Makibox issues

After I assembled my Makibox last night, I ran into a number of software issues.

Firmware: current 2.11
OS: MacOS
Host:

1 Lockups

Makible 5D Print in Google Chrome

In layer 5 of a test shotglass it just stopped and locked up. Wouldn't resume after power-cycling the printer and using the pause+resume function.
No error messages.

No manual jog controls for the extruder or Z axis => can't quickly move the head to a safe height and retract the filament after a failure.

Pronterface

Stock Pronterface/Printrun with Makibox settings.
I can jog each axis but loading any gcode and executing it, nothing happens.
It stays in line 0 (line 0 is a comment), doesn't heat up, doesn't home, nothing.

Update: gave Pronterface another try. This time it started printing but locked up just like 5D print did. So clearly this is a firmware issue or an issue in the common codebase (5D print started with a fork of Pronterface or so I've reat.)

Pronterface is incabable of stopping a print.
Pause, Reset, Stop, even Disconnect doesnt work. On Disconnect it executes the last 2 lines of g-code over and over again.This means the firmware has a bug and doesn't forget about already executed commands. At most it should finish the current command and then stop!


RepetierHost

I tried RepetierHost with these settings.
I cannot navigate to any but the first page in the preferences window. ("Host", "Slicer", "3D visualisation", "Colors")
It seems to not update the graph or textual output of the current temperatures.
Do NOT check the "( ) printer sends OK even after error" box for the Makibox to work.

....it just locks up like all the others.

Temporary solution: I switch from a Macbook Pro to a desktop iMac and a different USB cable.

Findings:

• On the MacBook Pro with MacOS 10.9 "Maverick" the printer just randomly locks up independent of the host-software used.
• On the iMac with MacOS 10.8.5 this doesn't happen.
• It seems to happen less on the iMac when I move the mouse every few minuted to prevent any kind of light sleep mode.

Currently investigating the USB port power requirements as a possible reason.
(Original Printsboard allows for being powered by USB or internal regulators from it's stepper/heater power supply.)
If so, a powered USB hub should solve the lockups.


Solution: Raspberry Pi with Octoprint

2 Stalls and filament grinding

At random points in the print, the filament stops.
The drive gear has ground a hole into the filament.
The filament above that hole is nearly half ground away by the drive wheel.

1. Repeated the bed leveling and all the tuning steps on the Makibox support page
2. The hole for the axis of the plastic wheel in the extruder was too small. sanded it and the wheel now spins freely. (all injection molded parts where oversized on this Makibox.)
3. Raised the temperature of the supplied "makibox HT test cube v0.2 PLA" from 192 to 218°C.
4. Retract the filament after printing. Don't leave the filament to cool inside the hotend. Don't re-insert it before the hotend is <120 br="">

 Currently testing....

Possible next steps:

1. Sanding down the surface of that springly plastic wheel to reduce friction.
2. lubricating the rods and the axis in the extruder.
3. Designing a different extruder with multiple drive-gears to increase grip without increasing pressure on the filament.

3. Slanted prints

...still working on this issue.

1. Check your stepper voltages. Maybe go slightly above the recommed 1.45V reference volrage
2. Check that your axis and hotend aren't fastened too tight.
3. Check the motor coupling for breakage or being loose
4. Check the bed-leveling bolts.
5. ...TODO

• Makibox forum thread about lockups
• Makibox forum thread II about lockups
• Makibox forum thread about jamming 
• Pronterface settings
• Some Makibox firmware modifications that could be interesting in the future 
• unofficial FAQ 
• Makibox Google Group

Assembling my Makibox

I received my Makibox HT and am trying to assemble it as per the
assembly instructions.

General improvements

• Label the bags!
• Numerate and label the parts! 
• Slot and tap contructions are always tedious. Especially in soft materials. It would have helped to tapper the taps (or the short side of the slots) of the slots to make insertion easier without affecting final stiffness. I get the impression that nobody ever thought about tollerances and how the parts have to move during assembly. A 5mm peg does NOT fit a 5mm hole. You make a 4.99mm peg and a 5.01mm hole!
• Other kits often include 1 extra back with 1 instance of each bolt type used. So you have 1 spare in case one gets lots and don't get confused with having bolts left. (because they come in an extra bag)
• The direct drive extruder is extremely prone to  get stuck and grind into the filament. It cannot transfer enough force, so a leadscrew or at least 2 driven gears on either side would have been a better idea.

Step 1

• There is a vertical feature on one of the 170mm parts that is missing in the drawing. Thus being confusing.
• Makible should label the parts by having a number be part of the mold. So you can refer to them.
• Step C looks like there are 2 of the parts used in step B. Image isn't clear enough.
• Use larger allowances for shrinkage in their taps

Step 2a+b

• My kit was missing the M4 washers. 
• The only M4x8 bolts in the kit have a different head (round) then in the image (flat).
• ...and need a different size of hex key then in the video.
• You need a broad screwdriver to hold the part for step B because you can't hold these with your fingers. (Too tight thread). In the video this is done by hand using the larger hex key...impossible in my kit because the bolts needs the smaller hex key and the part is made out of a denser material. One way is to clamp the screwdriver between your legs, hold the plate with one hand and very carefully turn the hex key with the other hand.

• The round parts have no thread to hold the M4x8 bolts but these types of bolts are not ment to cut their own thread (That's why you need pliers. To force them).

Step 2c

• A limit switch on the Y plate collides with a motor mount of the X plate. They don't fit. It looks like the limit switch on the X+Z plate is 4mm too long for it's position.
• The M4x8 bolt thus can't reach the box on the other side that it is supposed to cut into. :/

• Solution: That small, rectangular tab at the base of the cylindrical motor mount should not be there.
  If blocks the end stop and thus prevents the plates from matching up.
• Problem: That M4x8 bolt is way too short! If you can, get a longer bolt to replace it!

Step 3

• If you have magnetic (metric) hex keys, that could help in the fiddling with these tiny M3 bolts.
• As always with steppers, check the orientation of the cable in the drawing to sew how to insert the stepper.
• The image is not showing where one of the end-stop cables has to go when attaching the Z stage. 
• The z axis in step C has no bolts securing it in place. It falls right of then you lift the assembly to continue with step 4.

Step 4

• There are multiple parts that fit the drawing in 4a.
• The parts added in 4a will fall off during 4b-d all the time.
• 4d: You have to bend the thick wires to fit through the slots.
• In 4d the on-off switch is not shown in the drawing and it it completely unobvious where it has to go.
• Doing 4b before 4d makes fiddling the plug trough in 4d more tiresome then it needs to be.
• In 4c it is not obvious from the drawing what part to use. (2 parts fit,  4 parts look like the drawing. 1 part is correct)
• Seriously....Makible wants you to join about a dozen taps at once? In a tap&slot construction with no tapper and 0 allowances? Did they ever assembly any tap&slot or T-slot construction before? This is insane!
• Tools required: flat rasp and mallet!
• The bolts are too short to compress the parts together!

Step 5

• This step requires some force. It would actually be easier if the part was not pre-assembled because the rods block the way for the hex-key.
• For Step 5 B and C the flat part is one of the metal parts.

Step 6

• easy

Step 7

• 7b is tricky. The bolt has to cut it's own thread, yet there is not support form the other side. Why the plastic part had not been designed to fit is a mystery.
• Page 7 (like others) doesn't follow western reading traditions of left to right and then up to down. So make sure where steps A, B and C are and execute them in alphabetic order.
• The spring loaded platform is something I know from the old Repman. However this one is not good. The holes in the heated bed should be larger then the bolts! This way the plate tilts and then cants on the bolts and the springs have no effect anymore.
• The position of the bolts makes leveling the bed and adjusting extruder height a nightmare!

Step 8

• 8c : I'm not sure about their way of routing the cables. I usually try to keep cables away from heated metal plates. 
• Warning! The printed instructions forgets to tell you that the bowden tube has 2 DIFFERENT ends. Check this video to see what end goes where!(The end with the deeper support goes in the hotend)

Step 9

• I'm not sure that using bolts on acrylics without washers is a good idea.
• Attaching the edges is very difficult without using a thread-cutter first.
• Force the bolts into the edged first without the acrylivs, then remove them and then mount them. This requires enough force that the bolt actually gets hot (no powertools involved). Again this would be much simpler with another type of bolt that was designed to cut it's own thread.
• Do NOT put that cable through there like instucted. It's too short later!

Step 10

• Yes, you have to partly disassemble your preassembled extruder. That's correct.

Step 11

• If you mounted your Z stepper wrong...now you'll regret it.
• Also...what direction is X and what is Y? There is no imprinted hint! (Y is the longer axis)
• HE?
• HB? Maybe "Heated Bed" 
• The drawing shows 3 endstop-headers but the board has 4. 
• The video doesn't do the cabling in step 11 but aparently wires X and Y in some previous step. 

Step 13

• aparently as with any other hole nobody thought of allowances and tollerances for the hole for the power plug. Even though that hole could easily be 8mm bigger and still nobody would notice.
• I'm not exactly looking forward to installing Google Chrome just to run the damn printer software. A standalone Java program would have done exactly the same without installing such a thing on my Macbook. I don't see why I need to enter my Google account somewhere to use my local and offline 3D printer.

After Step 13 - Software

1. Power up the board with the power supply and plug in the micro-USB cable
2. Remove the jumper in the middle of the board at connector JP14

• Right after we are told how to screw the case shut above the board..
• I absolutely don't like that the host app is a Google Chrome app. Standalone Java already has a serial port API and doesn't require a new browser that auto updates without asking me, no knowledge of my Google Account and works!
• Installed the Google Chrome app.
• "Über die Chrome-App-Übersicht können Sie Ihre Lieblings-Apps direkt vom Desktop aus starten." Had to search a while to find that this damn thing has nested in my launcher WITHOUT ASKING FOR PERMISSION TO DO SO.
• 5D print says "no device". No hint how to tell it that my printer is already connected and running.
• bed leveling is a nigthmare.
• The PID control is not good. I tell the bed to go to 35°C...it goes to 35....then slowly to 36....37....38..stays at 38. 

Links

• official assembly instructions
• one inofficial assembly instruction

Fried 2 other FMD2740A stepper drivers

This is the second time I fried 2 FMD2749A stepper drivers on my TravelCNC.

Made 2 giant Hackcenter calls at the 30C3 hacker congress stink of fried electronics while the people from CCCfr ripped out all power cables they could find, called me and got a fire extinguisher.

The drivers are advertised to be operated at up to 50V.

The first one had a 5V DC converter (LM2576S-5.0) inside that is only rated up  45V instead of an " LM2576HVS-5.0" rated up to 60V.

I'm using them with a common 48V power supply. Well within the rating of the dealer.

Now, after 3 fried stepper drivers the dealer "Thomas Wu" answered that the drivers where actually only good for 45V!!!

I wouldn't ever but any power electronics from this Thomas Wu guy on Aliexpress again!

I turned down the power supply to 44.5V (can't go lower then that on the potentiometer)

and am waiting for some new 5A drivers with over-current, over-heating, under-current,... protection.

One of these drivers is working fine at 48V driving the super-strong 4th axis, so I hope they will hold up.

 

 

Reverse Engineering (Micro) Four Thirds lens protocol

Based on the work of Lasse Bayer, who modified an Olympus lens to attach a logic analyser,
I'm trying to reverse engineer the rest of the Micro 4/3 lens protocol.

I found an interesting macro adapter that sits between mft camera and mft lens and has electrical contacts.
Thus I'm trying to modify this adapter and analyze not only one lens but a large number of combinations of lenses and cameras.

I'm currently working on this.

This blog-posting will be updated all the time!
Here are the logs!

Setup 

 

 

 

Pinout:

VDD=microcontroller power (goes low on lens-release button)
VCC=battery power (possibly for AF-motor,...)

preliminary MFT pinout:

signal	logic analyser cable	logic analyser channel	position
ACK?	BROWN	CH4	Right when Facing the Sensor
ACK?	RED	CH2
Clock	ORANGE	CH3
Data	YELLOW	CH1
Lens2Body?	GREEN	CH0
Body2Lens	BLUE	CH5
	GREY	CH7
???	PINK	CH6
	WHITE	GND
	BLACK	GND
FEHLER(N/C)	---	---	Left when facing the sensor

 

Quote about the 2 additional contacts (from here):

In addition, the lens mount of the Micro Four Thirds System is equipped with two additional signal contacts for smoother Live View shooting with shorter time lags, faster higher communication speeds between the lens and body, and, of course, reduced lens and camera size.
These two contacts will also be used in high-speed data processing required for the movie handling capability expected in the future.

(Image again taken from Lasse Bayer)


SPI seems to be used at 500KHz with 8 bits LSB. Lens2Body and Body2Lens are pulled high by lens or body respectively to indicate that they are ready to receive/send the next byte and pulled low again to acknowledge the receipt.

The camera seems to pull the lines low (including clock) to pause communication.

• when the shutter is pressed
• when exposure has finished 

(confirmed for MFT)

General protocol

Everything seems to start with 4 byte commands sent by the camera to the lens,
followed by the lens answering.

Initial communication 

FT:  B0 F0 00 00  (first command on switch-on)

MFT: B0 F0 00 00 

FT: 

Followed by 2 values: A0 00 

MFT:

Followed by 2 values: 00 00 

Command: request lens data (MFT)

MFT: 80 ED 01 00

TODO: followed by some yet unknown data

Command: request lens data (FT)

FT:  C0 F0 20 00

There is quite a lengthy answer in 26 blocks: 

 Block 0

lens Id
lens designation(ASCII)
lens serial number
lens firmware version
number and length of the following blocks

 TODO : details

Block 1

 May be a list of all focal lengthes in mm - 1

Block 2

unknown

Block 3

unknown

Block 4

unknown

Block 5

unknown

Block 6

unknown

Block 7

unknown

Block 8

unknown

Block 9

unknown

Block 10

unknown

Block 11

unknown

Block 12

unknown

Block 13

unknown

Block 14

unknown

Block 15

unknown

Block 16

unknown

Block 17

unknown

Block 18

unknown

Block19

unknown

Block20

unknown

Block21

unknown

Block22

unknown

Block23

unknown

Block24

unknown

Block24

unknown

Block25

unknown

Block26

unknown

Regular polling (MFT)

 TODO: MFT seems to use multiple, regular polls.

81 E3 04 03
followed by yet unknwn data that does change while not touching the camera.

Regular polling (FT)

C0 F1 0 0 

Apparently the camera polls the lens every 104ms sending these 4 bytes.
The lens answers with 14 bytes including information on the current focus, focus mode, zoom distance, and a checksum.

(I'll insert details here)

Here is what Lasse found out:

 Byte Number:

 1 ,  2 ,  3 ,  4 ,  5 ,  6 ,  7 ,  8 ,  9 ,  10,  11,  12,  13,  14,  15,  16,  17,  18
----------------------------------------------------------------------------------------

    Infinity focus:

192, 241,   0,   0, 177,  11,   0, 160,   0,  31, 255, 255,   2,   5,   0,   0,   0, 196

    Focusing closer:

192, 241,   0,   0, 177,  11,   0, 128,   0,  31,  43,   1,   2,   5,  11,   2,   0, 223
192, 241,   0,   0, 177,  11,   0, 128,   0,  31, 155,   0,   2,   5,  85,   3,   0, 153
192, 241,   0,   0, 177,  11,   0, 128,   0,  31,  87,   0,   2,   5, 121,   5,   0, 123

    Close focus:

192, 241,   0,   0, 177,  11,   0, 144,   0,  31,  78,   0,   2,   5, 254,   5,   0,   7

----------------------------------------------------------------------------------------

Byte 8:   0 - 0000 0000 - Single Autofocus active
        160 - 1010 0000 - Manual Focus @ infinity
        144 - 1001 0000 - Manual Focus @ close-up
        128 - 1000 0000 - Manual Focus in between

Byte 10: Changes linearly while zooming from 31 @ 45mm to 0 @ 14mm

Bytes 11 + 12: FF FF - 65535 - infinity focus
               4E 00 -    78 - close focus
               Focus distance ?

Bytes 13 + 14: Seem to depend on the zoom position at switch-on.
               Do not change while zooming or focusing manually.

Bytes 15 + 16: 00 00 -     0 - infinity focus
               FE 05 -  1534 - close focus
               Focus mechanism position ?

Byte 18: Checksum of Bytes 8 to 17, maybe 7 to 17

The first 4 Bytes are sent from Body to Lens, the remaining 14 from Lens to Body.

Command: Start/Stop exporure (FT)

A0 B0 FE  0  ? stop of exposure ?
A0 B1  0  0  ? start of exposure ?

 TODO: different on MFT

Command: Change Aperture (FT)

Like the polling is a 4 byte sequence, this seems to be 4 byte too.

close aperture: 176  210 xxx xxx
open aperture:  176  215 0   0

 TODO: verify list of epertures

Aperture values as determined by Lasse:

aperture         command
value               data
                   
estimated
                   
1.0               0    0
1.1              85    0
1.2             170    0
1.4               0    1
1.6              85    1
1.8             170    1
2                 0    2
2.2              85    2
2.5             170    2
2.8               0    3
3.2              85    3
3.5             170    3
                   
determined                my own notes:

4                 0    4  (base value for comparison) 
4.5              85    4   (+85 )
5.0             170    4   (+170)
5.6               0    5   (+256)
6.3              85    5
7.1             170    5
8                 0    6
9                85    6
10              170    6
11                0    7
13               85    7
14              170    7
16                0    8
18               85    8
20              170    8
22                0    9

 determined
                   
3.5             157  131     my own notes:
3.6             171  131
3.7             198  131
3.8             211  131
3.9             236  131
4.0             248  131     (base value for comparison) 
4.1              17  132
4.2              28  132
4.3              51  132
4.4              62  132
4.5              84  132     (+77 instead of +85 =-10%)
4.6             106  132
4.7             116  132
4.8             137  132
4.9             156  132
5.0             166  132     (+162 instead of +170 =-5%)
5.1             176  132
5.2             195  132
5.3             213  132
5.4             222  132
5.5             231  132
5.6             249  132     (+257 instead of +256 =+0,4%)

Quote from a camera manufacuter I recently got:

"The electronic feedback from the lens tells the camera where it is in 
regards to f-stop, and the camera then interprets that information." 

Command:  shutdown (MFT)

Sadly the shutdown sequence seems to use different commands.
No clear command sequence to see yet.

-->

14-42PZ	14-42PZ	14-42PZ		14mm	14mm	14mm	14mm
23,89ms	12,32ms	32,09ms		14,06ms	29,83ms	21,92ms	23,86ms
SDO: A1 D7 00 00 11 05 7D 07	SDO: F1 CF 00 00 11 05 7D 07	SDO: B1 D7 00 00 11 05 7D 07		SDO: 31 CF 00 00 08 04 3F 02	SDO: 31 C7 00 00 08 02 BF 00	SDO: 21 9E 00 00 00 00 3F 00	SDO: F0 9E 00 00 80 00 BE 06
SDI: A1 D7 00 00 11 05 7D 07	SDI: F1 CF 00 00 11 05 7D 07	SDI: B1 D7 00 00 11 05 7D 07		SDI: 31 CF 00 00 08 04 3F 02	SDI: 31 C7 00 00 08 02 BF 00	SDI: 21 9E 00 00 00 00 3F 00	SDI: F0 9E 00 00 80 00 BE 06
lens folding up...	lens folding up...	lens folding up...
716,8ms	705,43ms	725,16ms		336,96ms	352,76ms	344,85ms	346,89ms
SDO: 42 FF 07 00 C6 16 02 00	SDO: 62 FD 03 00 C6 16 02 00	SDO: 42 FF 03 00 C6 16 02 00		SDO: E2 FF 07 00 60 08 00 00	SDO: 62 FF 03 00 C2 12 00 00	SDO: 42 FE 07 00 C0 10 00 00	SDO: E0 FF 07 00 C0 10 00 00
SDI: 42 FF 07 00 C6 16 02 00	SDI: 62 FD 03 00 C6 16 02 00	SDI: 42 FF 03 00 C6 16 02 00		SDI: E2 FF 07 00 60 08 00 00	SDI: 62 FF 03 00 C2 12 00 00	SDI: 42 FE 07 00 C0 10 00 00	SDI: E0 FF 07 00 C0 10 00 00
790,6ms	761,32ms	794,78ms		393,21ms	414,12ms	370,9ms	384,84ms
SDO: 7C 31 10	SDO: 7C 31 10	SDO: 7C 31 10		SDO: E0 00 08	SDO: C0 00 0C	SDO: E0 00 18	SDO: E0 00 18
SDI: 7C 31 10	SDI: 7C 31 10	SDI: 7C 31 10		SDI: E0 00 08	SDI: C0 00 0C	SDI: E0 00 18	SDI: E0 00 18
VDD going low	VDD going low	VDD going low		VDD going low	VDD going low	VDD going low	VDD going low

 

I also looked at this on the Blackmagic Pocket Cinema Camera...
not only does the #BMPCC not send the proper lens shutdown sequence, it also has some really bad noise on the VDD line while being switched off. 

Command: prepare for shutdown (FT)

176 160   0   0         prepare for shutdown

 TODO: verify 

Command: focus

176 161 xxx xxx  change focus to/by value
176 173   0   0  start single autofocus

 TODO: verify 

 Other commands

192 241   0   0  poll
192 246   0   0  ?(sent after initial command and before requesting the full lens data))

192 250 xxx xxx  ?
192 251   0   0  ?

  TODO: analyse these

• Previous work
• 30C3 project page  
• Arduino SPI library (Master only) 
• Existing thread about the pinout (photo)
• Photos (coming soon)
• Similar adapter for Canon EF
• Attempt to decode Canon EF (German)
• Magic Lantern lens.h (for comparison to identify similarly coded values) 
• Microcontroller rig to mount Canon lenses to Sony cameras

BMPCC LANC and Sysclk DX (USBee DX clone)

I received my Sysclk DX from China. It's an affordable clone of the USBee DX.
A 2 channel USB Mixed Signal oscilloscope paired with an integrated 16 channel logic analyser.
The first task I'm giving it is to analyse the LANC protocol on the Blackmagic Pocket Cinema Camera.
Why? Because I need to have a way to meassure what's supposed to be on the wire and what is actually there to implement LANC into an Atmel microcontroller and build a better LANC remote. ;)

First try:
On a breadboard I pulled the data line high and tried to analyse the signal. The camera is supposed to send data packets... absolutely no luck. :(

Second try:
I put the breadboard in between a working LANC remote and the camera and ran the logic analyser.
I may be doing something wrong but I can't see anything on the wire. The LANC remote however works fine!

UPDATE: Third try:
I dumped the Sysclk DX and got my old Usbee AX and it worked fine.
I found the following LANC commans on the Blackmagic Pocket:
//LANC(0x18, 0x33); //send REC command to camera
//LANC(0x28, 0x45); //send MANUAL FOCUS FAR command to camera
//LANC(0x28, 0x47); //send MANUAL FOCUS NEAR command to camera
//LANC(0x28, 0x53); //send IRIS+ command to camera
//LANC(0x28, 0x55); //send IRIS- command to camera
LANC(0x28, 0x43); // AUTO FOCUS  
Update: Blackmagic answered: IrisAutoAdjust =  0x28 0xAF
I search in the area of 0x55 and 0x47 but could not find anything that triggers auto iris or auto focus.
Aparently these commands do exist but I don't know the command code.

UPDATE:
Using the code from "Control your camera" it seems to work more reliable when removing the -8ms adjsutment for the digitalWrite command and replacing that with a macro from here:

#define fastWrite(_pin_, _state_) ( _pin_ < 8 ? (_state_ ?  PORTD |= 1 << _pin_ : PORTD &= ~(1 << _pin_ )) : (_state_ ?  PORTB |= 1 << (_pin_ -8) : PORTB &= ~(1 << (_pin_ -8)  )))
// the macro sets or clears the appropriate bit in port D if the pin is less than 8 or port B if between 8 and 13

my send-byte routing now looks like this:
void send_8(char dat) {
  char abit;
  char index;
  for(index = 0; index <= 7; index++) {
  abit = ((dat>>index) & 0x01);
  if(abit==1) {
    fastWrite(cmdPin, HIGH);
  } else {
    fastWrite(cmdPin, LOW);
  }
  delayMicroseconds(bitDuration);
}

//Byte is written now put LANC line back to +5V
fastWrite(cmdPin, LOW);

}

Also: I can successfully power the Arduino Uno from the LANC cable.

• Forum Thread on the Blackmagic website
• Others on the SYSCLK DX
• Open Source Firmware
• Arduino and LANC
• LANC protocol  
• fastwrite macro

Soundproofing the TravelCNC

The TravelCNC already has an inner lining of Basotect acoustic foam on the inside.
(Not all of it mounted yet because I'll improve some of the cabling hidden below the foam.)
This helps a great deal with high frequencies.
For the low frequencies I have a heavy, 10mm thick rubber sheet below the actual machine.

It turns out that this is not enough. I get a great deal of vibration that can actually be felt on the wooden stage box.
Next idea: there are 4 bolts going through the rubber and mounting the machine to the case.
I'mm try to widen the holes in the wood and install sound insulating bolt anchor fittings to insulate the bolt from the hole in the wood.

...let's see of that helps.

2013-11-13 UPDATE
I finally finishd the sound proofing

Fixing the 4th axis on the TravelCNC and breaking my tools

Since I mounted a larger stepper motor to the 4th axis on my TravelCNC (portable hobby sized CNC in a custom flight case), I needed to raise the 4th axis by 10mm.
Else the longer stepper would collide with the frame.

This means that the tailstock is now too low.

So yesterday and today I milled myself a spacer out of some wood to keep the tailstock at the correct height.

It worked...but when I tried to use it, I made a terrible mistake.
I wanted to use the edge finder to find the right and left edge of the cylinder I mounted.
I knew it was perfectly round, so the center of rotion would match the center of the cylinder.
....so I mounted my mechanical edge finder...switch on the spindle to rotate it slowly

...and forgot that the spindle was still set to 10'000RPM.
It ripped my trusted edge finder apart.

Now you can get mechanical edge finders for 3-4eur in China, no problem.
You can even get cool electronic ones for 10eur.
However all the mechanical ones are 10mm shank and the electronic ones 20mm shank.
The maximum my ER11 collets can mount is 7mm.

I would really like to modify one of these electronic ones. Attach it to the electronic probe input and find edges extremely accurately without jogging back and forth.
However the back is also the battery cover. So I can't simply mill it down from d=20mm to d=7mm. :/

MY BMPCC review

I just received my Blackmagic Pocket Cinema Camera last night.
In this blog posting I'm trying to collect my findings about it to help others decide on it's maturity.

untested

• 
  
• LANC focus control
• LANC IRIS control (command codes unknown)

confirmed features

• LANC start+stop
• very fine and organic grain

confirmed, known issues

• very low audio levels with internal or external microphone (Sennheiser MKE400 with levels "+")
• constant white noise in audio
• Focus button is extremely slow (2-4 second for a Leica lens that focuses in less then 1/10s)
• no OIS on 14-42PZ
• extremely short battery life
• seldom dropped frames (3-4 frames on a total of 45min of 1-5 minute recordings) using Patrion Memory EP-Series 64GB UHS-1 SD-card

new issues

• 14-42PZ does not shut down properly on BMPCC (lens doesn't collapse)
• LANC cannot control the zoom on the 14-42PZ, ,14-42 II, 42-175 PZ and 100-300 using the Manftotto 484LNC (my fault. DAQ states that ZOOM via LANC is unsupported. :( )

feature requests

• manual aperture control in the menu
• you can do that. You have to press IRIS and the up+down buttons right below. Imposible with one hand. Difficult with 2 hands.
• display of remaining minutes on SD-card
• audio levels
• deletion of clips in camera
• formating SD-cards in camera

Links 

• technical discussion about the LANC commands the BMPCC supports
• LANC using an Arduino

Tailstock height adapter for my TravelCNC

The original 4th axis on my TravelCNC was mounted so low, you can't even extend the arms of the chucks beyond the diameter of the chuck itself without hitting the table.
It was also so weak, that you could rotate it with 2 fingers while it tried to rotate the other way with all it's force.

So now I have a NEMA23 longneck stepper with tons of holding torque...and I can't mount it because it collides with the table.
Luckily it already came with a 10mm high alumimum plate to be mounted on the T-Slots.
What I did was to mill a small spacer out of wood to get the tailstock to the same Z+10mm height because obviously both have to match.

As it turned out, I had the wrong tool diameter in my CAM and it was 2mm too short.
..but that's easily fixed with hand tools while cutting it out.
(I don't want to cut it out with the CNC because I have no scrap-wood below it to ensure a flat surface.)

The perfect multitool

I think I found the perfect multitool for me.

What I found I need are:

• LED lamp
• pliers
• phillips #1 screwdriver
• flat screwdriver
• knife
• sometimes a bottle opener (although Germans open bottles with just about everything...except bottle openers.)
• wire strippers would be handy but not needed.

I like full sized multitools but they tend to stick out in an office environment, so I can't wear them all day (wich is kind of a point of having such a tool on your belt).
Instead I carry 3 tools:

The first thing I carry on my belt is a foldable and very tiny xacto-knife. Extremely tiny, extremely sharp and with a number of replacement blades inside. I think I'll keep that.
It's similar to this one except in color and name and that it comes with a tiny belt holster:

I had to ditch a small chinese tool at the airport, having forgotten to put it into the checked in luggage
(Not because of the knife...but because it had a screwdriver. Aparently airlines are afraid Germany Engineers can do just about anything given a screwdriver and duct tape.).
In San Francisco I got myself a Portland Coast  C2899.

I also carry a small 1xAAA flashlight but can't find a belt holster tiny enough for it.
The Coast was supposed to replace it because it already HAD a flashlight.
...only that it was way too weak, required a size of batteries you don't easily get in continental Europe and constantly empty because the button was accidentally pushed by just about anything in your pocket.
...and it didn't come with a belt holster.

Now I found something else.
A chinese dealer named Yuki Guan sells a RUISS RS1016A.

It doesn't show but he answered that it's also avaliable in black and it comes with a very sturdy belt holster.
It has the same, nice grip that the Portland Coast one has but the LED lamp is bigger and cannot be activated accidentally anymore.

custom XL45 pulley for 8mm D-shaped shaft

My CNC milling machine has a weak 4th axis.
I want to attach a larger stepper motor with seriously mor torque.
However that motor has a D-shaped 8mm shaft while an XL45T10 pulley always has a 6mm shaft.
I ordered a conventional one, hoping to drill it up but it was stuck in transit (damn Hermes. An envelope would have worked better.).
So I had to design a custom pulley that could me machined without a rotary axis (less of a top guard) and fits snugly onto an 8mm shaft.

In fact, I'm 3d printing a first one and plan to do the final one with a full sized top guard on the, thus fixed, 4th axis.

• Design files on Thingiverse
• My free 4+5 axis CAM software