Hello Guest it is July 21, 2019, 09:37:21 AM

Author Topic: homing accuracy  (Read 9242 times)

0 Members and 1 Guest are viewing this topic.

Re: homing accuracy
« Reply #20 on: September 10, 2008, 02:26:41 PM »
Hello again,

I did shut everything down on Friday after getting everything set at a point I could make a part. I started everything back up Monday, loaded the G code I used on Friday, loaded a part and hit start. The part did not come out at all like on Friday. After checking things out carefully still no success. I did go to the single step mode to see what is happening line by line. 

I did load the tuning & settings from the Y into the Z. I don't have information on the current limiting resistors at this point. The motor temperature did not seem to change, it just remained barely warm to the touch. I haven't put a thermocouple on it yet.

As far as the line in the drivetest being straight, I am not sure I understand. The stepper is just spinning the drive screw back and forth and line on the screen graphic was straight.

Needless to say I have run more than 50 parts this week and not made a single one even close to acceptable. From the pictures you can see I am taking an aluminum ring .006 thick by .176 wide. The lathe and tooling is then suppose to form flanges on each side of the ring. I have during the past three months been able to make about 400 parts so I have some confidence the system will work.

I have been reading other posts to get some ideas. Thanks.

B.E.N.

Offline Hood

*
  •  25,849 25,849
  • Carnoustie, Scotland
    • View Profile
Re: homing accuracy
« Reply #21 on: September 10, 2008, 04:19:14 PM »
OK the drive test needs Mach shut down, you then look in the Mach3 folder and you will see Drivertest.exe, double click on that and it will do a countdown, take over the system and run the test, after a while it will stop. If the line is smooth you are OK if its full of huge spikes then you have problems.
Sounds however like there are other problems with your system but do the driver test and see what its like.
Hood

Offline jimpinder

*
  •  1,233 1,233
  • Wakefield, West Yorks, UK
    • View Profile
Re: homing accuracy
« Reply #22 on: September 10, 2008, 11:22:10 PM »
BEN

You still do not seem to be getting anywhere.

I think your problem is that you have no idea what the problem is - and this makes it difficult to find a solution.
1 - The problem appears intermittent.
2 - As far as you can see, where is the problem - is it always on one axis, or does this affect both ??
3 - Your set up appears compact so it seems unlikely that any stray signals are getting in there.
4 - From what you describe, the problem is the axis ( or both) suddenly seem to do their own thing.

Try and narrow down the problem - BUT - if you have a spare drive I would change it - if this means just switching the four wires and reconfiguring ports and pins, all well and good.

The problem when you are starting up is different, and Hood has already mentioned this. You must, when starting, take some action to "zero" the machine - so it knows where it is. This does not mean that physically it is in the same position as when you left it, it means that all the counters and DRO's have reset on start-up to the correct settings. I am not talking about configuration - since this does stay set. I am thinking of offsets and such, which can easily become altered.

You should really "home" your machine when starting. You do not need switches on it, but you need to put the machine at the same position everytime, on start-up. If you can attach a couple of blocks on the bed, so you can jog the tool holder up to these blocks (slowly) until the motors miss, and then stop, you can then "RefAll Home" or if you are using the lathe screens, "Set Home x" and "Set Home z". This position can be anywhere on the lathe bed, and might be better at the tailstock end, so they do not get in the way of normal operations.

If you check the DRO's, the Machine Co-ordinates should now be 0,0. ( The machine co-ordinates button should be lit  - or on the lathe screens - selected) The machine always keeps track of it's position in  machine co-ordinates

I must assume that your work is always in exactly the same position - and your tools fit in the toolpost in exactly the same position. If so, then jog to your position  0,0 on your program. This takes a bit of doing on a lathe, particularly finding the lathe centre for the X axis. Press the machine co-ordinates button and the light will go out. The machine is now displaying "program co-ordinates" (Or on the lathe screens these are selectable on their own).

Type in the MDI line G54 ( The default offset ) and zero both DRO's

You will see the display got to zero, but if you have a quick look back at the machine co-ordinates, they will not have changed. This is the "offset" for your program. If you look in the  Config/Fixtures you will see that G54 contains these figures. The other offsets should all be zero - if they are not, then clear them all.

Copy the figures from G54 to say G55, or any of the other slots. G54 is prone to change if you are not careful (indeed they all are) so do not zero the program co-ordinates again, or it will change the offset.

You can either add the G55 offset to your program, or type it in manually when you start the day or indeed, at any time after you have homed your machine. It is more professional if you add it to your program. Make a note of it somewhere, because, providing the stop blocks stay in the same place, that is always the offset for that program.

And This is Why.
At the beginning of the day, run your table up to the blocks, and zero the "Machine Co-ordinates" That is the job done - your machine knows exactly where it is.

If you have added the offset to the beginning of the program, there is nothing else to do, when you press cycle start, the offset is entered and the machine then moves to the correct position for that program. You must make sure, of course, that there is a path for the table, and tool to the start point, without hitting anything else.

If you have not added the offset to the program, type it in on the MDI line. The program co-ordinates will change to take account of this, and if you were to type in the MDI line G0X0Z0 then the machine would move to the 0,0 position of your program. As it is you can just jog roughly up to the start position, it doesn't matter exactly where, press the cycle start and the machine will go to the correct start position anyway.

Note: Once you have zeroed the "Program Co-ordinates" when setting up the offsets, you do not need to zero them again at all. You must not zero them while the program is running, or the offset table will alter.
On a lathe the x offset is usually always the same, so only need be fiddled with once. The Z offset might need altering for different program.


Thats a bit of a diatribe - but I hope it helps.
















Not me driving the engine - I'm better looking.
Re: homing accuracy
« Reply #23 on: September 11, 2008, 10:17:31 AM »
Hood,

Find attached a sketch of the drive test. The fuzz on the right side only occurs occasionally otherwise a straight line. I will follow procedure outlined by Jim Pinder to see what I am missing there. This is a very good explanation of ref all home, machine coordinates, and G54, etc.

If the drive test shows something I should look at let me know.

Thanks,

B.E.N.

Offline Hood

*
  •  25,849 25,849
  • Carnoustie, Scotland
    • View Profile
Re: homing accuracy
« Reply #24 on: September 11, 2008, 10:20:29 AM »
Yes your drive test is bad, it should be simlilar to the first part of it with maybe an occasional spike. Have a look on the downloads page and right down at the bottom you will see optimisation steps you can do, that may sort out your pulse problem. It still may not cure things but it should at least eliminate one problem if you are getting a nice clean pulse.
Hood
Re: homing accuracy
« Reply #25 on: September 11, 2008, 11:31:31 AM »
Hood,

I looked under downloads on the website. I see a couple selections. I went into Mach and found some stuff on XP documentation. I did not find optimization steps. Can you be more specific. I am running windows 2000.

B.E.N.

Offline Hood

*
  •  25,849 25,849
  • Carnoustie, Scotland
    • View Profile
Re: homing accuracy
« Reply #26 on: September 11, 2008, 11:48:49 AM »
B.E.N.
The optimisation is XP specific but a lot of it can be done for W2K, you will find the steps here
http://www.machsupport.com/downloads/XP_Optimization.txt

Hood
Driver Test?
« Reply #27 on: February 11, 2009, 11:41:36 AM »
Hello again and Happy New Year.

Things had been going very well with Mach 3 for awhile. I managed to figure out a zeroing technique and I moved the cable for one of the motors that was erratic to a spare driver and life was good. Now I am back to that same motor doing same thing again. I can run for 20 minutes sometimes and others 2 minutes. While the motor is turning it has seizures and looses steps. As a result, the zero position is lost.  I have slowed down the drive no help and it was working.

I will verify that I did the optimisation, I think I tried that.

B.E.N.