Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: m.marino on March 26, 2012, 04:56:30 AM

Title: Calibrating (tuning) twin axis
Post by: m.marino on March 26, 2012, 04:56:30 AM

I have X and A axis together and A IS slaved to X. Problem the steps per millimeter is not the same on both axi' and having issues with getting calibration done so that I can move forward with final set up and cutting. Tools being used are Steel engineers rule, a Vernier Calipers (150 mm) and a calibrated Dial Gauge (certified to .01mm on one rotation or less and 5um on repeatability in that range).

Now what I have been doing is swapping the driver control from X to A and back and that has gotten me closer but still get error after to single axis run out of greater then 10 movements of the x axis plane. When doing air cutting tests will get .2mm error approx' in x+ and the same or slightly less in y- from the point used as 0,0. Velocity is set at 4250m/min and acceleration is set at 225mm/min^2. Movement quadrant is set up as X+,Y+.

Motors are 3Nm Nema23/4 (off size 23's) and being driven by a 60VDC 10.0A ASTEC power source Drivers are set at 3.7A (as the next setting is over the 4.2A that the motors are rated for). Drivers are rated for up to 80VDC and are based off the M752-G type series just a bit bigger and older model. The Screw are ball screws from linearmotionbearings on ebay and are 16mm dia and 5mm pitch. Micro stepping is set at 1600 on all axi'.

Hopefully that is enough information and hopefully someone can help me on this as have clients that are being very patient with me for parts to be made and have convention season upon me very soon.

Title: Re: Calibrating (tuning) twin axis
Post by: Hood on March 26, 2012, 05:22:18 AM
If you have an axis slaved then the slave has to be the same as the master, it receives the same  steps per unit as the master.
If the axis are not the same then you would have to use them as separate axis so that you can use individual tuning and then you would have to command a dual axis move.
It seems as if your axis should be the same so possibly it is something like Step active state that is set wrong or maybe even you need a wider pulse.
What I would suggest is you detune the axis by at least 50% and see if you can get it working that way as it may simple be a case of you pushing the steppers too hard and thus are losing steps.
Title: Re: Calibrating (tuning) twin axis
Post by: m.marino on March 26, 2012, 05:41:59 AM

by detuning, which would be better to start with 1st, acceleration or velocity? the other part makes sense but I am currently getting jamming when they have matched steps per and the A axis is coming up about 1mm short (don't have 2 dial gauges and stands to get exact measure on it versus the x axis for calibration)when the x axis is showing about .03mm over on the dial. I will start with bringing acceleration down and see how that goes. Also if you want send me a PM and will send you my number as I am in Edinburgh. thanks for the information and will be keep plugging at it.

Title: Re: Calibrating (tuning) twin axis
Post by: Hood on March 26, 2012, 12:15:17 PM
I would reduce both for testing and if it helps then wind them up one at a time.
To me acceleration is more important than velocity, if you have to wind down the velocity to get a slightly better acceleration it will usually pay dividends. My experience however is rigid metal cutting machines so that may not be the case for everyone.
Title: Re: Calibrating (tuning) twin axis
Post by: m.marino on March 26, 2012, 01:18:47 PM
Thanks Hood,

Cut acceleration to 175 and velocity down to 4000mm/min and things running a good bit better. testing out y axis now as I got the calibrated Dial gauge on saturday and had calibrated y before with a non calibrated gauge and vernier calipers. Shaping up much better, x is now spot on after 1000 lines of code run and Y was out by .2mm after the same. which for me needs to be fixed. I cut plastics (Acrylic, HDPE, etc.), Hard and soft woods, and will be getting into brass, bronze in the near future (PB102 to PB105 bronze).

looks like I will be takingthe advice of the man who built it for me and setting up a few profiles to deal with different materials.

Again thanks and progress made, hopefully this will still keep nice tight corners with the system set at CV.