What's the Next Feature You Would Like?

[rungemach]

Hello Stephen

The shark has gone through several variants of controllers over the years, Originally the had a parallel port controller (Xylotec) with Mach3 software as the interface.

Subsequently, NWA developed their own controller and software, based on the Ncpod board, which brought in the USB interface and eventually dropped Mach3 support.
A lot of features were lost when Mach3 was no longer supported in favor of the NWA product. I suspect that you may have a heat related problem with the x axis driver circuit in your controller. If the machine functions well cold and does not have an exhaust fan in the case, I would try adding airflow past the circuit board(s) as a quick check.

If you are out of warranty, feel comfortable working with basic electronics and wiring (as you seem to have done already), and would like an industry standard interface and controller, I would suggest you consider moving away from your troublesome NWA controller and replace it with a Gecko 540 driver using Mach3 on a PC to drive it. If you would like a usb interface you can add a smoothstepper board. IMHO you will end up with a much more full featured interface and a pretty bulletproof motor driver system (Gecko 540). I use this setup on 4 machines now, ( router, lathe, mill, 3d printer) and have had zero problems. On each machine I have two instances of Mach3 , one metric and one imperial. Depending on which I chose, it is like having two different machines which you can select. Perhaps most important, you can easily add limit switched to prevent "hitting the stops" and causing undue stress to the machine. That feature ( and many others) went away when the controller design was changed to NWA proprietary.

[4DThinker]

I would like to have an x axis that doesn't "glitch". I'm throwing away a lot of blanks due to this issue.

With no marks on the Shark beds indicating the cutting limits I found I was occasionally and mistakenly clamping down my work with part of the toolpathing running outside those limits. This would leave some projects cutting fine, but the occasional one (even if it cut fine before) getting screwed up. No feedback to the controller to tell it the router has bumped into a wall, so it'll keep cutting but be "off" by the amount past the limits any tool path ran. If the router bumps against the limits often it'll eventually wear slop into the t-nuts that run on the threaded rods. It may also loosen the coupler nut enough to show some play changing directions. I eventually had to replace the t-nut on my X axis, as about 1/16" of play would ruin any cut I needed accurate dimensioning in.

I've got tape on my blue t-track bed now that delineates the front and back (Y) limits. I generally clamp a guide board along the left limit to clamp my project boards against. Since doing this I've stopped running out of bounds and have seen no "glitches" in either direction.

[stephenb79]

I left the same blank on the table re-establishing zero and raising z up clear of the old work piece. I loaded the same program I ran last night while losing the blank. Pressed run and it ran perfectly traveling back to my zero position or home that is. I did nothing besides disassemble the t nut and remove x axis motor inspecting for coupler slop. Both the ball screw and motor have flats milled so slippage is unlikely. I grabbed a pair of pliers and grasped the motor shaft while attempting to twist by hand the ball screw. Nothing. I'm starting to lean towards the heat issue. However, yesterday my x lost position on the first run. Will that much heat be present 3-4 minutes into the 1st run? To discuss the limit idea I would like to describe my setup. Whenever I replace the table which is rare, I add new 3/4 a/c ply and then a layer of 1/4 luan on top of that. With the outer limits and a v groove bit I manually drive the machine around making the 25x25 square. Now I know exactly where to place my work pieces. In the prior issue described, my +X had 10 inches of travel left on the slide and 5 inches of -Y. It is odd though. When the x acts up, it sounds exactly like it is at the stops trying to over travel. The whining or struggling sound. The controller does have a fan installed drawing good air flow into it and out the side vents. Well My story is told. Really do love love the machine. Just wish I had a wand to wave over it as I sit and scratch my head. Thanks for all the posts guys!

[rungemach]

Hello Stephen

Thanks for the explanation of what you have tried and how your machine is set up. it helps eliminate possibilities.
Since you have been inside your control box,and it has a fan, from what you have said I am assuming you have two circuit boards inside , not just one.
If that is the case ,one will be the usb Ncpod controller and the other a stepper driver board.

4 minutes should not be long enough to develop heat problems in a good circuit board , especially if you have fan circulation. I was thinking more like a half hour from dead cold.
The newer controllers I have seen did not have circulation fans, so I am wondering what driver board you have. The older units used a Xylotec board.

Since the motor stops driving the screw and still makes a pulsing or chattering noise (like if it is run against a stop) I am thinking one phase of the x driver circuit is intermittent. If you changed motors and the problem stayed with the controller x axis, it could be a bad solder connection or a bad component. If you lost a step or direction signal from the controller board to the driver board, you would have no drive at all or drive in one direction only.

If the axis is not binding and moves freely, the only other time I have seen "stopping while still making pulsing sounds" is if the axis is driven to too high a speed too quickly and stalls because the drive train can not accelerate that fast, or if the motor & driver hits a bad resonance frequency with a driver circuit that does not compensate for resonance. I had problems when running over 75 ipm on my old shark, but if this is happening below 50 ipm it is probably not an issue.

Intermittent problems are hard to diagnose if you do not have the luxury of just letting a system run for long periods of time. You usually end up taking your best guess and substituting/replacing the most likely component. If you have the older Xylotec board, you may want to consider a newer driver.

[stephenb79]

I'm convinced... I tore the controller down and soldered my x stepper leads. The problem persisted. The white flag went up. I even inquired about a new controller. A thought resurfaced from the early stages of this nightmare. "Take a flat head screw driver to each one of your punch down block terminals for the respective axis and give it a slight tap to maybe seat the wire a little deeper." The flag came back down, and I disassembled the box for the 10th time. Tapped each terminal. Reassembled. Completely lost my X. Tore the box down and eliminated all punchdown block terminations. Soldered every wire one by one directly to the board. Reassembled. It runs like a dream now. Moral of the story if you have punchdowns in your controller, get rid of them. They aren't worth the hassle. Keep carving friends...