Problem keeping my gantry square, using a dual drive slaved set-up with Gecko’s.

Problem keeping my gantry square, using a dual drive slaved set-up
with Gecko's.

Hi all,

I have been out in the shed for the last few days trying to get the
gantry square and also set the correct steps per unit. This is what
I have been doing.

Although this is for a plasma cutting table, I can not find any way
of working out the steps per unit in the SheetCam .set screen for
mach3. So what I have been doing is going into the mill, Mach3
screen, which is called 1024, I think. Once in there I can go into
the settings page and enter which axis to check/set-up etc.

Once I have done this I just transfer the number from the mill
screen motor tuning page back into the SheetCam plasma screen motor
tuning page.

So far so good,

Now what is happening is that I will get the gantry set up square.
I will reference the xyz, then zero the xyz. (Note: I do not have
limit or home switches on the table yet.) Then I will open up the
set steps per unit box in the setting page and I will select the x
axis to calibrate. I enter a value of 3000mm, the machine will run
up to the 3000mm mark, it might be out a bit say 2997mm. Then I
enter the 2997mm into the box that asks me how far it travelled. It
then say's it will set this number as the steps per unit. As this
is a dual drive set up I have been opening up the motor tuning page
and making the slaved axis A, the same number as the X axis.

So now when I go back and home the gantry is not square. It is only
clamped together at the moment so I can see that it has moved out of
square. And not just a little bit but I'm talking 3-4 inches.

The only things that I can think of that would be causing this are:

I am using 2 gecko's. One for each motor. For the current set
resistors, I need to use 270,000ohms. One gecko has the correct
resistor, but the other Gecko has a bunch of resistors twisted
together to make a value that is close to the 270K mark. I did
measure it with a meter when I first made it up and it was not
exactly 270K. I only had one 270K resistor at the time so that's
why I made a McGyver one up. I have just left the home made
resistor there.

Now I am thinking that if one has the correct resistor value of 270K
and the other has say 250K. Will this effect the synchronisation of
the two motors that are slaved together in Mach? I guess that one
motor will be getting say 6amps and the other might be getting 5.5
amps, or what ever it might be.

Will this make one motor move slower than the other motor and
causing it to go out of square? I'm pretty sure that it will effect
the motors acceleration as the other motor has 0.5amps more power
that the other.

I have even made my acceleration figures really low so that they do
not miss steps, but this does not do anything. The only other think
that I can think of is that the one motor on the one side has more
resistance to drive than the other motor.

On the plus side, It looks like that If I can fix this problem I
will be able to get my machine square, set the steps per unit and
then I should be able to move anywhere on the table within 1.0-0.5
of a mm. Pretty cool!

So any ideas with this problem of mine?

Cheers,

Peter
Australia