PDA

View Full Version : Breaking bits



jerrymorgan
02-24-2014, 08:59 PM
I've broke a 3/16 in one cm of red oak and 1/8 in poplar and I don't understand why?

fwharris
02-24-2014, 09:05 PM
Are you making these cuts at full depth or at a maximum pass cut?

bergerud
02-24-2014, 09:57 PM
When making deep cuts in a single pass, one has to watch out for climb milling. This is when the bit cuts only on one side and tries to pull itself along and climb out of the cut. (The direction of the cutting edge and the direction of travel opposite.) Climb milling can easily happen when cutting out boundaries of patterns.

Digitalwoodshop
02-25-2014, 01:25 PM
From what I am seeing, it looks like the bit was broken by a movement in the X Direction since it did not look like any side to side stray cut... a better picture with the bit removed would help determin that... Sort of a CSI investigation.... CarveWright Version,

To snap that bit, I would venture to guess that the Y would not have the Torque to do it... but I could be wrong.... We know that the X Torque can very well snap that bit and based on the investigation of the cut marks at the time of the break, it would be logical to expect the X Jumped.... That brings in my Friend the Brass Roller.... It could be the sole source of the problem.... The Board lost contact with the Brass Roller for what ever reason... From the board tipping up or bumping into a in feed or out feed tray roller... BUT COMMA... That does not look like the real problem as the CARVING is PERFECT.... Any Brass Roller Problems would be reflected in the Carving TOO and that is NOT the Case.... So NEXT....

Then we have something as simple as the feed rate for the Cut Path was JUST TOO FAST for the Bit's ability to cut the wood... I saw this when I was Vector Cutting plastic tags on a Carrier Board and using a 1/16th end mill. The machine used a feed rate that was faster than the bit could cut the plastic in one pass and since I used double sided tape to hold the tags to the carrier board, they would sometimes be pushed off the tape. The fix was to tell the machine I was using a 1/4 inch bit but actually use a 1/16th end mill and that locked the feed rate into it's 1st gear. It never shifted into 2nd gear the faster feed rate causing the problem...

Since you are dealing with solid wood, and IF the FEED RATE was TOO FAST.... The Cut would react in one of 2 ways... Cause the Board to SLIDE on the Drive Belts.... OR if locked in on a Rubber Belt..... SNAP the BIT.....

That is my take on this..... A Feed Rate TOO fast..... for a one pass Cut Path... Multi Pass Cut Path would solve the Broken Bits....

AL who

atldal
02-27-2014, 02:00 PM
Snapping bits was my first indication of my x tracking problem.

DJKnutsen
05-11-2014, 03:40 PM
I've just broken my second 1/8" cutting bit, on the same MPC, at exactly the same time-position in the project (months apart) and had a third instance where I was lucky enough to hit the stop before it broke.... In this project, it's just finished the initial carving, and requested the cutting bit... located and indexed the newly installed cutting bit, and moved to the spot to start that process... dropping the Z truck and slightly starting to move in the Y direction... but it never started the router motor!! and didn't stop the process or profuse any error message... so as I said, it happened three times, in the same MPC, the same point in the process, weeks and many other successful runs in between on other MPC's, twitch I was slow (but saw what happened) and broke the bit... once I was quick enough to kill the power and saved the bit... It seems to me that there is an error in the compiled (G code?) loaded onto the memory card for that MPC... but with no way to look at what is being sent to the machine, I can only guess. But I've not had any problems like this with any other MPC's, and yet... three time with this same one...??? Any thoughts... I'll call CW in the morning, but I suspect they'll try to deflect the blame away from a software or firmware problem... BTW..I'm running Designer 2.005, build 11800 with the latest firmware update, on a Mac desktop.

lynnfrwd
05-11-2014, 03:57 PM
You mean deflect like maybe saying it could be a project problem?

lynnfrwd
05-11-2014, 04:35 PM
Did you try re-uploading the project to your card? Maybe you can post that mpc file, so someone can look at it.

DJKnutsen
05-11-2014, 05:09 PM
OK... call me cynical... by deflect, I was implying that they might not want to consider that their software could be the problem.. I did empty the memory card, and re-load the suspect MPC.. between results one and two.. no help. I can't upload the MPC here as it was a purchased package from Dick Bipes (the Synchronicity clock) and have been in contact with him and was assured that the MPC works. I'm new to CW... and have lots to learn, and lacking updated (Designer ver 2..?) printed reference material, I keep trying other options... but I'm running low on tool bits and lumber..!

lynnfrwd
05-11-2014, 05:39 PM
Then I guess the question becomes, anyone else have same issue in 2.005 with Synchronicity Clock? Which mpc? I think there may be several.

DJKnutsen
05-11-2014, 06:14 PM
There are a lot of MPC's in that package. The one I was working with was the "half_parts.mpc"... I bought it hoping to learn how he did, what he did, so I could understand by example and results... and I am learning.... just traveling a bit more laterally than I intended..! Thanks

DJKnutsen
05-11-2014, 06:23 PM
I'm sure I should clarify... I am extremely new to the CarveWright world, and am undoubtedly doing everything wrong... and I'm well used to learning the hard way, I've just become accustomed to fall back on documentation to fill in the gaps and goofs... and when you get a firmware update, but no explanation as to what was changed, and how it might effect your operations, it's hard to identify the cause of some unexpected effect... me or thee??? I don't really care much which, I'd just like to not repeat the same thing unknowingly!

lynnfrwd
05-11-2014, 06:25 PM
See Changelog on update page.

DJKnutsen
05-11-2014, 07:17 PM
Thanks for the info... a bit had to find (http://www.carvewright.com/assets/downloads/readme_2_005.txt) but a lot of information there!

fwharris
05-11-2014, 07:44 PM
Are your cut outs using multi pass or in full depth? At the bit change did the bit/cut motor spin at the start of the bit measuring step? Most times the cause of the cut motor not coming on is the cut motor switch (switch on the cover on the bit plate end of the machine) is not going closed.

DJKnutsen
05-11-2014, 07:56 PM
Yep... it did spin up when entering the new tool measuring routine. I've never gotten to actually cutting anything at that phase... The first time it broke the bit (and ended that board-try), next time I killed the power (and ended THAT board-try) and this last time... I stood there and watched as it broke the second bit (ending that... and so on!!) ???

DJKnutsen
05-14-2014, 03:55 PM
ANSWERS!!!:D I talked to Lynn at CW, and she told me a lot about the evolution of the Designer program, and how there were mostly additions to the original coding, and that there really shouldn't be any hardware conflicts between the two (1.87 and 2.xx), and that the only problem encountered with Mavericks was due to the NAP function designed into the system for portables and battery life.... So, I set up my system with an empty sled, and in as much as I wasn't actually carving anything, I didn't turn my vacuum on.. (I have the Dust Devil Collector installed). I got everything purring along, and after about 30 minutes, the router motor shut off..!!! Using the OPTIONS 8 motor test function, it said that the motor should have been running, but it wasn't! SO.... poking, prying...twisting I discovered that the cover's motor interlock switch on the right side wasn't making, and if I pressed hard on the cover, I could get it to run... Long story-short... I'd had to cut the cover down to accommodate the Dust Devil, and (being a machinists at heart!!) I made it a nice tight clean looking job... and it worked... right up to the point the things heated up and then that switch wouldn't make... My shop is quite cold, so running Dick's MPC took about the same amount of time to heat up and fail (all three times).... if it had been much warmer for my last run, I'm sure it would have stopped sooner. I am waiting on CW to deliver my replacement cutting bits, so I can't actually run for real 'till then, but I'm sure it'll work as designed.

It's interesting that the motor interlock is hard wired into the motor control circuit, and not monitored by the program... so the program thought the motor was running therefore no error existed..!? Ta..da.. broken bits!!! It's all on me, but boy am I happy to have found the answer!!!

lynnfrwd
05-14-2014, 04:03 PM
This is about the 3rd or 4th time I've seen this be the issue. It is better to overcut on the cover than to be perfect or undercut.

DickB
05-14-2014, 04:26 PM
Why does the software not monitor the cut motor speed and stop all movement when the cut motor interlock switch shuts the motor down?

fwharris
05-14-2014, 06:27 PM
Why does the software not monitor the cut motor speed and stop all movement when the cut motor interlock switch shuts the motor down?

We get "check cut motor" errors a lot of times so why wouldn't it be possible for the system to check that the motor was not running?

bergerud
05-14-2014, 07:02 PM
I have wondered on this as well. Why not monitor the rpm sensor?

DJKnutsen
05-15-2014, 09:08 AM
It seems like it would save a lot of heart ache if there was a check for RPM's before a "Z axis down" was initiated... No RPM's equals an "Check Motor" error, and a stopped process..? It sounds simple... if you say it fast..!

bergerud
05-15-2014, 09:32 AM
I think the software does the rpm check before a vector cut starts. I have had the check cut motor halt before vector cuts. It just needs to also do it before raster cuts.

(It would be even better if the rpm could be continuously monitored.)

DJKnutsen
05-20-2014, 02:59 PM
I had to resort to buying 1/8" cutting bits from eBay because CarveWright was out of stock, and couldn't even forecast when they'd be getting more in... Fortunately I had already procured an ER-11 collet chuck, so the bit's 1/4" shaft was no problem.... So, armed with cutting bits, I ran the MPC that had given me problems, and as I had surmised previously, it all worked just fine!! YaHOO!! I also noticed the the little tabs that actuate to two cover switches are different... the right hand (Motor) switch is much more sensitive the the left hand cover detection switch, so that accounts for why the computer didn't detect the 'not fully closed' cover.... stick that in the food for thought folder... Cheers!

Metallus
05-20-2014, 03:15 PM
motor interlock is hard wired into the motor control circuit, and not monitored by the program
It is a gov't safety requirement that the cutmotor has a physical override.


(It would be even better if the rpm could be continuously monitored.)
The software continuously monitors the cutmotor rpm. The problem is we have a threshold for failure, and we can only read the rpm sensor so many times per second.

bergerud
05-20-2014, 03:32 PM
The software continuously monitors the cutmotor rpm. The problem is we have a threshold for failure, and we can only read the rpm sensor so many times per second.

Monitoring the rpm is not enough. It needs to halt the operation if the cut motor is not spinning.

Metallus
05-20-2014, 03:52 PM
Even if we get a reading glitch? Nah. You're just opening up a different issue.

bergerud
05-20-2014, 04:14 PM
I see your point. It would be just another reason why someone's machine stops working.

bergerud
05-20-2014, 05:40 PM
How about if it was one time only. Zero rpm fault stops operation. Abort or Continue? If continue is selected, further rpm faults are ignored (until reboot).

Metallus
05-21-2014, 08:05 AM
How about if it was one time only.

Latency kills that thought... it could be more than 1/2 second before the abort.



If continue is selected, further rpm faults are ignored


that's what we already do.

GVigue
08-14-2014, 01:30 PM
There is a glitch in the software because I have opened my cover and the bit stop spinning but the program kept running and the bit broke. I had to hit the stop button to stop it. This has happened just once but it indicates a problem in the software to me. The cover safety switch should be hard wired to stop all motors, not just the bit. The software would have to back up a few steps when restarting the project because of the delay in receiving the input signal but that is easy to overcome. Unfortunately it seems we have to purchase the Designer upgrade to get any firmware fixes for the machine. Someone correct me if I am wrong please. If that is the case this won't likely be fixed unless you buy the latest or a future Designer upgrade. I suggest that you always hit the stop button before opening the cover to be safe.

dbfletcher
08-14-2014, 01:43 PM
There are two micro switches on the cover... one is the return path to the cut motor... and one is the "cover open" switch. It should like your cover open switch isn't working correctly. If that switch was faulty, I believe you would see the exact symptoms as you describe... you life teh cover and the cut motor stops.. but everything else keeps going.

GVigue
08-14-2014, 02:15 PM
Well that does make sense Db but it does show another problem with the design. It indicates to me a poor execution of the software and safety switches. They should have used magnetic reed switches or other type of sealed switches given the environment they are in and as I mentioned earlier they should be hard wired to cut off all the motors. Safety switches should never be software controlled. Dividing the task between two separate switches is a bad idea too. One switch will always lag the other which could cause a timing issue especially since the software cycle is very slow - "could be more then 1/2s to process a signal"?. Two hard wired switches in series would have provided redundancy however. This would have saved me a bit as well as the OP. Now I have less faith in the cover then before :P

CW-HAL9000
08-14-2014, 02:24 PM
You will be told for safety sakes to leave them alone but I bypassed the two cover switches in the first month I owned my first machine. I just push stop if I want to stop the machine. Anyone who opens the cover and sticks their hands in while it is running deserves the ouch they will get. LOL.

mtylerfl
08-14-2014, 02:49 PM
Well that does make sense Db but it does show another problem with the design. It indicates to me a poor execution of the software and safety switches. They should have used magnetic reed switches or other type of sealed switches given the environment they are in and as I mentioned earlier they should be hard wired to cut off all the motors. Safety switches should never be software controlled. Dividing the task between two separate switches is a bad idea too. One switch will always lag the other which could cause a timing issue especially since the software cycle is very slow - "could be more then 1/2s to process a signal"?. Two hard wired switches in series would have provided redundancy however. This would have saved me a bit as well as the OP. Now I have less faith in the cover then before :P


The switches are mechanical, not 'software controlled'. Also, the original design had sealed switches. You can thank Underwriter's Laboratories (UL) for forcing CW to use the switches you see now. The folks at UL thought 'their approved switches' were better.

In any case, it is important to clean the switches every now and then. Blow out with compressed air and a squirt with WD-40 if necessary. If they were sealed (like CW originally designed) we wouldn't have to clean them at all. Thanks again, UL!

Metallus
08-14-2014, 03:05 PM
"could be more then 1/2s to process a signal"

This is for the rpm sensor, not the cover switch. The cover switch is instant and raises the head immediately.

fwharris
08-14-2014, 03:20 PM
The proper way to "pause" a carving is to press the stop button ONCE. Doing so tells the computer to halt/stop all operations at the same time so you will not have that delay that you get when you open the clear cover.

GVigue
08-14-2014, 03:23 PM
Metalus, if that's the case then this should not have caused the broken bit for the OP should it? The head would have raised the bit when the switch changed states thus saving the bit? I dunno... too much contradicting info here lol. Mr Tyler says the switches are mechanical only. I had assumed at least one was software controlled based on the previous comments. It appears that there is one micro switch on each corner of the cover so I can see how if one corner was raised a little the other switch may not detect it. A double pole switch at one location would make more sense in that case. I do agree; UL has some wonderful ideas! Maybe Mr Tyler could tell us which switch was in the original specs? I would like to replace mine next time I clean the unit. It might be available in a double pole version too. CW-Hal has a point too... The cover should be backup for emergencies only and not for daily use. Pressing the stop button is safer.

mtylerfl
08-14-2014, 04:35 PM
I have no idea exactly what the original specs were...this information about the original sealed switches and the UL swap was shared publicly by a CW Hardware Engineer during a presentation at a CW Conference two years ago.

In any case, the switches are mechanical and the 'power off/on' status is detected by the machine controller (software?). No matter...it is what it is and we just have to deal with it. I'm very sorry you broke a bit, but FW is correct...the lid opening to stop the machine is only to be used in an emergency situation - which could potentially lead to bit breakage due to the impossibility for both switches to disengage at precisely the same time. In non-emergency scenarios, use the Stop button on the keypad and you won't run the risk of breaking anything.

EDIT: Please call me Michael or MT or anything except 'mr. tyler' ...I feel old enough already ;)

Digitalwoodshop
08-14-2014, 05:55 PM
The left switch likely was jammed with sawdust and that is the reason it never released and the computer never got the message to do it's software routine of stopping the Cut Motor Electronically AND Raising the Head.

The purpose of the right side Cut Motor Return Path Electrical Safety Switch is a Backup to the computer in stopping the Cut Motor when the cover is open. The computer turns off the Q1 Transistor Switch on the X Termination Board. That Q1 is the electrical on off switch.

As said before, the Right Switch is just a backup and it did it's job of stopping the Cut Motor when you opened the cover. Because the Left Switch stuck the computer never got the word to stop. So it is nothing to do with the software and having to buy something "soapbox".... Once you understand the machine and what happened it is not a sotware fault.

It is sad that the bit broke but I would rather hear about a broken bit then a mangled finger.

This just comes down to knowing the machine and being ready to react when we see something unusual like opening the cover and the head keeps moving and a little luck. This SAME thing happend to me with the more Robust Carving Bit once years ago and my only option was to quickly turn off the power switch before something broke. I got lucky.

I would recommend a replacement of the switch and recommend the LHR version. Retro fitting a aftermarket switch could result in another broken bit if not installed properly.

Remember you can always test the left side switch with Sensor Data on the LCD. The right side switch cannot be monitored on the LCD, it is just wired on the return path to the motor.

Transistors, SCR's, Triacs, all power control devices like the Q1 on the X Termination Board have been known to fail and they either fail open or shorted. When Shorted they would start the Cut Motor. With the Cover Open, the right side switch would prevent the motor from moving. Had someone bypassed the Right Side Switch and the Q1 Failed the Cut Motor "Could" Start as you were changing bits.... BY BYPASSING the Right Side Safety Switch and just press STOP you are putting your TRUST in the Q1 Transistor NEVER FAILING and turning the Cut Motor ON while you are changing the Bit.... It's like changing the Fan Belt on your Car while you Child plays in the Drivers Seat with the KEYS in the Ignition..... He or She would never turn on the Key.... RIGHT...

If you search out old posts around the time the pictures were taken shown below with the C1 Capacitor Snapped off one end you will find posts where the Cut Motor came on without even running a project.... cover closed, tap on top of machine, C1 bounced and Q1 Turned on... It was like using a paper clip to short over the transistor turning on the Cut Motor. I had 3 of my 4 machines have the C1 Snap.... That was WAY back in the QC Days when I ran a BAD QC long after it was BAD and vibrating not knowing... The Vibration broke the C1 Cap... They are glued now.

So with all that knowledge, you can make an informed decision as to if you want to bypass ANY safety switch. Knowing you could loose your fingers over it.

AL

Here are 3 old posts as "I" was learning the machine and also learning that a few of my "Guesses" were wrong.... The Cut Motor HAS been shown to be able to RUN BY ITSELF.... So beware bypassing any Safety Switches..... Here is proof from 2008.

Here is the first post in 2008 with the Cut Motor running by itself... http://forum.carvewright.com/showthread.php?7433-Cut-Motor-Runs-When-It-Wants-too-ALL-THE-TIME&highlight=Motor+running

And http://forum.carvewright.com/showthread.php?7462-Anatomy-of-the-Cut-Motor-X-Termination-Board&highlight=Motor+running

And months later AGAIN... http://forum.carvewright.com/showthread.php?7903-Cut-motor-still-running-at-bit-change-and-end-of-carve-in-home-position&highlight=Motor+running

Digitalwoodshop
08-14-2014, 06:47 PM
You will be told for safety sakes to leave them alone but I bypassed the two cover switches in the first month I owned my first machine. I just push stop if I want to stop the machine. Anyone who opens the cover and sticks their hands in while it is running deserves the ouch they will get. LOL.

Respectfully, It's your machine and your fingers. Just beware that you are putting all your trust in the C1 Cap and Q1 transistor never failing when you change bits with the right switch bypassed. See my 2008 post links in my other posting. Bumping the machine in 2008 caused the Cut Motor to turn on like the act of changing a bit.

Good Luck,

AL

GVigue
08-14-2014, 07:14 PM
The manual states that "At any point during operation the CarveWright machine can be stopped by pressing the STOP key or by lifting the cover." so people are going to use one of those two options. Considering if you accidently press the STOP key twice you will abort your project I think people tend to use the cover instead as I used to.

While there are several different opinions on how this cover safety mechanism works all agree that you should not rely on it functioning properly. That indicates a problem; a design issue, be it hardware or software, that should be addressed. Perhaps someone will take notice and look in to it.

DickB
08-14-2014, 07:38 PM
Interesting. Hard to believe no one makes a UL approved sealed switch.

dbfletcher
08-14-2014, 07:48 PM
I typically use the "push the stop" method.. but I will admit I have been burned several times by either my finger bouncing slightly or the button bouncing and aborting.... thats never fun.

lynnfrwd
08-14-2014, 08:40 PM
I panicked one time and hit stop twice. Lifted lid after that, but now know not to. We all are still learning.