during initial test of power supply, parts are ordered/shipped for repair.


Close-up of the printing on the Zenbot's spindle motor

Oct 29 Mike L

Info found on motor:

12 VDC MOTOR
Mabuchi # RS-555VD.
12Vdc, 750mA (no-load).
Operates over 8-18 Vdc range.
38.5mm diameter x 65mm long.
3.17mm (1/8") diameter shaft. Solder lug / 0.18" qc terminals.
CAT# DCM-354
$3.50 each

From http://www.scribd.com/doc/66758727/42/DC-MOTORS page 79

In order to get something going on the belts, I ordered:

Products
------------------------------------------------------
2 x 3/16" (4.8mm) Orange 83A O-ring Belt = $8.00
	Cut-Length 5
2 x 3/16" (4.8mm) Orange 83A O-ring Belt = $7.78
	Cut-Length 5.5
2 x 3/16" (4.8mm) Orange 83A O-ring Belt = $7.86
	Cut-Length 5.3125
2 x 5/32" (4mm) Clear 83A O-ring Belt = $7.10
	Cut-Length 5.25
------------------------------------------------------

from Ebelting.com


Oct 28 Mike L

Working to figure out which belt to acquire.

The durabelt people have a calculator for spec'ing out a belt at http://durabelt.com/beltlengthcalculator.php. Method three seems the most promising given the measurements Chris and I took.

After two pages of trying to work out the math, I came up with a length of 146.35mm (5.76181") for the unstretched length, durabelts calculator gave me the same thing (if I said 0 percentage stretch.

Still not sure if I have this right though. The belt vender sells belts in 1/16" increments.


Oct 26 2012 - Mike L

Diameter of belt grove on motor side
Diameter of belt grove on spindle
Distance grove to grove

Zenbot has been moved to a desk in the front (broadway facing door) room.

It is missing one of the rubber feet (that might have been added aftermarket). May be easiest to just replace all four - the three that are still on are clearly "stick on" type. There is no special size or shape required.

PC originally wouldn't boot. Chris P. futzed with the ram in the machine (added/removed/cleaned-with-eraser) and the machine became willing to boot.

Once machine was booting the EMC2/linuxcnc?/axis-software was willing to run, and upon turning on the X,Y,Z controller, the software would drive the armature.

The machine is sitting with its orignal QR code facing out, this seem to be a good way point it so-as to have the x driver-motor pointing out at the user as it slides left to right BUT the config on the machine currently has both x and y configured for the machine rotated 180 degrees. I'm working to understand the config, with the expectation that I'm going to reverse both the x and the y (the z is fine)

Also critically important, all the axis are WAY out of scale. Moving the head 1.4 inches in the controlling software really moves the head very rapidly 10 inches in real space. Given the machine only has 10-20 inches of travel on x and y this makes a really easy to crash the head into the end of the rails its mounted on. Given the machine has no sensors for the end of the rails, if the software commands it there it will simply slam the head into the rails and grind the motors trying to keep moving on. (specifically moving 10" in real space read 1.3892" on the software, moving the head 10" on the y axis in real space read as 1.3762" on the software, z wasn't measured)

To the right are pictures of measurements of the spindle belt groves, with the goal of computing what size belt we need to acquire

Chris found a "O ring style belt" custom belt maker. dura-belt durabelt.com is the manufacture, but they don't sell retail. ebelting.com is one of the retails for them.

--

Chris Pettus did some googling and found another zenbot user using a Wolfgang spindle and powering it with a computer power supply: https://plus.google.com/103014027224354212001/posts/9EELxYozG4W


Oct 25 2012 - Mike L

In a back-and-forth with Chorgy he said: "The spindle I ran using a generic 13v regulated bench top that I bought from radio shack for like $30-$40." Which I'm somewhat unclear if he donated but regardless is not in the pile of equipment with the Zenbot.

Chorgy also added that he really can't see the motor drawing more then 20A 12v dc max (.. through the motor controller board) and so he though a computer power supply would work.


QR code label on the side of the Zenbot

Oct 24 2012 - Mike L

Warped photo of the QR code that attached to the Zenbot


Oct 23 2012 - Mike L

Following up on what I found out about the zenbot:

  • Chris connected me to Gerrit, he is the writer of the note (that I sent earlier) on the zenbot and seems to be on top of what needs to be done to get it going. (I.E. he does "have the ball")
  • Current status
    • Believed capable of high precision operation other then a non-operating spindle (the router part) because of multiple missing parts, and likely incompletely configured control software.
  • Parts needed to get the zenbot operating:
    • Power supply for the spindle motor
      • either found by chorgy or an expense of $75-150
    • Belts for spindle
      • Gerrit had an exchange with Wolfgang Eng (maker of the spindle part) and thinks he may have enough info to order from a parts house (price currently unknown, though expected to be low)
    • More bits (there is one with the zenbot now)
      • Both Gerrit and chorgy agree that ebay is the way to go. Gerrit's comment was that the reason that ebay bits are so cheap is because they are partly used and have become slightly out of dimensional spec. With effort, the bits can be re-measure there-by enabling them to be used at full precision.
    • Config for EMC2? (x,y,z stepper motor control software)
      • The current assumption is that the EMC2 on the linux box doesn't have a fully configured config for the zenbot. Gerrit has contacted the zenbot and they have said they will send him a (customer supplied) EMC2 config, though they haven't yet so he is planning on bugging them again.
      • Brian W posted that he had x,y,z working a few months back, Brian if you think the config is complete then please reply so Gerrit doesn't have to keep hassling the zenbot folks

If the zenbot gets up and running, there are number of improvement that Gerrit is already thinking about ("stops" to calibrate the arms - likely requiring a different controller, enclosure for the control systems, vacuum system, etc..)

The consensus seems to be that machine is in good shape, and should really be capable of high accuracy milling if the parts/config are done and the machine is treated gently.

I have some extra time over the next few weeks and I'm going to try and help Gerrit move things forward, I suspect that may not be enough to get the system operational, but if nothing else I'll make sure to document what happens.


Pre Oct 20 2012

Had written notes from Gerrit:

Gerrit's handwritten notes on the zenbot
Personal tools
Namespaces

Variants
Actions
Navigation
Toolbox