The post processor will need to be edited as well. Open it in a text editor, I suggest Microsoft visual studio code (free). Find this line:

Comment that line out by putting "//" in front of it (so you can keep it for reference) like this:
// var lengthOffset = 1; // optional, use tool.lengthOffset instead
​and paste this line below it:

In theory this should be all you have to do, but this causes the 808D to alarm when you call a NEW tool number, and the D number is not 1. In order to make this work I found you need to change the actual tool change code to send the D number after the M6 call. Find this code in the post processor, and comment it out as shown using the "/*" above and the "*/" below:

Then paste this code below the code above:

UPDATE: You also need to change this in another area of the post, that handles if you change WCS in the program (like two parts in different vices and each has their own WCS (G54.. G55 etc.). This was found in the "// WCS" area of the post, about line 1328.

It is a good idea to leave the old code just in case. The /* and */ make what is in between not process (commented out). Change it to this:

'Now in the control you can add additional "Edges" in the tool offset table by hovering over the desired tool number then pressing the "edges" key, then "new edge" and enter in the length offset, and diameter (optional). I am using 3D printed tool tags (print in PETG so they aren't too brittle) to keep control of what tools are loaded in the machine:

I just switched my machine over to the latest head version so I decided to document it as a bit of an assembly guide. Hope you find it helpful!

STEP 1:
Install the tensioner blocks on the main plate using 4 M3x35 screws, and 8 M3 nuts as seen below. The tensioner blocks fit a little snug so you may need to press them in. I agree that the screws are a little long....

STEP 2:
​Install the main head plate on the X axis bearing block using 4 M3x10 screws. This is the best time to align the head. Since I used a dial indicator attached to the X bearing block to level the bed, I simply used a machinist square to level the head front to back. I used shim stock between the X axis bearing block and the head plate. To align the head left to right, simply loosen the mounting screws to rotate the assembly around the Y axis. Since we machine a reference edge for the Z rails to square up against, you only need to perform this on one side. Take your time to align the head properly.

STEP 3:
Install the NEMA 17 Z stepper with 4 M3x10 screws. Align the motor so the connector is easily accessible (unlike in these pictures). Tighten the screws in an alternating pattern to ensure the motor sits flush on its small round boss (this will make sense when you go to install the motor). Install the spur gear.

STEP 4:
Connect belts (if you are using belts). You will need the supplied two little delrin pins to wrap the belt around.

STEP 5:
Install the proximity sensor. Only use one lock washer, on the back side of the head plate. Leave the sensor retracted as far as possible until the head is fully assembled.

STEP 6:
Install your ELP camera in the camera mount (ELP camera is not supplied, nor is the M2x10 mounting hardware). If you purchased or printed our 3D printed light panel mounts, now is the time to mount them as well using 2 M3 screws. Install the camera mount to the main head plate using 2 M3x14 screws and nuts.

STEP 7:
Install the top plate using 2 M3x14 screws and nuts.

STEP 8:
Install the left and right carriages using 8 M3x8 screws. Note the machined alignment marks that line up with the spur gear. Important note, run the carriage up and down to adjust the mesh on the gear. Find a good balance between low friction and excessive play. YOU WILL HAVE A SMALL AMOUNT OF PLAY BETWEEN THE GEAR AND RACK (WHICH RESULTS IN PLAY IN THE Z AXIS). THIS IS PERFECTLY NORMAL AND WILL NOT AFFECT PRECISION OR ACCURACY, ESPECIALLY WHEN USING SPRING LOADED JUKI NOZZLES.

STEP 9:
​Admire engraving. 

STEP 10:
Install the NEMA 8 motor mounts using 4 M3x14 screws and nuts. Align with carriages.

STEP 11:
Install NEMA 8 motors using 8 M2x8 screws. Do not over tighten and strip the heads. [If you purchased or printed our vacuum adapters, install them now. Remove the 8 phillips head screws that hold the motors together. Optionally, apply a thin film of sealant on the face of the adapter. Use the supplied 8 M2x35 screws to install. Do not apply excessive sealant as it may bind the motor shaft.]

STEP 12:
Adjust proximity sensor. Try to get the gap to 1mm or below.

STEP 13:
​Wire and plumb.

STEP 14:
To use our head you need the -edge branch of the smoothie firmware. [Please buy genuine smoothie boards (like the 5x), I didn't, I'm going to correct that mistake shortly]. This enables the unique homing method. Download the firmware.bin file from our product page here.
​Copy and paste the .bin file onto the smoothie SD card:

Reset smoothie. After reset the .bin file should have disappeared and a config and firmware file (type cursor) file remains:

STEP 15:
​Copy and paste our example config from the product page here. Paste the example config file into the Smoothie SD card. It will ask you if you want to replace it, select yes. The config file is where you set a bunch of parameters, like steps per mm, homing, etc. You will definitely need to play inside this file, I suggest using a code editor like notepad++.

STEP 16:
​Inside OpenPnP, click the "Machine Setup" tab, then expand "Driver", then click on the "GcodeDriver" heading. In the Setting drop down box, select "Home Command":

Paste the following code in the box and click apply:

Home the machine. You will notice that the head is not likely homed level:

Use the machine controls to jog the head in Z, until the two carriages are level. I use a ruler across the top edges. Note the OpenPnP DRO reading (in this example, I jogged Z +4.8mm):

Now modify the homing Gcode to move the Z axis the amount you measured with the DRO to balance the head, like this (then hit apply):

Press the home button again, you should see that the head is now level and the Z DRO reads 0.00.

STEP 17:
Revel in the fact that you just built a pick and place machine.

​
THANK YOU FOR YOUR CONTINUED SUPPORT. SALES OF THESE TYPES OF PRODUCTS ALLOWS ME TO CONTINUE APPLYING TIME AND RESOURCES TO GIVE BACK TO THE AMAZING OpenPnP COMMUNITY!!! 

Update! 2016-04-28: The head is built and everything went smoothly. The pick and place isn't running yet as I had to order a new smoothie board to run it. Hopefully it will be testing in the next 2 weeks.

We are making a pick and place machine to better serve our customers. More information will be available here but for now this is more of a repository for people looking to build the head we designed. Stay tuned for more!

View the most recent drawing online here: http://a360.co/1pcXVkM

The drawing below assumes 9mm thick parts. If you use other thicknesses you will have to adjust the hardware lengths:

Use the file below to Laser cut the entire head, so you get nice clean gears. File works for ponoko, 9mm acrylic.

A360 3D browser and CAD model download: http://a360.co/1QCDeLH

More info on the Machine Design (Courtesy of Anthony Webb) and the head development here:

https://groups.google.com/forum/#!topic/openpnp/-XGpRTdaTrk[1-25]

https://groups.google.com/forum/#!topic/openpnp/w46THEk0Mjc[1-25]

We are pleased to announce that BETZtechnik has been selected to be part of FirePick DELTA - Beta test group 1.

This machine will be able to greatly improve our in-house capabilities and lessen our reliance on some of our technology partners (For pick and place machine assembly (BGA- QFN designs etc.)). 

Click on the image above for more information about what the FirePick DELTA is, and why you should be excited!

Here is the code, but we will dissect it further:



/*
This code is written to test the BETZtechnik.ca RFduino port expansion shield.
All three solder jumpers for address selection must be soldered so that the adress is compatable.
Note that the board was supposed to create 8 additional I/0 pins, but do to some design errors the
board in its current level supports 5 additional I/O's. This will be corrected in the next revision.
 This example code is in the public domain.
 */
 
 #include <Wire.h>

void setup() {

Wire.beginOnPins(2, 3);

 Wire.beginTransmission(B0100111);      // this is the adress of the chip
  Wire.write(0x00);                                     // this is the mode for input/ output
  Wire.write(0x00);                                    // this sets all ports to output
  Wire.endTransmission();

}

void loop() {

     Wire.beginTransmission(B0100111); // address of chip
  Wire.write(0x09);                                    // this is high/ low mode
  Wire.write(0xff);                                    // this sets all pins to high (3.3v)
  Wire.endTransmission();

  delay(2000);                                          // 2 second delay

       Wire.beginTransmission(B0100111);
  Wire.write(0x09);                                 // high/ low mode
  Wire.write(0x00);                                // set all ports low (0v)
  Wire.endTransmission();

  delay(2000);                                         //2 second delay

}

First thing first. i2c is handled by the "Wire" library, so be sure to include it in your sketch like you see at the start. RFduino expects you will be using it on other pins so we must define the pins that match the board itself, so instead of "Wire.begin()", we will use "Wire.beginOnPins(2, 3)".

The first important step to communicate with the chip is to determine its address. You should close the three solder jumpers on the board to pull the chips A0, A1, and A2 high. Lets look at the data sheet:  http://www.mouser.com/ds/2/268/21919e-53245.pdf

Here it shows the address selection. You can see the A0, A1, and A2 in the slave address (the chip is the slave). Since we closed the solder jumper, these are all now high. So the address is B0100111. If they were all low (which our board cannot reliably do at the moment since there are no pull down resistors) it would be B0100000.


You can see in our code where we have indicated the address of B0100111.


Now we need to look at setting the ports to input or output:

This table shows how you select as input (1) or output (0). This is all in binary, so if we wanted them all to be inputs, it would be 1111 1111 (this is represented in HEX for the code, so "0xff"), if they were all outputs it would be 0000 0000 (in HEX it is "0x00"). if you wanted only port IO0 to be input and the rest output it would be 0000 0001 (in HEX "0x01"). Each 1 or 0 represents each port, left to right, IO7 to IO0.

Notice the "register" for this mode is (in bold, above the table) "0x00".

So the code above directs the chip to set all ports to output (in setup if they are not going to change again) in this block:

 Wire.beginTransmission(B0100111);      // this is the adress of the chip
  Wire.write(0x00);                                     // this is the mode for input/ output
  Wire.write(0x00);                                    // this sets all ports to output
  Wire.endTransmission();


Now we will look at setting the ports high or low:

Here we see that the mode "register" is "0x09"

1 is high and 0 is low, so for all ports high it would be 1111 1111 (HEX is 0xff). For all ports low except port 6, it would be 0100 0000 (in HEX it would be 0x40).

In our code we set all high by sending this string of commands to the chip (in the loop because we want it to switch back and forth):

   Wire.beginTransmission(B0100111); // address of chip
  Wire.write(0x09);                                    // this is high/ low mode
  Wire.write(0xff);                                    // this sets all pins to high (3.3v)
  Wire.endTransmission();

Then we set them all low using this:

   Wire.beginTransmission(B0100111); // address of chip
  Wire.write(0x09);                                    // this is high/ low mode
  Wire.write(0x00);                                    // this sets all pins to low (0v)
  Wire.endTransmission();


As you can see there are many, many more functions available in the chip such as interrupts and read capability, but I hope this proves to be a useful tool to help you dive in to this port expander!!

I decided to go ahead and install limit switches so I could do the TTS repeatable tool holders, probe and MSS (Mach Standard Mill) implementations at once. I am very sick of touching off after every tool change!!

It was Taking me forever to come up with a design that I could machine on the mill, and then it hit me, 3D print them!

I have make simple mounts for the TEMCO limit switches found here.





X axis mount:

This was the easiest because the adjustable stops were already on the cross slide. I also extended the bottom so that it would protect the Y axis lead screw that gets exposed.

Then the Y axis mount and rail for adjustable stops:


(Pics to follow!)

Z axis mounts and stop rail:

Requirements:

- BLE112 or BLE113 system with required external SPI flash memory

- BLED112 or our BLE112 breakout board to use a as BLE dongle on the updating computer

- The NEWEST SDK from Bluegiga. This also installs all required drivers on your computer. Also the required BLEGUI program is inside the SDK.

Procedure:

1. Ensure your target device is programmed with Bluegigas OTA example firmware contained in the SDK. For advanced users: Ensure your custom firmware contains the required code found in the APP NOTE or copied out of the OTA firmware example.

2. Power the target device and connect the BLED or BLE112 USB enabled breakout board to your computer.

3. Open Bluegigas BLEGUI program found in the bin folder of the SDK.

4. Bind the BLED112 or BLE112 breakout board to the GUI by selecting the required COM port and pushing the connect button (top right).

5. Enable active scanning and start a scan for devices.

6. Connect to the target device (which should appear in the search results). Click the GATT button. Then expand the services and descriptors. Notice the numbers beside the two 3rd party descriptors. These are the attributes required for the OTA process. On my boards it is #25 and #27.

7. Unselect the COMM and TEXT check boxes at the bottom of the page to speed things up.

8. Go to COMMANDS > DFU, then select the OTA tab. Select your target firmware and then enter the control and data points, again on mine it is control- 25 data- 27. Hit the update button.

9. When complete you will get an "OTA update process complete".

That's it, hope this helps some of you!

This is my custom board incorporating an ATMEGA 328p MCU, lipo battery charger with USB connector for charger powering, 8MHz crystal oscillator, lipo battery powered, SPI programming header and header for a 1.8" SPI TFT display.

The device is for my "Bubba Keg", a creation made for a few friends. Essentially it is a control system for the 70oz Bubba Keg insulated mugs. It uses an ultrasonic sensor in the lid to determine quantity and a PTC thermistor in the cup to determine content temperature. For the old design I used a custom PCB that joined an Arduino Nano and 1.8" TFT display. It ran off a 9v battery and was housed in a generic enclosure. it was far more bulky than I wanted it to be. That is why I went for a SoC solution.

The display has gauges for quantity, temperature and voltage. The display also shows you calculated L/Hr consumption rate and time to empty.

Programming the ATMEGA without USB support was the challenge, but I found some tutorials on using another arduino (UNO) as an ISP (In Service Programmer). I wired the two together and used the Arduino IDE to program the chip. I was very pleased to see it working! Not having the bootloader on the chip means it starts up lightning fast and has more room for programming space. I chose 8MHz to conserve power as it was more important than performance.

Next step will be to custom design and 3D print a nice compact enclosure for it and test fit it on a mug!

Seems I wasn't immune to having the plastic high/ low speed gear (p/n P0704212) strip on me. I crashed the mill in low gear cutting steel. It was hard enough to rotate the head about 20 degrees.

Grizzly was good enough about supplying me one under warranty and I purchased an extra one for good measure.

I believe I know a contributing factor to the gears stripping at least slightly before they should. I initially chalked it up to poor rigging as the low gear did not engage 100% of its tooth width. The more I looked at it however I realized there is no possible way to adjust the gear mesh quality and it does not engage 100% in low or high gear.

A possible solution would be to re manufacture the arm of the high/ low speed fork arm. If you were to make the fork arm p/n P0704217 a little longer than it currently is it would allow you to have more throw on either end of the travel, allowing the gears to mesh completely.

Another option would be to make another detent plate (item 273 on the left) with the detents located further apart. I am not sure if something else limits the throw of the knob but I do not remember being able to rotate it past the detents.

Since I already have the CNC mill (G0704) in the shop, the next logical choice was a 3D printer. I wanted something relatively affordable yet capable. I went with a MendelMax 1.5 Deluxe kit from 3D Factory LLC.

My free time is limited but I felt the value of having built it from scratch was worth the extra time vs. buying assembled.

I am already well versed in Arduino and CNC in general so I feel I am at an advantage to most, time will tell!

Box one of two arrived, the other was delayed through customs. That is ok because it was the mechanical assemblies, which need to be completed first.

I had realistic expectations about the quality of the printed parts, these in particular were printed buy a MendelMax 1.5 so I was eager to asses them. The quality was greater than I expected, the Y axis carriage had failed during shipping but I should be able to repair it with some ABS cement.

I did a quick inventory and was surprised to have more than I needed, unlike most kits you buy. Everything was there (minus box 2) so I dug in immediately.

This was a bit of a challenge because I did not yet have the build docs from the seller. I did some searching around on the internet and found a few sources that helped. This printer has a larger than normal build area (X 9.75" Y 11.00" Z 6.00") so I was not able to determine the correct position for some of the extrusions, this led me to first assemble the frame with the vertices in the Y position then have to disassemble to correct. In all this was fairly simple, everything lined up well and was square. I went with black extrusions, black printed parts and black hardware.

Bottom line, it looks good!

To get to this stage was about 3.5 hours. Doing it over again (and now having the build docs I could do it much quicker.