need help in pcb drilling machine!!

[crys]

hai, i need help! we have this pcb drilling machine idea for our thesis requirement and we are in need of help.

the basics of our idea are:

1. we use polar movement instead of the usual x and y
2. we want it to be computer interfaced
3. the size of the pcb is limited to 1/2 short bondpaper

for the program, well, we're stuck because were electronics and computer engineering students, and we haven't gone through hard core programming.. my bf, one of the member, said he'd study short course java, but i doubt if he can finish in time. and its time consuming.. T_T so we have little clue as to build our project. we also dont have any ideas what PIC we should use.. our ideas, thanx to Engr Rafinan, came from the website below by elektor electronics:

**broken link removed**

 

[3v0]

You want to reverse engineer one of these.

**broken link removed**

How long do you have to do the project?
This project is about 25% EE, 25% ME and 50% programming.

You need someone on your project who can program in C or has the time and ability to learn it. On the other hand if you build a device compatiable with the one shown above , I think the software can be downloaded if you look for it in the internet achives. The company that was selling the kits (or whatever) closed their doors.

 

[crys]

Re:

well, we only have 2 months for our proposal(first semester) and another 4 months for the actual implementation of the proposal(second semester). if we can't make a proposal now, our whole project will be ruined..

and we are all Electrical and Communication based so programming is really difficult for us.. i searched the internet though and found this really cool site that has a step by step instructions but then, its not our project. i also found a software used called Mach3.. im still searching for it right now. hope its a freeware..

 

[crys]

and also, im concerned.. mach3 have x and y coordinates.. is there any other same program with mach 3 that uses polar coordinates?

 

[3v0]

I have to question if you are aware how complex the software for this project is.

Most CNC milling machines use X Y Z axis. To drill a hole at xy you move Xaxis to x and Yaxis to y. With this machine you have to find rotate the work table and the arm to get the desire xy under the bit. You will not be able to use any off the shelf CNC software, at least not without modification.

Do not take my word for it try asking you question regarding this at
CNCzone.com-Machinist Community Forums
They are all about CNC and can better answer your question.

Good luck

 

[crys]

waa.. okay.. thank you very much! ^_^ i'll go there right now!

 

[yngndrw]

Mach3 has a trail version, it is limited to so many lines of G Code but it should work fine for your use.

However, Mach3 is only for the Cartesian coordinate system. Why do you want to use the Polar coordinate system, can you not use Cartesian instead ? That would make the programming side of the project a lot easier.

I'm actually currently in the process of building a PCB milling / drilling machine, it uses the Cartesian system and Mach3.

 

[crys]

someone from the past class already made a pcb drilling machine with cartesian coordinate system. so we opted for polar, and since its more easy in a sense for the drill to move about.

and yeah, i cant use mach3 because the objective of our thesis would be to make a program that would run a pcb drilling machine with polar coordinate system.

 

[Boncuk]

PCP drilling machine

Hi crys,

the machine shown in your picture is the "Tanbo" (Tangentialbohrmaschine) developed by Radix GmbH in Bayreuth/Bavaria (Germany)

The kit includes drilling software written in GFA-Basic and usable for Windows OS.

Radix GmbH stopped marketing because improvement was too expensive, e.g. using fast servo motors instead of hard kicking solenoids for the Z-axis movement, and sales weren't much. Marketing then was given to "Bear-GmbH", but they failed having no qualified personnel.

Supposedly Radix-GmbH still exists. Try their web page **broken link removed**.

If you don't have any success contacting someone in charge of the machine I'll help getting in touch with the designer. You should try to get in touch with Thomas Müller, who has developed the machine.

I have all plans and firmware as well as the GAL-content of the machine and also redesigned the control board PCB for cleaner signals from the GAL. There were some design mistakes like too weak ground connections which made the whole thing a "perhaps machine".

The mechanical design however is thoroughly engineered allowing high manufacturing speed for drilling with very fast X and Y movements.

If all your efforts are unsuccessful I will provide the formulas used to convert X-X coordinates into the Tanbo-system, which is completely different from a linear system. The table rotates and so do the arms with the drill motors.

Regards

Hans

 

[3v0]

Hans,

Hobby types could use a inexpensive CNC drill. What do you think of building this sort of machine but favoring low cost over performance/speed. How cheap could one make it and still have it work ?

 

[Boncuk]

Hi 3v0,

the Tanbo kit is indeed comparatively cheap. €990for the basic version with one drill arm (can be fitted for a total of three). Compared to the BUNGARD CNC drilling machine in the basic version (manual tool change) it was really cheap. For the BUNGARD you must calculate 12.000€

I was thinking about the design of a hobby machine as well. Getting precision spindles for X-Y and Z movement is no problem in Thailand. You can purchase any length and diameter second hand spindles at Chinatown/Bangkok for a few hundred Thai Baht (1 €=56 THBT). Also all kinds of stepper motors are offered cheap and it shouldn't be a problem to make an own CNC-drilling machine.

There is just one small problem arising when it comes to mechanical design. You can't get strong alumimum profiles in Thailand.

Thinking about automatic tool change, which would be desirable, lets the overall costs explode. Using a KAVO-drive with pneumatically initiated tool change takes already almost 1.000€. (impressive data though: 3 phase regulated motor ranging from idle to 100.000 rpm, motor controller included in price)

It isn't the problem to do manual tool change, but with every tool change you will risk misalignment in at least one drive axis. (During tool change the drive motors are disenergized and small movements jerking with a wrench won't be recognized by the system.)

Next month I will however take a ride to Bangkok to purchase the basic stuff for such a machine.

There is a big disadvantage to all hobby machines on the market: Their basic accuracy is not better than 0.025mm. Imagining a board with fine trace technology the vias would be drilled "miles" away from their original position.

It is just a matter of sufficient reduction gears to achieve a basic accuracy of 0.005mm (even better than industry standard).

Take a sharp look at a professionally manufactured PCB. No drill is exactly centered within a pad. My adjustable eye ball system works far better.

Regards

Hans

 

[Gayan Soyza]

This is my very basic PCB drill helped me to make millions of holes.

 

[Boomslang]

This is my very basic PCB drill helped me to make millions of holes.

Just add a small drill stand and carbide bits.

 

[Boncuk]

This is my very basic PCB drill helped me to make millions of holes.

... whatever satisfies you.

 

[Boomslang]

Have a look on CNC zone, these guys will help you in the right direction.

 

[3v0]

Hi Hans,
I change tools as little as possible because it is too easy to break them. Most of the time I use a single drill for the entire board.

One could not use a multi arm setup on a XYZ cnc mill. But by the time you add the arm and electronics I am not sure it would be reasonable. It would take a longer X axis to do the same work. The only software change would be to add a different offset depending on the arm in use. Easy if the software has provision to make such an adjustment during tool changes. Or if on had the source. Maybe a better method would be to generate Gcode and then process it to add the offsets.

On of the problems is lash or play in gears and screws. If one always approaches the target from the same direction it is better. Also if one were to provide a slight load in one direction it would help.

 

[Boncuk]

Hi 3v0,

adding an arm only makes sense using the Tanbo. Since the plate rotates each arm is equidistant from the center of the plate. Adding arms to a "linear" drive makes absolutely no sense.

I need at least 3 different drill diameters making a board. Screw terminals have a pin diameter of 1.2mm. Drilling a hole of 1.2mm for an IC pin there wouldn't be much left to solder with a pad size of 1.6002mm. Box headers and jumpers require 1mm drill size and ideally IC pins require 0.6mm (most DIY-PCB makers use 0.8mm instead).

My most simple PCB drill was made by Isert automation about 1970. It had a strong 12VDC motor with very little vibration and perfect alignment of the drill chuck and had an adjustable power supply. Max speed was 30.000rpm, and using carbide hardened drill bits the holes were smooth and shiny inside, good enough for elelectrolytic through contacting. Using a small magnifier lens directly mounted to the drill stand I could drill holes hitting "bulls eye"

The spindles I was talking about are salvaged parts of industrial CNC machines and they require little adjustment only to move a sledge or a tool carrier without any loss of reference no matter which direction they have to move.

Old lathe machines are fitted with cone shaped heavy stabilizing bars, whereas nowadays linear rollers are used travelling along a precision rod and fixing the moving part so it can only go two directions - forward and backward with absolutely no side movement.

I have already calculated for the X and Y speeds using a precision of 0.005mm. It is still 40mm/second which is fast enough for hobby applications, and much more accurate than any industrial machine. (I am perfectionist, pilots have to be that way to survive 6,000+ flying hours in a combat jet)

Years ago there was an article about a selfmade CNC drilling machine which made me smile mildy. The drives were steppers though, but the spindles were of the kind you can buy at home markets for construction work, max length 1m. The principle was simple: Start at the zero/zero point and move to the destination, drill a hole and return home beyond the zero/zero point. Slowly approach that point again, autozero position using optical sensors and then "hustle" to the next coordinate. (Working off coordinates in sequence would have caused offset errors of 0.5mm in each axis, adding to a total error of 0.7mm)

Looks like you are interested in a cooperation developing an affordable CNC drilling machine for our forum members.
Yes, I go along with that. I'll try to get all the good scrap when I travel to Bangkok.

There should be enough of that stuff to make at least one good prototype. If I can get a contractor there repeated supply with the same kind of spindles should not be a problem. (some of the spindles I saw there were still wrapped in oily paper and I assume they were brandnew.)

I'll have a good look for suitable drill motors too. They must be 100% balanced for vibration free spinning. (Depending on the amplitude of vibration a drill planned to be 0.6mm can quickly become 0.7mm being 0.05mm out of balance)

Hans