User:Mcmaster/cnc microscope

From Microwiki
Jump to: navigation, search

First section is about what the system does as of today. Below that shows how it arrived there

"pr0nscope"

Current configuration

Driver controller: New Focus ethernet controller

Joystick: New Focus MODEL 8754 PICOMOTOR JOYSTICK

XY stage

Klinger/Micro-Controle/Newport variable reluctance UT100-50PP (50 mm). I've read these are rated to 0.1 um accuracy. IIRC they use UE30PP motor but you can get a stepper version.

However, instead of using the variable reluctance motor a stepper motor was attached via a custom machined adapter plate, flexible shaft coupler, and 4-40 standoffs from discarded computers

Optics

Olympus BH2-UMA illuminator

Olympus LH50A lamphouse

Think Nikon makes the lamp power supply

Misc Mitutoyo Plan Apo, a couple Olympus NeoSPlan DIC

NeoSPlan DIC prism set

Sensitive tint plate

Polarizer, analyzer

AmScope MD800E camera (using Linux GSPCA driver I wrote)

Misc

CD4069 to convert 3.3V indexer logic level to 5V stepper drive logic level

Unknown who makes the base but its very heavy duty

mk0

Failed attempts that never made a working system

0.0

Mcmaster cnc microscope mk0 overview.jpg

Above: manually operated system. Clicked point and shoot camera to take pictures. t-slot aluminum lines it up to the monocle.

0.1

Mcmaster cnc microscope mk0 linear stage springy planning.jpg

Above: stepper motors with (to be added) springs to couple them to the linear stages. IIRC it took way too much force to turn the stags leading the springs to turn the stages in unpredictable bursts.

0.2

Mcmaster cnc microscope mk0 stage almost there perspective.jpg

Above: Instead tried to spring load some gears onto the stages. Slightly non-linear but very predictable. I don't recall what the specific problem was with this system, but probably just difficult to keep the gears aligned

mk1

Purchased a Unitron inverted metallurgical microscope because it was the least expensive metallurgical microscope I could buy ($200 shipped IIRC).

1.0

Mcmaster cnc microscope mk1 a unitron n.png

Above: an unmodified microscope of the same model

1.1

Mcmaster cnc microscope mk1 overview2.jpg

Above: basic idea of setup. Microscope stage has been removed and the microscope has been turned upside down to make it right side up. Point and shoot similar to mk0. I soldered wires onto the shutter controllers and routed them to EMC2 (LinuxCNC) software that ran g-code to move Sherline 2000 mill stages (something I already had…and the only "nice" thing I had at the time). Pictures were snapped by sending coolant mist (focus), coolant flood (snap picture), followed by coolant off (release shutter). $20 laptop with a wireless card from a scrapped access point.

Mcmaster cnc microscope mk1 z axis.jpg

Above: since the original stage had been removed I needed a way to focus it. Linear stage removed from a broken spectrometer found in school dumpster and spring loaded with a rubber band to maintain position (intended to be gravity spring loaded but I'm holding it the "wrong" way)

mk2

2.0

Original microscope without computer control

2.1

Mcmaster cnc microscope mk2 stage.jpg

Above: very first test before installing the tip tilt. Note riser block is still present below stage. This takes a way a lot of my clearance making it so that I couldn't use Mitutoyo objectives

Mcmaster cnc microscope mk2 tip tilt.jpg

Above: tip tilt installed shortly thereafter.

Used motors, mixed types. Minebea size 17 frames

PMC VS series drivers that I found in the trash

"USBIO DAQ" board used to bit bang step/direction signals

2.2

Mcmaster cnc microscope mk2 picomotor prototype.jpg

Added picomotor theta X/Y. have drivers and stage setup for Z but never installed picomotor mostly becasue I never replaced the cable

Ultimately it turned out I didn't really need the picomotors. The real issue was that the original leadscrew was poor quality. I just use them as leadscrews and don't use the motorized capabilities. At first I had problems bumping stage and so used the joystick to avoid touching things but now I'm quicker just adjusting by hand without issue.

2.3

Got new motors: 17PM-K318-04V

  • 17PM-K: [1]
  • General motor guide: [3]

Tried OEM650 drivers but burned out a motor and ultimately went back to old drivers

New motors: TYPE 17PM-K318-04V

  • Yellow A-1
  • Black: A-C
  • Red A-2
  • Blue B-1
  • White B-C
  • Orange B-2

Driver settings (PMC VS series):

Motor current S2-3:S2-8: 001111

   0.95 + 0.19A

switch off selects it

Seems to be stalling...upgrade to 0.38A

   1
       1 1
       2 1
       3 1
       4 1
       5 1
       6 1
       7 0
       8 1
   2
       1 1
       2 1
       3 1
       4 1
       5 0
       6 1
       7 1
       8 1

2.4

Mcmaster cnc microscope mk2 3.jpg

As of 11/09/2013

Recent changes:

  • Use of hardware DIY indexer has enabled microstepping. Significant noise/vibration reduction
  • Implemented acceleration/decceleration logic in hardware indexer
  • On UPS (I haven't had a problem, but its right next to my room UPS)

Y motor drive box may be beginning to see mechanical failure but move to microstepping and hardware indexer has smoothed acceleration profiles significantly which seems to have sufficiently mitigated suspected stripped gear for the time being

Control software

Revisions:

  • pr0ntools: 1dea224a62bef5f0a02e2c41e81d85553db30ab0, Sat Nov 9 20:21:30 2013 -0800
  • uvcada: c550026a99c50a63e70f278c2e98e98d6e0486ef, Sat Nov 9 19:51:01 2013 -0800
  • python: 2.7

Indexer

Revisions:

  • pr0ndexer 010652d4ddc8f06d54ff01062372cb9ba916f12e, Sat Nov 9 19:50:36 2013 -0800
  • Toolchain: arm-none-eabi-gcc (GNU Tools for ARM Embedded Processors) 4.7.4 20130913 (release) [ARM/embedded-4_7-branch revision 202601]
  • STM32 eval borad. CPU rev L1

Driver

Drivers (2): Precision Motion Controls PMC VS series

Switches

 1
     1 0
     2 1
     3 0
     4 1
     5 1
     6 1
     7 0
     8 1
 2
     1 1
     2 1
     3 1
     4 1
     5 0
     6 1
     7 1
     8 1

Switch/current: off means add current

  • S2-3: 0.095 A
  • S2-4: 0.19 A
  • S2-5: 0.38 A
  • S2-6: 0.76 A
  • S2-7: 1.52 A
  • S2-8: 3.04 A

Microstepping:

Level S1-1 S1-2
50000 OFF OFF OFF
25000 ON OFF OFF
*10000 OFF ON OFF
2000 ON ON OFF
51200 OFF OFF ON
25600 ON OFF ON
6400 OFF ON ON
400 (none?) ON ON ON

Theta control

Mirror mount: Newport P100-P2

Motors: Picomotor, model ???

Drivers (2): New Focus something

Adafruit serial adapter

Wires:

  • red: 5V @ 500mA direct from USB port
  • black: ground
  • white: 3.3V RX into USB port
  • green: 3.3V TX out of the USB port

2.5

Mcmaster cnc microscope mk2 4.jpg

07/19/2014: is currently out of service due to broken laptop (and has been for a while). replacement parts in the mail. cleaning up electronics while its down

07/28/2014: back online! Down for a few months but rolling once more

built t-slot cage around desk. mounted pegboard to which I attached the various electronics to clean them up and get them off the desk (the idea cabinet thing just wasn't working very well)

want to add:

  • finish rewiring electronics. either make pcb or use phoenix contact DIN mount screw terminals
  • get a cabinet to store spare carasols loaded with objectives
  • Can fit the confocal microscope next to it now that I have more room on the desk?
  • would be nice to get a real anti-vibration table
  • consider adding continuous running hepa filter at top (make it like laminar flow workstation)
    • problems blowing dies away?

2.6

Mcmaster cnc microscope mk6 bbb.jpg

Controller replaced with BBB

Motors replaced with IMS NEMA 17 integrated drivers

Circa 2015-09-15

Discovered my MDrive17 units are set to 51200 microstepping. This is a bit high...may pose issues getting BBB to step fast enough especially given already high gear ratio

Created high performance BBB configuration. Although my step rate is way smoother and higher than pr0ndexer, the high microstepping meant that I stayed at about the same max velocity (possibly losing some). About 4 rev/s max

"g0 x5120" 5 mm

  • 1000 steps / unit
  • 51200 steps / rev
  • 100 rev => 5 mm
  • 1 rev => 50 um
  • 1 real motor step => 250 nm
  • 1 microstep => ~ 1 nm

Appears to be either missing steps or I have the ratios wrong. Consider setting up a repeatbility test using microscope itself to target something.


2015-08-29

NEMA 17 motor pinout

Original connectors:

  • P1: DB25 controller interface
  • P2: limit switch?
  • P3: limit switch?
  • P4: motor encoder
  • P5: motor main (motor P1)
  • P6: motor power (removed)

P1: controller interface


DB25

Pin

To
Wire color Function
1 P4.1 Red
2 P4.6 Green
3 P4.8 White
4
5
Old: P2.5

New: P5.1
Black
Old: lim sw

New: ground
6
7
P3.2

New: P5.1
Black
Old: lim sw

New: ground
8
9
10
11 P5.3 Red OPTO

5V

Orange

BBB P9.5-6
12 P5.5 Black EN


N/C
13
14 P4.5 Brown
15 P4.7 Blue
16 P4.4 Black
17
P2.1

New: P5.2
White
Old: limit sw

New: +V
18
Old: P2.3

New: N/C
Black
Old: shield

New: N/C
19
Old: P3.1

New: P5.2
White
Old: limit sw

New: +V
20
Old: P3.3

New: N/C
Black
Old: shield

New: N/C
21
22
23
24 P5.4 Green
SCLK

("step")


X: BBB P8.13

Y: BBB P8.15

Blue
25 P5.6 White DIR

X: BBB P8.12

Y: BBB P8.14

White

NOTE: original cables had these two ports which were repurposed into power lines:

  • P2
  • P3

Probably intended to be used for limit switches

P2/P3: limit switches

P2 (original wiring):


P2

Pin

DB25 pin
Wire color Function
1 17 White
2 5 Black
3 18 Black Shield N/C as power

P3 (original wiring):


P3

Pin

DB25 pin
Wire color Function
1 19 White
2 7 Black
3 20 Black Shield N/C as power

P4: encoder



Mot

P1
P1 Wire color Function
1 1 Red
2 N/C N/C
3 N/C N/C
4 16 Black
5 14 Brown
6 2 Green
7 15 Blue
8 3 White
9 N/C N/C
10 N/C N/C

P5: motor main

2.4.3 Motor P1 - Power, communications and logic (universal inputs)



Mot

P1
P1 Wire color Function
1
5

7
Black Ground
2
17

19
Red V+
3 11 Red OPTO
4 24 Green SCLK
5
12
Black EN
6 25 White DIR
7 N/C N/C +5V
8 N/C N/C CLK
9 N/C N/C CGND
10 N/C N/C MOSI
11 N/C N/C CS
12 N/C N/C MISO

P6: power

Removed. Never pinned out but was just red + black wire

Misc notes

T7 star to take off BH2-UMA

7/64" allen to remove nosepiece adapter from Z assembly

5/32" allen to remove fine/medium focusing assembly from Z assembly

3/16" allen to remove fine/medium backplate to expose gears

7/16" socket to remove fine/medium knobs (remove circular cover first)

  • left side comes off, right side doesn't
  • remove by holding left fine focus in place and turning