Gecko tuning and System Measurement results

[SMW Precision]

Gecko Tuning X-axis-

One of the reasons for my CNC adventure is to work with Scale and measurement data towards a goal of true closed-loop. I have done the initial mounting and testing on X-axis using a 1 micron Certified High Precision Scale. I’ll cover what was used for equipment in my testing and early results. System installed was a standard CNC retrofit kit supplied by Industrial Hobbies.

Scale- Foic Certified High Precision Glass Scale 0.001mm resolution +/-0.003mm deviation within 500mm travel, repeatability +/- 0.001mm the metric equivalent to 0.00004”
Gecko 320 Servo Card
¾” Precision Rolled Ballscrews specified at 0 to .003 +/- deviation per foot Encoder ballscrew shaft mounted with a 4 to 1 reduction
Industrial Hobby High Performance Servo motors

I was unable to measure any backlash from the double nut ballscrew at 0.001mm registration. Ball nuts are set at approximately 100lbs of torque as per IH specifications.

These are initial results while not definitive they are suggestive. Perhaps other group members will suggest other testing parameters.

Gecko 320 unit grossly tuned Damp and Gain were set in the 11 o’clock position. Testing done was with using current Mach 3 software. With these settings it became clear that system would not hold a repeatable Zero point. System was variable in repeatable accuracy and deviation showed increase in measurement deviation as table was operated in both directions multiple times.

In tuning I was able to get a consistent and repeatable Zero registered to within measurement system deviation specifications. While ballscrew deviation in length was only grossly tested; it did not exceed 0.0015” at any point.

[SMW Precision]

Originally Posted by SMW Precision While ballscrew deviation in length was only grossly tested; it did not exceed 0.0015” at any point.

Ballscrew deviation is easy adjusted out with Mach 3; I was asked to mention that point.

[Shepard]

In tuning I was able to get a consistent and repeatable Zero registered to within measurement system deviation specifications.

Forgive my ignorance but do you mean dialing in your geckos or some arcane software tuning (i'm using mach3) ?

[SMW Precision]

Originally Posted by Shepard Forgive my ignorance but do you mean dialing in your geckos or some arcane software tuning (i'm using mach3) ?

This was done using Mach3 and its DRO readout. I was looking for several different elements.

1. Was how critical is tuning on the Gecko drive in positional accuracy. What showed from my testing is for positional accuracy Gecko's must be correctly tuned. You need to get a repetative zero at the same point or you are in trouble accuracy wise.
2. Was to see what the uncorrected accuracy of IH ballscrews are from my testing they are quite good. With minor correction I should be able to dial system in at about +/-0.0002".
3. What is the backlash on IH ballscrews. It is below the 0.00004" resolution scale I was using. In other words I could not measure any backlash at that resolution.

While it is normal to set Gecko's with a dial indicator going back and forth between table and PC is time consuming. It is also limited to indicator movement distance. With a scale display which was also at PC and I could conveniently measure accuracy for total table movement rather than a bit more than an inch.

[Cruiser]

Gecko's are not set with dial indicicator ! They can be set pretty darned close without a vom or scope, just by driving the axis back and forth and nudging the damp & gain till it runs smooth and will sit still with minumal buzzing. After the gecko's are set then the settings in motor drive are set with indicator to assure they move the asked for distance. If i understand it correctly, the gecko tuning is critical for proper locating on pulse edge, and for smooth running without bucking, and growling, which would be indicator of how tight or loose it is when driving the axis. is this what you were after here ? Does your glass slide act as encoder or just run a DRO ?

[SMW Precision]

Originally Posted by Cruiser Gecko's are not set with dial indicicator ! They can be set pretty darned close without a vom or scope, just by driving the axis back and forth and nudging the damp & gain till it runs smooth and will sit still with minumal buzzing. After the gecko's are set then the settings in motor drive are set with indicator to assure they move the asked for distance. If i understand it correctly, the gecko tuning is critical for proper locating on pulse edge, and for smooth running without bucking, and growling, which would be indicator of how tight or loose it is when driving the axis. is this what you were after here ? Does your glass slide act as encoder or just run a DRO ?

Crusier,

Procedure you mentioned is where I started. Gecko's were set for that response, movement and Stop looked and sounded good. Though when scale was attached it was easy to see that they were not correctly set. I was not actually returning to the same zero point. Each pass showed increased deviation in measurement. While the amount was fairly small it kept growing.
After tuning I was consistently hitting the same Zero point for the Axis. At this point I am using a DRO display. Later I will connect Scales to a Sound Logic Interface board and load in Mach3 add-on software for board.
While this system is not what I want it is the closest thing out there. You can have true DRO position displayed in Mach3 with the ability to set a deviation parameter that will halt Mach3 if exceeded.
How well this is going to work I have yet to experience. Though scales make setting Gecko and mapping table an easy process.

[mvcalypso]

Hi - from the thread, I'm not sure what you did to set the geckos up. Have you tuned them via the procedure in the Gecko manual (which requires a signal gen and scope to set the transient response)?

Dave