Monday, December 31, 2007

PICKIT2 Clone

Just received a PICKIT2 equivalent from the courier service few minutes ago. Works on first plug in... uses Microchip standard software (PICKIT2 and MPLAB) without any other software or driver... real value for money for those developing with PIC...
Had a double AA carrying case next to me... nibbed a small rectangle for the mini USB connector and fixed a double-sided tape to the back... voila... an instant casing for the programmer/debugger... protecting it from short circuit as well as handling abuse...

While I try to find time for the millions of hobby I have, I wish a VERY HAPPY AND SAFE NEW YEAR to all...

A very productive weekend...

Managed to accomplish a lot this weekend... Went for grocery shopping, took the kid to the park, had some nasi lemak, and, of course, worked on the hobby...

Finally got the time and peacefulness of mind to make the mount for the high speed water cooled spindle. Milled out of 0.5in 6061 aluminum, it worked like a charm. The riser was from the old proxxon riser, so I only had to make the horizontal piece. I also took the time to replace the water pump to a bigger unit with about 3m max head pressure... no problem so far...

Photo showing the spindle on the sherline... apart from the color, it is a nice match... no problem for the z-axis stepper to move it up and down...
The spindle bottom view. I mounted the magnets as per previous installment... tested and worked fine... may explore closed loop spindle speed control later...
The enclosure and the water cooling tank at the bottom... after about 2 hours of cutting with reasonable load, the water temperature rose about 5 deg celcius... nothing to worry about...
The first piece totally milled using the high speed spindle... from the center hole to the middle and end slots to the drilled holes... What is it... a new riser for the sherline to replace the old ones... so far so happy with the results...

Thursday, December 27, 2007

Quick update on PWM-to-analog for Mach3

I was using the schematic from here to make my PWM-to-10V converter. After changing the PWM frequency from the default 5Hz to something around 200Hz, I was able to somehow control the VFD input. Still got some issues but currently using a workaround.

  1. Converter not giving full 10V at maximum frequency. At full speed, the converter only provides around 8.5V thus limiting my top speed to around 20000rpm. Workaround: Set maximum operating frequency on VFD to around 30000rpm thus at 8.5V, I was achieving 24000rpm.
  2. Frequency control not linear. I notice that the accuracy of the PWM control is around 5%. This is within my acceptable 'error rate' :-)
I believe the problem really lies in Mach3 configuration somehow. Will continue to search for optimum solution.

Also found out that the aquarium water pump I have does not have enough pressure to push the water up the tube (about 2.5 feet vertical) when tube is empty. Need to get a bigger pump soon...

Wednesday, December 26, 2007

VFD in Driver Casing

Had some free time over the holidays so I decided to give a go to locate the VFD into the current driver casing. Initial test shows that the VFD isn't as noisy so there is no false step on the stepper when the VFD is operating. Photos show the results...

This is how my front cover looks like. The VFD is under control of Mach3 (except for the spindle speed. Still awaiting suitable optocoupler for PWM-to-10V conversion).
The internals. The VFD is mounted on the drive bays using a bent 4mm acrylic.
The rear plate. Separate socket for the VFD power input and additional socket for the VFD 3-phase output. VFD connectors on top whereas the stepper connectors at the bottom.
Photo of my breakout board. The two relays to control the VFD forward and reverse switches under Mach3 command.

Next to do is the PWM-to-10V converter. There's a spare parallel port pin available for that function. The first converter I made had two problems: unable to hit 10V at maximum output and fluctuating voltage during intermediate pwm. Probably the optocoupler I have isn't suited for the intended purpose. Back to component sourcing...

Wednesday, December 19, 2007

Final version of the cooling system

Got some free time last night so I managed to tryout the new setup for the cooling system. It's basically the same approach. Main difference is the size of the container, all the interconnect tubes are plastics and the cooling fins resting horizontal on top of the container. The 6-liters cooling water should be able to absorb significant heat should the heatfins fail to dissipate it efficiently.
Wanted to stick the inverter for the high speed spindle into the driver casing... waiting for suitable free time... also need to acquire some connectors for the mains power and the spindle 3-phase wires... later on that...

Monday, December 17, 2007

Been a busy weekend...

Managed to find time to work on the hobby during the busy weekend...

The stepper drivers repacked into a new casing. More space, easier for maintenance and upgrade. A full height PC casing makes modding very easy...

Broke the cover of the old meters but mounted it anyway...

Current contraception for the cooling system to check for flowrate, pressure drop, etc. Basically an aquarium pump passing through a CPU cooler heatfins... Yellowish green came from the car radiator coolant... So far so good, able to take the heat of a stalled 800W spindle... Most likely not the final setup... not happy with a few things...

Wednesday, December 12, 2007

New toy... Water Cooled High Speed Spindle

Picked up this spindle yesterday. Should be able to handle my milling requirement... overkill for the Sherline but can be used later should I decide to replace the Sherline. Need to make the mounting as well as the water circulation system...

Spindle Spec: 800W, 150V max, 400Hz max, 24000rpm
Inverter Spec: 1.5kW, frequency upto 1.5kHz, voltage upto 400V (3-phase)

Thursday, December 6, 2007

Accomodating noisy Sherline spindle...

For the past 1 week, my system was acting funny. Whenever the Sherline spindle is on, my steppers would step by itself. Changing to different spindles (Proxxon/Dremel/Router) doesn't replicate the problem. I presume something is wrong with the Sherline or my electronics somehow became more sensitive to noise.

After several trial to fix the electronics (rearranging the wires, putting a noise filter) and the spindle (checking for worn-out brushes), I gave up. Details here at Yahoo group

After digging further and several internet searches later, I decided to make myself a low-pass filter to be fitted at the input (step/dir/enable) of my stepper driver - PICStep. This site triggered me to do it and I used the values given there (1k0 resistor, 1n0 capacitor, ~150kHz cutoff). It worked like a charm from the first power up. Not a single step on the steppers either when the spindle is running or start/stop switching. Everything back to normal now... Photo of the contraception with schematic shown below.

Monday, December 3, 2007

Engraver infinite hands...

Every engraver know that sometime it is really PITA to get the material to be engraved to be 100% flat. I was helping a friend to set up his Sherline CNC last Saturday and and, as luck set in, I noticed he was repairing a top end engraver unit (probably a Roland machine). The material to be engraved was placed flat on the machine without any clamp whatsoever... Upon my enquiry, I was 'introduced' to the adhesive plate the unit had. It is basically a foam like piece of rubber/plastic that was kinda sticky (think of Post-It note). The best part is that it is reuseable if taken care correctly... Quickly acquire some for myself, comes in sheet about 2mm thick and 200mm x 125mm rectangle.

Shown below, the adhesive foam on an engraving jig for my Sherline. I mounted it on a piece of 4mm acrylic coupled with basic ruler made from normal engraving plate.

The same jig in its' storage box... the small piece on the right is the balance from the original piece... anyone read Japanese?

Wednesday, November 28, 2007

Update on upgraded motor installation

After fixing the motor to Sherline, I notice that the max rapid with bipolar series configuration was around 30ipm only! After a quick rewire to half coil excitation, I was able to get over 50ipm with ease. The result just goes to show that spinning air is very different from actual load... frictions, vibrations, leadscrew efficiency, so on and on... Currently still running at 2A/phase @44V... Will re-set the driver for 3A/phase when I have the opportunity...

Photos below show my current setup and the motor tuning parameters...

Tuesday, November 27, 2007

Upgraded stepper for the Sherline

Took delivery of new set of motors yesterday. It is model KL23H276-30-8B (8-wires, 3A/phase, 280 oz.in), ordered direct from China. Several US online sites also sell identical motors.

I'm yet to fix it to the Sherline. Tried hooking up to the driver (PICStep+Mach3) and result as follows:

  • With 2A/phase @44V, bipolar series, 1/4 step: reached more than 2000rpm (80ipm) before stalling.
  • With 2A/phase @44V, half coil, 1/4 step: reached more than 1250rpm (50ipm) before stalling.
With the above results, I'll be fixing it to the Sherline as bipolar series for high torque at low speed yet yield acceptable (>50ipm) rapids. I'm just too lazy to change the current setting resistor of the PICStep to try half coil with 3A. Probably give better, or equal, rapid speed compared to the bipolar series.

Photo below show the two motors side by side for comparison.

Tuesday, November 20, 2007

Build CNC in a flash...

Found the following video from youtube.com which may be of interest to some of you...

Thursday, November 15, 2007

Spindle Index Addition to Sherline/Mach3

I've been wanting to add a spindle speed sensor to my Sherline/Mach3 setup for a long time. After several failed attempt of using reflective IR optical sensor, I managed to locate and procure a hall effect sensor (from RS Electronics) to settle the matter once and for all. I had two different sensors, proportional output sensor and latch sensor. Looking for easiest way to setup the sensor, I went with the latch sensor approach.

For such setup, I only needed the sensor (Allegro UGN3177UA), a pull-up resistor (10k) and two small magnets. Three wires goes from the sensor to my driver board and PC (+5V, GND, Sensor Output). Since the sensor output is toggle using magnet polarity, the magnets are mounted to the spindle as shown below, one polarity facing up per magnet, giving a consistent 50:50 duty cycle on the output. The pocket for the magnet were milled in the acrylic using the Sherline, of course. Works on the first power up. Mounted to the Sherline using scrap acrylic, double sided tape and cable ties. Now... the DigiSpeed controller looking very tempting... hmmmm....

Thursday, November 1, 2007

Video of my Sherline in action

Uploaded the video below to show the speed of my current Sherline setup. Though cutting air (the artwork was cut earlier... forgot to capture the video) the rapid speed can be seen. Current rapid is 25ipm with Sherline-supplied motor. Driver is bipolar, 1/4-step, 2A per phase, 35V supply (PICStep).

Tuesday, October 30, 2007

Something old but maybe useful...

The following links are things I did in the past which may be of use to others...

How to make a PCB using toner transfer method...
My first try on DIY CNC... works but disassembled now...
My second (or is it the 3rd) try on CNC... also disassembled now...

It was posted there way back when the website just got started... look at the thread numbers...

Monday, October 29, 2007

Tale of the deburring tool and dial indicator holder...

Was at the Machine Tool Exhibition yesterday evening and picked up the three items below...
a) Rotating deburring tool (for metal) - blue handle + 10 spare tips
b) Flat deburring tool (for plastic) - gray handle
c) Dial indicator holder, for use with clamp...

9M4RKK & 9M4RKL Change of Cabinet...

Several stations led by 9M2CJ were at Bukit Lanjan on Saturday 27th October 2007 morning to relocate the repeaters into their new (but old) cabinet. All went as planned. The only surprise was when we found that the UHF cable is full of water... the water was dripping from the cable when the connector was removed... the cable is yet to be replaced... UHF still usable with SWR or about 1.2

Photo showing the black core due to the water leak...
The equipment in the new cabinet...
View from the rear...
And where the actual antenna were located...

Monday, October 22, 2007

PIC-Based Timer with LCD...

Latest update 071022-1615:
The codes have been completed and tested.
For full set of relevant files, please leave comment and I'll email it to you.


At the request of a friend, I quickly design and build the prototype of a PIC-Based Timer module with LCD display. Picture below show the progress thus far...

The parts before assembly.
The PCB was made using the toner transfer method and yet to be cleaned for soldering...
All parts assembled except for the reverse mounted LCD.The component side showing various connection points (Relay, Speaker, Power)
Coded and tested for board functionality. Works on first try... yeaaaaa

As in most project, getting the main function working is quick. Now need to code it for 'nice' user interface to allow selecting/saving preset timing as well as allowing manual duration setting...

The Sherline Cleanup... after months of use...

After months of abuse with actual work, various form of lubricants and grease, aluminum chips, melted plastics and so on, I decided that my mill need the much needed cleaning. I finally find the courage and the time to get it done. The mess it was in... the off-white stuff is lithium grease... yuck...The weapon... degreaser and grease pump with Slick 50 One Grease with PTFE.
The anti-seize lubricant pictured was not used.
Yeaa... after a couple of hours of soaking, scrubing, rinsing and drying... I am very pleased...
The mill ready to be put into abuse... ooopss... service again...

Wednesday, October 10, 2007

The 9M4RKL Morning Crew TT session

Just a brief recording of history... The 9M4RKL so called morning crew had our first informal teh-tarik session last night at Bukit Rahman Putra, Sungai Buloh. It nice to be able to put faces to callsigns and to meet friends again, as always.

Five stations were there including me. The others were 9W2ZSH, 9W2JIP, 9W2ALT and 9W2LYZ. Unfortunately, the macho man 9W2VA is away overseas at the moment. BTW, may need to find suitable nickname for 9W2ZSH and 9W2ALT because everyone else has their coined nickname.

9W2VA - Macho man
9W2JIP - Itu Macam Cerita
9W2BSR - McGyver
9W2LYZ - Sunny

Long due updates...

I haven't been updating this blog for a month already. Been taking it easy on the hobby during the Ramadhan month. Nonetheless, managed to cook up some project during the 'take it easy' period.

Pictured below is the homebrew battery eliminator for Yaesu VX-6R made for 9W2ZSH. It is the 2nd version which is able to deliver continuous 2A@7.4V. The first version burned up due to excessive current load on the switching power supply unit. The clear plate is an acrylic piece to connect to the transceiver which is milled to the dimensions of the battery compartment.

After milling my last callsign plate for the mobile phone holder, it came to mind that I could do much finer detail using normal PCB. Tried the new design using UV and FR4 PCB and result shown below. So far happy with it but need a good method to blacken the PCB for aesthetic, otherwise the contrast is very minimal. Anyone with practical suggestion, do drop me comments... Currently using permanent marker ink but that will rub off when I polish the copper with polishing chemical...

Tuesday, September 18, 2007

Laminated Moxon Antenna

Haven't been updating my blog for some time. Not much activities towards my hobbies lately. Too tied up with other commitments...

Anyhow, managed to try fabricating Moxon Antenna for a friend. Unfortunately I forgotten to scale the dimensions to account for the PVC velocity factor of the laminating sheet resulting in very high SWR... Nonetheless, the friend is still able to smile a bit as the Slim-Jim I fabricated for him is giving him much pleasure in his QSO...

The wrong sized UHF (70cm) Laminated Moxon Antenna...

Monday, September 3, 2007

New 144MHz Foxhunt Transmitter

Hello again all,

At the request of several of my amateur radio friends, I've redesigned the previous foxhunt transmitters to make use of a more easily obtain PIC16F84A instead of the PIC12F675. Photo of completed transmitter below. Setting for FoxID and delays are done via jumpers. Overall transmitter power is about 300mW and frequency of 144.010MHz.

For those building the unit, the label below shows how to set the jumpers correctly for various foxes. For sync start, jumper JP2 could be connected in parallel to a single reset switch which can be removed after reset/sync is done.

Tuesday, August 28, 2007

Photo of previous spindle mount

The spindle mount was cut from MDF board using the same CNC unit... a testament of its' functionality...Front view. Meant to hold Proxxon spindle shown in earlier post...
Rear view showing the aluminum plate for stiffness and mounting to the Z-Axis...

CNC from storage...

Here are the photos of my previous homebrew CNC taken out from storage...
The modified milling table. The ruler is 18" long and placed there for size reference.

The Z-Axis is also about 18" long. Two THK linear slides on the inside and driven by 1/4" threaded rod.

The backside of the Z-Axis to show the motion related parts. All stepper motors are bipolar and coupled to the leadscrew using flexible couplers.

Tuesday, August 21, 2007

Almost sold already...

Will be departing with my dear old CNC contraption very soon... after coming back from my holiday... The photo shows the elements...

Milling table modified and hooked to stepper motors. Total travel after modification is about 6" by 6"

My old setup with the Z-Axis shown. The z-axis bolted to the wall for rigidity... works super well... except for the coupling of vibration to the whole house...

Close up of the Z-axis. Has about 5" of vertical travel shown with the simple dremel mount...

Tuesday, August 14, 2007

Mobile Phone Pouch

I was having a very busy week last week. With 4 days of training session, I couldn't spend as much time as I want on my hobby. Nonetheless, managed to get something done during the weekend.

I have always wanted a means to show my callsign at several ham meetings. Originally I thought of using something similar to a name tag or tie clip. Then I realise that I carry my mobile phone everywhere I go thus a callsign plate on my mobile phone pouch would fit the bill.

The plate material is brass... the lettering were milled to 0.05mm... darkened with permanent markers and the top side was sanded with fine sand paper... Result below...

Thursday, August 2, 2007

Iambic Keyer and Current Setup

Hello again,

As mentioned before, I'm not happy with the iambic pedal design shown below. Too cluttered and too many parts. Scraping this design... and the aluminum as well... Will be designing new pedal... when I got the time...

This is my current stepper driver and PC setup. The PICStep drivers (4-Axis) are housed inside the old PC casing with voltage and current meter mounted at the floppy disk space. The PC is an IBM NetVista, Pentium 4 1.6GHz with 512MB memory and 40GB harddisk. Both EMC2 and Mach3 installed for my testing... as well as DOS apps such as TurboCNC and CNCPro... Still trying to figure out why my Sherline cannot go reliably above 10ipm... BTW, the mini keyboard is 'velcro-ed' to the rack... allowing removal if needed...