Messages

871

General Mach Discussion / Re: video problem

« on: February 29, 2008, 03:44:24 AM »

Hood - I would appreciate that - jimpinder@tiscali.co.uk.

I will check for service pack 1 - I thought it was installed - I am sure SP2 is in, but I will check it all again.

Jim

872

General Mach Discussion / Re: Help on setting up ToolOffsets

« on: February 29, 2008, 03:34:59 AM »

Putting G43 in the start up line is fine - but remember - if you use an absolute co-ordinate for some reason - that is a MACHINE CO-ORDINATE as opposed to a program co-ordinate, then you need to a) change to the absolute co-ordinate first (G90), then change back to incremental distance mode (G91).

Tool offsets are all incremental from the figures you enter in the tool table. So Tool 1 can have offsets as well.
If you want to use this as a base then leave all it's offsets at 0. Set the tool DRO to 1, then set up your machine in relation to this tool as you want to start the work. This might be at the bottom left hand of the work, with the tool tip just touching the work.

All other tools should then move to their new offsets as you change them.

Most of my work is with steel - so I usually put a piece of wood in first. If it chews that up, I am not bothered !!!

873

General Mach Discussion / Re: inconstant speeds?

« on: February 29, 2008, 03:21:04 AM »

As Graham siad - Is this a curve - i.e. a G2 or a G3 generated by the G Code, or is it in fact very many straight lines just joined together to look like a curve.

The first is easy to produce since the two axis accelerate/decelerate at a predertermined rate together and the whole move can be computed at one time.

The second one is different, and whether you like it or not, the machine has to stop between each move, pick up the new information for the next straight line, then accelerate away again. Since the angles on the curve are different, the rates of movement of each axis, in relation to each other are different for every move. If some of your lines are millimeters long, this takes time. The CV mode cheats in that it picks up the next move and as axis slow to a stop from the previous move, they accelerates again up to speed for the next one - the overlap can be set.

I don't know what is writing your G Code - but it may be quicker if, instead of straight lines, you did away with say half of them, and used G2 or G3 to join up the curves - that, however is quick complicated.

874

General Mach Discussion / Re: homing limits problems

« on: February 28, 2008, 12:26:23 PM »

Your maths seemed alright when you were calculating the pulses per unit. I assume your unit is millimeters. You are saying your motors are 200 steps per rev and the drivers are set to 1/2 steps. Therefore your motor will need 400 pulses to turn one rev and one rev of the motor turns the shatf forward by 1 mm - so you are correct - 400 is your answer - but you seem to be saying your axis is moving 8 mm instead of 5 mm when you put in GCode.

There can be many reasons for that, because Mach3 can take into account all sorts of offsets, so move the axis from 0 to 10 first, then move from 10 to 20 - measure the second movement - which should be 10 mms. This should be correct. Do not come back, or you will have to take into account backlash - which is another problem.

With your switches, you must understand the difference. LIMIT switches are what they say. They limit the movement of the axis, and, as such, however they are activated, should stop the axis dead. There are various settings that then allow you to back off the switch to get going again. (There are excellent videos to watch which explain all this)

HOME switches are different. As Hood said you can have the same switch act as a LIMIT switch and also as a HOME switch - the computer switches off the "LIMIT" action if it is using it as a home switch. The way a "HOME" switch works is that the computer will run the axis up to the home switch. The switch then switches and stops the axis. The computer then reverses direction and moves the axis off the switch untill it opens (or closes) again, then stops. The computer can, then, set the Digital Read Out's for the axis at 0 if you wish (or at any other figure). The computer then knows exactly where the machine is, and can compute all it's next moves from there.

You will find fitting and setting up Limit and Home Switches quite tedious. Home switches must be accurate. The way to test these is to home the axis, then knock off the auto zero facility and try homing the switch again a few times. The answer should always be 0 of course, but it is a test of your machine to see if that is actually the case. A few thous might be a more realistic answer.

I do not bother with limits or homes on my lathe - which I set up for each job - and I am staisfied then that it is accurate.
My lathe also has a milling head, and whilst I do not do too much milling, I prefer to set each job off from a know position, usuallly the bottom left hand corner of the job, with Z resting on the work.

Since I am not trying to run off thousands of parts, this satisfys me, but each to his own!!

875

General Mach Discussion / Re: Help on setting up ToolOffsets

« on: February 28, 2008, 11:40:42 AM »

You are getting into quite a disciplined part of Mach 3. You mention the Z axis so therefore I assume you are talking about a milling machine application.

When you say you now have a quick change, this means that each tool will fit in the holder at the same position (or length).

The first thing you must do is decide on a start position for your tools. This means a position where Tool 0 is at, and normally axis Z is at 0 with this tool. The tool can be imaginary - in other words, there might not be a tool - just the tool holder. and this is zeroed by touching it on the work piece.

How ever you do it, all other tools in the tool table are set from this. I personally have an empty tool holder as my 0 position, and all tools in the holder then have a length protruding from the holder - and all those lengths are positive.

Remember on your GCode that G43 must be activated to take into account the tool length offset - if you are plunging, it could be that you are missing the appropriate G43 - if in doubt - put it in again, it can't do any harm.

You can enter all your tools into the tool table. Even ones you are not using on a particular job. You might just have a few, but the big boys might have a hundred all in a big drum - so they can be entered - information about them is stored. You, of course, must remember which tool is which - and I would suggest, get a tool holder for each tool and label them.

If you are milling, then tools also have a width, or diameter, and this can be entered.

That is it really - but your scripts must all start the same - cancelling all offsets etc, then building them back in as you need them, other wise the machine looses track of where it should be.

Tool changes are M6 T** where the star is the new tool number. It is best to move to a tool change co-ordinate to change your tool. This should be an absolute co-ordinate. You change your tool, the apply G43 again and then return to your position. Because you have left your position, gone to an absolute co-ordinate, then returned ( with a g43) the machine will take the NEW tool co-ordinates into account as it returns, and be in the correct position to start again. ( See the video tutors about Scripting and moving to a position for a tool change)

I think you should get the idea from that diatribe.

876

General Mach Discussion / Re: How to test a parallel port.

« on: February 28, 2008, 11:01:38 AM »

You can test both your parallel ports quite simply, with a volt meter. You will  have to make some sort of lead out to a piece of printed circuit board (rather like a DIY breakout board). I used an old printer cable and a piece of printed circuit card. I identified all the pins on the 25 pin plug, and soldered them in order onto the card.

Testing the port after that, is childs play.
The printer port - LPT1 has the following sets of pins.  OUT  pin2 to pin9.  OUT pin 1,14,16,17 and IN pin 10 - 13, 15. Pins 18 to 25 are signal return (computer gnd) wires.

All the output pins can be tested by allocating each pin to output 1 on Config/Port and Pins/Output signals.  If you then go to Spindle set up and allocate M3 to Output 1 and untick disable spindle relays,  if you click "spindle" for example on Mach3 Turn, the output on the slected pin should change from 0v to 5v, or 5v to 0v depending how you have set "active low". You can use this, or type M3 into the GCode input - which will turn the output on, or M5 which will turn it off.  Since you can allocate any pin to Output1 you can test all the outputs.

Similarly with the inputs - allocate them to a "limit switch or home switch". If as has been said, you then short the input to ground, you should see the appropriate LED change.

You can test you second port in a similar way, but I assume you can allocate the pins to be inputs or outputs, so it is up to you how you test it

877

General Mach Discussion / Re: video problem

« on: February 28, 2008, 10:39:25 AM »

I must be the odd one out. The PCMCIA to USB2.0 card came and was installed on the laptop. I plugged the camera in, but got nothing, just a black screen on "My Computer" Funnily enough I plugged it in the old USB 1 port(s) and got a picture on "My Computer".

Mach3, however, still failed on USB1 port, and now fails on the USB2 port (which is not suprising).

I found an upgrade on Microsoft for the PCMCIA card to USB, but that failed as well, saying that it could not find some information it needed.

I will keep trying when I get some more time.

878

General Mach Discussion / Re: video problem

« on: February 26, 2008, 02:12:26 PM »

Rodney - your reasoning seems very logical - both my computers have USB1.

I have ordered a PCMCIA to USD2.0 card for the laptop (my lathe computer). The chap says it will come tommorrow. I will report on it

879

General Mach Discussion / Re: Brians new addition

« on: February 26, 2008, 02:07:10 PM »

Yes - he certainly looks to have all six axis attached at all the right points.

BUT - be warned Brian - they only get worse as they get older !??

880

General Mach Discussion / Re: 5 phase stepper and Mach 3

« on: February 26, 2008, 02:03:21 PM »

You seem to be getting things in a bit of a twist. The "steps" are nothing to do with Mach 3, they all depend on your motor and driver, If you have the motor and the driver, then you can find this out yourself by connecting up the motor and driver, then putting a number of pulses into the "step" input of the driver card, and seeing how far the motor moves. The advantage with a stepper motor is that it will move for one step. So you can mark the output shaft, then put pulses into the motor. The shaft will move round step by step until it comes back to where you started. You could even use a push button to put the pulses in.

What Mach 3 does need to know is how many "pulses" therefore the number of "steps" Mach 3 has to put out to drive your axis either 1mm or 1 inch, depending on which measurement your wish to set up in.

I have not heard of a 5 phase stepper, most are four, or two  - but for example - my drivers are set to 1/8 step, my motors need 200 steps per rev, I have a geared down drive to my axis of 3 to 1, and the axis has to turn 10 time to move one inch. Therefore I need 8 x 200 x 3 x 10 = 48,000 pulses to move one inch. As you can see then, my best resolution is about 1/5th of a thousanth of an inch.

If this a a bit difficult to work out, there is a setting system built into Mach 3 to set the correct number of "pulses per unit" when you are setting up each drive (all drives need not be all set up in the same way).