my experiance G540 VS Xylotex

[Fixittt]

first I want to point out that this is no way a xylotex bashing.......
I currently have a xylotex on my maxnc and it is rock solid!!!!! Now.
My first xylotex setup on the maxnc took a dive (power surge due to ice storm) over a year ago. It was replaced with the plug and play 4 axis Xylotex kit. it has been rock solid after I learned a few things about the nature of small desktop CNC.

So that left me with a 24V power supply, 3 269 OZ steppers LPT cable. Did some research while working on the minitech cnc frame. I decided to go with the G540. Read the specs thought all I needed was the drive, seeing as I had everything else, boy was I wrong. not fully understanding the limit resistors ect. ordered the g540, excited and waited. I did more research and finally understood the resistor setup, luckily before the drive showed up. So a run to the local rat shack reviled that the 2.8K resistors I needed, were not there, so I sacrificed and got some 2.2 K reisitors. Got the drive, got the soldering iron out and burn my fingers soldering up some cables. Got it all wired up and setup mach 3.... Go to plug in the lpt cable and POOF, nothing goes smoothly. Seems the cable for the Xylotex is male female. You need a male male cable for the G540. Another oversight on my part. Tear thrue my box of cables and no dice. So I gear up jump on the ninja and blast into town. $5 later I have a gender changer and Im back in business. Plug it all up and Fault light....... some emails to Marcus and wow I finally understood the charge pump. and how to setup the pins. Still fault. more digging....... Estop.......... jumper ed the estop and GREEN LIGHT BAYBE!!!!!! Setup the motors steps per...... and start testing.
now realize that I am severely under powered in 2 areas. 24V power supply and 2.2K resistors. During the motor tuning I was able to max out the motors 3450 mm per minute at 45,000 kernel speed and no stalling. finally decided to give the machine a max feed of 2000 MM per minute. Less wear and tear.

Now with the xylotex driver and 269s on this machine it would rapid at 700mMM per minute with dampeners on the motors. I am currently running a test wax with the g540 stroking the machine at 2000 mm per. There is no resonance the think of and the machine is smooth as glass. Even with the small power supply I am ecstatic with how this drive performs.

I will keep the old faithful Maxnc in its current configuration with the xylotex kit on it. Like I said, its rock solid and I can leave it to run and not worry about it. The new setup on the other machine is still new and will take some time before I am comfortable with it. but so far....... It was worth the investment.

Xylotex support is top knotch! And they have a great product! I am pleased with the performance on the maxnc over the stock junk config.

Gecko has great support as well as all my emails were answered. The drive so far is performing more then I had hoped with the power limits.

Ramble over.

edited to 700 MM per minute... OPPS

[Fixittt]

ok, first experience cutting a wax model. I am assuming that at 1500 mm per minute cutting speed I am saturating the PSU. (Could be wrong) because there is distortion on the model in all directions. my girlfriend has my camera so I cant show you. but the Z axis is making a dome in the Z + direction My guess is loosing steps downward and it machines. Also features that should be straight are bent....... and Z zero position is lost. My guess is that all home positions are lost.

I checked backlash and repeatability of the machine, Its spot on. So my guess is that I just done have the power needed.......... any other ideas?

[Crevice Reamer]

If you are still using the 2.2K resistors, then you are running a 60V 2.8A motor on 24V & 2.2A. This would make it both under amped and under volted.

The under amping is the least costly to fix.

CR.

[Fixittt]

right and I agree........ but at the time, it was all that was avaialable, and im the type of guy that when I get something, I wanna make it work, not sit on the shelf. So new resistors are going to be on the way.... and hopfully a new PSU 48V 7.7a will be soon after.......

[Mariss Freimanis]

Fixitt,

Set up a test program that moves the z-axis up 50mm, then down 50mm at your 1500mm/min feedrate. Put a mark from the motor shaft to the motor body. The marks must line up after every up/down cycle. Repeat the test cycle many times to see if there is any drift to the mark position. If there is no mark line-up position error then the drive and motor is doing exactly what it was supposed to.

Mariss

[Fixittt]

ok, second attempted. Changes made...... I lowered all the motor tuning values. Max speed was set at 700MM per. Re ran the same toolpath....... almost mirrored results from the first test. It is spiraling from the inside to the outside of the model. Z is loosing height as it progresses thru the program. (Cutter gets deeper and deeper)

[Fixittt]

great idea.... take out all the machine components as variables..... Am running a raster in X test code now........ I will do your testing after this is done......
Thanks for the input!!!!!

[Fixittt]

I feel like I must say this, this is just like any new build, you have to work the bugs out of the system as a whole. This is the debugging stage which I knew was coming...... if it was easy it wouldnt be fun... but does it have to be this fun?

[Fixittt]

some testing that i did do what setup a dial test indicator and a block screwed to the X table. Put the tip of the DTI on the block zeroed out the dial and ran some back and forth up and down tests to test for repeatability. They were short moved inch and a inch and a half, I did this about 5 to 10 times and it always came back to within .0005. I will do the longer strokes as suggested and more reps. I will use the DTI and the makr on the stepper body. backlash is under .003 in.

[Mariss Freimanis]

Fixitt,

Getting the same results the second time says a lot: Error in g-code or error in Mach3 setup.

Step motors make funny sounds when they stall that are hard to ignore. You are not running your motor very fast so 24VDC and 2.2A is reasonable. Going to 48VDC will double your speed, going to 2.8K instead of 2.2K will increase your low speed torque 27%. Neither will have any effect on the problem you're seeing.

Mariss

[Fixittt]

I agree that the same results twice at drastic different speeds says something. As for the G-code. It is a proven code that I have cut before.

The computer that mach3 is installed on is a fresh install, optimized for mach 3 (as per optimization sheet) fresh install of mach 3. (Driver test is a near dead flat line) No network access ect.

I cannot tell you which version yet of mach3 it is, but I am pretty sure it is one of the new builds as the drivertest screen is unlike my other mach 3 on the other cnc mill.

[Fixittt]

ok, this is very interesting to me, I am sure the solution is going to be easy, I just dont know what it is.......
I did a small program. basically a square rastered toolpath from left to right small Y movments down the model. Here is what is interesting. As Y would move thru the model Z put a wave on the surface. imagine this a box. and as you fro from the front of the box to the rear. the surface height is a symmetrical wave. Sure wish my girl would get home with my camera. Pics are alot better then words.
The wave starts at the front of the model. As the program progresses thru the model the z height gradually goes down, then back up to peak then back down then back up.

Im sure its a setting.


BTW, this thread was intended to possibly help others who might develop the same issues that I am having to face. No judgment or ill feelings are meant to anybody, company or product.