Erratic Destructive Behavior

[Tim Owens]

Drew,

Did you send your unit in for us to look at?

The controller has a double check to make sure that the data gets over to the controller correctly. Usually you would get a communication error if it was with the USB Cable. However in your case with the erratic issue it might be a memory card. if your memory card (micro sd) has a problem that could be an issue. This is located in the tiny slot on the back of the controller a penny will eject it if you want to try to replace it. Do not go over 4 gig.

But if you would like we can test it out and look to swap it out. That is not a problem if it is defective. What we don’t do is ship out a new one before we receive the other one back anymore ( unless we are give a credit card on file if the other was not returned) . We had issues where people seem to have trouble returning the other components.

Thanks

Tim

[Drew MacDonald]

Tim:

I have not yet sent the unit back to be evaluated. In all honesty, I have not been able to afford the resulting downtime on the machine, as I had, and continue to have, jobs to process. I am a one-man shop, without access to another machine in the interim.

Behavior continues to be hit or miss - however I have been able to mitigate problems by taking specific steps before running each job. So essentially, the issues still exist, I have just been working around them for the most part. When the machine does decide to do its own thing, it is now usually apparent at the start of the run, allowing me to intervene before damage is done (usually).

I will be more than happy to put a credit card on file, to receive a replacement controller before sending mine back. That is not a problem whatsoever, and I would like to do so as soon as possible. Please let me know who I should contact to accomplish this and get a new controller shipped. I will continue to check this message thread. Alternately, you can reach me directly at Drew@SheepDogWoodworks.com

[gordon1s]

Drew , Please call the support line for help. This may be a very simple fix. Best regards Jim NWA support.

[jeb2cav]

Hi Drew - did you ever get a solution you're confident in? If so, can you post some information on that? Since joining, I've watched some discussions on the SD card in the controller. Some of them I'm a little skeptical on, but this one seems to identify a similar problem, and a solution that worked for this member - http://www.cncsharktalk.com/viewtopic.php?f=8&t=773.

In either case, knowing what worked for you will be useful to all.

[Drew MacDonald]

Joe:

Still in the process of sorting everything out. I did finally get a new controller box headed my way, after putting a credit card on file with Next Wave (just to keep me honest and ensure I return the old one).

In the interim, I have been able to mitigate most of the problems - but not eliminate them completely. I go through a series of steps every time I run a job (such as powering completely down, waiting 30 sec, powering back up, etc) - all suggested on the forums - in an attempt to prevent the erratic behavior. This has reduced, but not eliminated the issues. I also started using particular settings, which allow me to determine early in the job if the machine is going to go off on a tangent. For example, observing that my erratic behavior typically involves X axis travel opposite of the intended direction at some point, I can try to use an origin point that will take the router away from the cutting piece if a problem occurs (now usually apparent at the start of a run). This has allowed me to E-Stop the job before any damage is done - for the most part.

As I indicated above, so far the "solution" has simply included steps to work *with* the issue, and has not corrected it. I greatly appreciate the information on formatting the SD card - I went ahead and did this. Unfortunately, right now, there is really no way to determine if this is the fix or not - based on my experience I am still waiting for the machine to screw up again.

I am hoping that the controller box replacement will take care of this permanently - as I strongly suspect that the root of the problem lies in the controller and/or card. I would imagine that Next Wave will bench test and troubleshoot the controller upon return - and I will try to follow-up and see what the final verdict is.

Will definitely keep you posted. Thanks again to all that have offered guidance and information - it has been greatly appreciated.

[Drew MacDonald]

Just wanted to post a quick update.

NextWave replaced the controller box. I have run multiple jobs with the new box, and have not experienced any issues whatsoever. Whether it was the box itself, or the memory card I don't know. Hopefully NextWave will post a response indicating what issues were found with the original box?

Thanks again to everyone who offered guidance. For the time being, the Shark Pro Plus is running smoothly.

[drueth]

Erratic behavior is the hardest problem to find in any system/machine/computer If the problem happens every time you do this ask your doctor and he/she will tell you don't do that. I have 30+ years working with very large computing systems and I have never seen a Shark pro. I want to by one so I have been following this site, but here are some ideas I have.

Please try these ideas with no bit in the router.

1. I am assume this is a time sensitive job ie the computer MUST supply new data to the controller as it is needed. And in my mind Windows is not so good at this. If you bring up the Windows task Manager and switch between the performance & processes window do you see your computer is doing anything else or the cpu usage spikes up. The one thing I have found over and over is the developement folks always get the fastest and newest machines. They should get the oldies and slowest like we all have. Try starting any program you have that is slow to open. Does cause the router to freeze or stop???

2. With the router running and no bit in the router gently tape all the connectors, move the control cables around and gently tape the controller . Plug an electric drill into the same outlet as the controller. If you start the drill does it behave erratically?

[Drew MacDonald]

DREUTH:

I am no expert in this arena, however I will share my understanding of how the process works, and what I have found through my own "failure" testing.

Regarding the data flow, I believe you are correct that data is supplied from the computer "on demand" - that is at run time, the computer systematically feeds the controller the G-code. I believe that the memory card in the controller is an intermediary of sorts, and serves the purpose of queuing a certain amount of code for "instant" access by the controller during the job. This makes sense to me - as data access on an SD card would be much quicker than accessing the actual compiled G-code file directly on the computer.

Depending on this card queue size (we know the entire file is not loaded, because the computer appears to run through the code for the entire job), it would seem that slight hiccups on the computer end, or bogging from services and programs running would not be an issue - at least in theory. I have run jobs with an without virus protection on, and even while playing an HD video solely for the purpose of trying to cause a failure as you suggested (this was when I was troubleshooting my erratic behavior). From my experience, services, programs etc. running on the computer had no effect whatsoever. I have repeated this "test" with the new box - no problems.

As for the power fluctuations - I did, in fact, check all connections to the controller and the stepper motors. My dust collector, for a time, was plugged into the same outlet as the Shark Pro - firing it up during a run appeared to do nothing but cause the router speed to audibly slow a bit. Job progress and accuracy appeared unaffected. I have a theory about this and why such power fluctuations should *not* cause problems...but I am far from an electronics expert. In terms of Amperage draw at startup, I would think that the dust collector would have caused an issue, if there was an issue to be had. Notably, just to be safe, now the only thing drawing power on the circuit to which the CNC Shark is attached, at runtime, is the laptop which controls it.

One thing I forgot to mention - on a whim, I removed the SD card periodically when experiencing issues, put it in a card reader, and checked its contents. Not knowing when data on the card is purged, or how this is accomplished, I can not say for certain if this was an issue - however - at the completion of a run, sometimes there would be remnant data which was still on the card. Again, I don't know when or how the controller "clears" the card. If it is at runtime of the *next* job, then this remaining data makes sense. If it is at the end of a run, then this may have been the issue. Not knowing the order of operations for the data flow to and from the card, it merely raised a suspicion that leftover "junk" data on the card *might* have been causing the problem...and not knowing if it is a software line item in the g-code, or a hardware routine in the controller which clears it, I was basically left with a theory and some head scratching.

Again, the new controller box appears to have solved my problems. Multiple jobs - some simple, some quite complex (graphic based diamond drag etchings) have been run without the slightest issue that was not based on user error on my part.

[Bags]

Drew, I have carefully read all of your posts since unfortunately I find myself with the same problem. A brand new Pro plus which I have been unable to use for a real project yet. I have a fairly large volume of some simple and repetitive signs to make. On setting up the Shark I made a quick sample and it worked perfectly. Did a few things just trying it out and all was fine. So I set up to start my signs--that was the beginning. Y axis problems which are erratic and z axis with bit destructive plunges, etc. You have described it all very well.
I have tried all the things that were suggested to you, lead screw connections, wiring, usb, etc. Finally tonight I set a fresh panel on the worktable and put a series of patterns to be engraved very shallow to check z axis integrity (and also check table). A large rectangle engraved at a depth of .02 resulted in a half rectangle with a faint line then a descending cut that was at an inch by the time I hit the stop. I am pretty frustrated at this point and will call NextWave. Your last post talks about SD card problems and the lack of information we have on how code is handled. I ran the same program multiple times and sometimes it would run fine, other times it would skip large sections of code and other times it would take on the destructive behavior that we both have seen. It makes me think that there is a definite control box/cpu interface problem. This is obviously a problem that continues to happen, hope there is a fix in the works.

[jeb2cav]

One thing that hasn't been mentioned in this particular thread - and you may have already addressed it - ensure that you have disabled the screen saver, and if using a laptop, ensure that you have changed the power settings for all devices on the laptop to "always on". Early on in my use I've seen this result in some of these behaviors. Although your case sounds pretty severe, and certainly a call to tech support is the best idea - and you paid for it as well. I hope you get it resolved and will post the solution back here for others to learn from.

Good luck.