PDA

View Full Version : Board Sensor, really stupid question



John
03-22-2008, 10:52 AM
I think LHR should do away it. All it does is confirm what it should all ready know. If an operator can figure out how to "build" a project with the software, surely they are capable of using a tape measure to tell the machine the correct starting size of the project. If the machine can find the top, left hand corner of the board it should be good to go. This could be easily accomplished by manually jogging the end of the board to a reference point, and then telling the machine "you're good to go." What am I missing here??

newcarver
03-22-2008, 11:02 AM
I have went through several board sensors and would have to agree, or else at least design a better one. It could be made like to the probe and simply jog to the specific points on the board.

Jeff_Birt
03-22-2008, 11:49 AM
The board sensor does way more than just measure the board. It also tells the machine where each edge is w.r.t. the axis encoders. BTW, they have recently been shipping SEALED board sensors. I think that is a step in the right direction.

John
03-22-2008, 12:10 PM
The top edge is always at the same place. Shouldn't be too hard to find where that is without a board sensor. Just a reference point for the Y axis would work.

Calibration, you ask? Could easily be done with a test pattern like a printer does, ie. which lines, line up, A, B, or C?

Jeff_Birt
03-22-2008, 12:27 PM
The top edge is always at the same place. Shouldn't be too hard to find where that is without a board sensor. Just a reference point for the Y axis would work.


?????Just watch the machine work...It finds each edge of the board. This not only tell the length both WHERE it board is W.R.T. the axis encoders. Otherwise it has no idea how big the board is or where the board is.

John
03-22-2008, 01:33 PM
?????Just watch the machine work...It finds each edge of the board. This not only tell the length both WHERE it board is W.R.T. the axis encoders. Otherwise it has no idea how big the board is or where the board is.

How can I make this any easier to understand?? For argument sake, lets say a board has 4 edges. A top, a bottom, a right and a left. The top will start (and hopefully remain) against the rail on the keyboard side of the machine. The left will start at a TBD (to be determined) reference point, which will be verified by the operator. Now, for the sake of argument, this board is 10" top to bottom, and 20" right to left. Is it too much to ask for the software and hardware to figure out where 10" from the rail on the keyboard side is? Same scenario, for the right and left. The TBD reference point will be X distance from the board position sensor so with a little addition and subtraction both right and left edges should be easy to locate.

Now, I certainly don't want to insult my own intelligence, I just want to have some fun with this and try to understand why a board sensor is necessary. I can't come up with a reason / situation where one is necessary.

Jeff_Birt
03-22-2008, 02:00 PM
John I see where you are going but I think you are still missing a few key points on what the board sensor does. I'm also not trying to be contentious, rather I'm trying to bring to light some of the benefits of having the board sensor. I've design automated machinery as part of my job and can say with confidence that machinery is quite stupid. Things that you and I might take for granted as being quite evident are not so from a machines point of view.

I use a small bench top mill at work that operates very much like you are saying. You home the mill clamp down your workpiece and then you have to move the mill to one corner of your workpiece and zero the mill. Then you can start cutting. Getting all three axis zeroed properly is not trivial.

When you start up the machine it has no idea where any axis is at. It will go through the homing routine which lets it know where the Y and Z axis are in relation to a fixed point on each axis. The X axis of course is the traction belts and has no reference to compare against.

Now when you put a board in the machine it will no where the Y and Z axis are in relation to a fixed point on each axis; but the machine will not know where those fixed points are in relation to the board. That might sound funny but due to manufacturing tolerances you can never know unless there is some method to calibrate the offsets and then that calibration needs to be regularly checked. Also you have no idea where is the X axis the board is or what size the board really is.

So, the CW machine starts to measure the board. It finds the top edge then the bottom edge; now it knows where they are w.r.t. the Y axis encoder reading. Then it measures the length of the board and knows where the board edges are in relation to the X axis encoder. When you put each bit in it checks each on to find its position relative to the Z-axis and relative to the top of the board. If you used fixed reference points as you describe then you open up lots of problems with operator error and continuously trying to calibrate the offset for the Y axis.

The side effects of having the machine measurement are that it can verify that you are using the correct size of stock, if you are using a different size stock then it can automatically scale a project for you.

I hope that all makes sense...

John
03-22-2008, 06:26 PM
Jeff, you have convinced me I could not be more correct. One sentence confirmed my thinking.

"I've design automated machinery as part of my job and can say with confidence that machinery is quite stupid."

I could not agree more. I bought a high priced hammer (yes, a quite simple automatic machine that is very stupid). I swing the hammer at the nail, the hammer hits the nail, and the nail automatically goes deeper into the wood. However, one of the first times I used that hammer it automatically just about smashed the end of my finger to smithereens. Now granted, I was a lot younger, and at the time didn't realize how stupid the machine was that I had just bought. Since that experience I have been much more careful how I use that simple machine.

So, how does this relate to the CW? (I'm glad you asked!) The CW is very stupid. It is a combination of belts, bearings, screws, leds, lcds, transistors, diodes, etc. etc. etc. Together they make a nice boat anchor. HOWEVER, the smart software developers at LHR can make it sing (so to speak). Just as the maker of the hammer I bought, had confidence I could program myself to hit the nail, they will have confidence that I can tell the CW the board size and load it with the top against the rail and one edge at a TBD reference point. Too simple. Not at all. I can think of even simpler methods, but will not bore you all with them. Thanks for listening, John

Begin Edit:

The above is to be regarded as a compliment to the LHR software development team. I have nothing but respect for them. Do I think they are looking for input from users to improve the software? I hope so. At least two readers have misunderstood this whole discussion. I hope this clears up my position on this matter.

Jeff_Birt
03-22-2008, 07:36 PM
The reference point is the problem John...

optionman
03-22-2008, 08:52 PM
the reference point should always be (0,0,0) bottom,right corner, or bottom left corner. from (0,0,0) if i tell the machine that my board is 10x14x1, i assume the bit can find any point on the surface from there on. Sorry for my 2 cents but that's my first impression.

newcarver
03-22-2008, 09:05 PM
The board sensor does way more than just measure the board. It also tells the machine where each edge is w.r.t. the axis encoders. BTW, they have recently been shipping SEALED board sensors. I think that is a step in the right direction.

That is a great step in the right direction

SandBuoy
03-23-2008, 01:55 AM
I can see both sides of the coin here. For us that have programmed NCNs before its a no brainer why we need the sensor that seems to cause a heck of a lot of problems for a lot of people. We sure didn't have them on our metal CNS in the 70s and 80s. It was the operators job to position in all the axis. With cutting oils, solvents and chips it would be a full time job to clean them all the time.

But on the other hand, I have to argue that the user friendly sensor makes it so easy for people that never ran a CNC machine before. They put the piece in, the sensor finds all the correct positions and make changes for mistakes that might have been made by the designer. Heck this machine is so easy to use that my wife can now run it. And she has no machine and very little wood tool experience at all.

I do agree that by eliminating that troublesome sensor it could solve a lot of problems for both the company and us. But the simplicity that it brings to the machine sure seems to out way the elimination of it.

But, I would also like to say that I can't see why this devise can't be made to perform for a HE!! of a lot longer then it does. All I can say is very poor quality. With that said, something else I would like to vent on is why better quality can't be brought into this machine.

For such a remarkable piece of equipment it sure has been hindered by poor quality, shoddy service and poor representation from Sears. Hey I looked into a couple of the other CNCs before and even after I bought this machine. I can't afford another $4-5000 bucks for a Cox or even Cox machine.

This machine may have its problems, but it will cut a piece 10' long while the machine 3-4 times its cost still only cut the table size pieces. The way I see it, if some of the faulty pieces were made to hold up for some period of time they could have charged $400-500 bucks more and couldn't keep up with demands. Its fast, does a fantastic job of carving and its only about 2" x 2" in size.

I was going to buy another one a few days after I got this one but now I am holding off to see if this one going to carve the next job. At this time I am getting a cut motor error within 2 hours of the last job. The only thing I can think of is I got the shop vac to close to a wire and sucked it loose or it snapped off.

The part that really stinks is, I just got 5 orders from the Indian head clock face and 4 lithopanes from a guy across the street that just lost his mother. I even got invited to a craft show after showing some pieces to the chairman running it that is usually booked 1-2 years in advance. I can get the work to pay for this machine but I can't get the work out.

Eagle Hollow
03-23-2008, 06:03 AM
Yup, So simple a cave man can do it!

badger
03-23-2008, 06:27 AM
I can get the work to pay for this machine but I can't get the work out.



Not sure how many times this has to be covered, but I think some people are expecting too much from this machine. They have already covered this in the warranty that it is NOT meant as a commercial machine but as a hobby machine. I do however agree that I believe that some of the parts could have been made a little stronger. You put a fast moving spindle like this on plastic wheels and parts and its bound to go bad sooner or later. I also would have paid a couple hundred more bucks for better parts.

If your expecting to use it to make money you better have enough for at least two machines to start with to keep the work going while one is down.

ChrisAlb
03-23-2008, 06:54 AM
I can see both sides of the coin here.

Just my take here but I think this whole issue is pretty ridiculous. Why on earth would you want to go back to a "manual" task when it's already done automatically and very accurately??

I mean, if this is just about all of you "showing your smarts" fine. But to me, technology is supposed to move forward, not backward. If the sensor is of poor quality (which to me it's not "that" bad), then improve the sensor. Don't eliminate it! After 160 hours cut time I never had a problem with the sensor (other than not keeping it CLEAN in the beginning). After 13 hours of a 15 hours carve mine coughed up it's diodes. Hey, If I worked that hard for 13 hours straight, I'd probably cough something up as well..lol


For such a remarkable piece of equipment it sure has been hindered by poor quality, shoddy service and poor representation from Sears.

The common denominator here is SEARS!! From LHR, My machine was NEVER misrepresented (it does exactly what I was told it would do). I find it be of very good quality and it performs very well so long as it's used and cared for properly. As for service, I can't speak to that on either Sears or LHR because I've NEVER had to send mine in due to how reliable it's been.

Now there are a few things I'd like to see improved a bit on the machine but it works exactly as advertised to me. The improvements I'd like to see most are in the "software". If you ask me, that's the weakest link in the chain.

Just my 2 cents.

Chris

ChrisAlb
03-23-2008, 07:09 AM
If your expecting to use it to make money you better have enough for at least two machines to start with to keep the work going while one is down.

If you just use and care for the "one" you have properly, it will do fine. I have only one. Never got interrupted in my work flow for more than a few minutes. I just sold my fourth big piece and with that, the "one" I have just paid for itself.

Now I'm not advocating that you can go stamp out a thousand plinth blocks a week or keep the thing running 24/7 but on average, mine runs "Carving" about 30 to 40 hours a week. No problem. An additional 10 to 20 hours scanning. Again, no problems.

optionman
03-23-2008, 11:01 AM
wouldn't it be relatively easy to have a manual mode, so that if i put in a 20x14x.75 piece and only wanted to drill/carve a small test area, i could tell the cw to ignore measuring and let me jog the bit to starting point (the upper left corner of my project, inverted and reversed to the lower right corner). then all the cw has to do is find the surface.

ChrisAlb
03-23-2008, 11:32 AM
wouldn't it be relatively easy to have a manual mode, so that if i put in a 20x14x.75 piece and only wanted to drill/carve a small test area, i could tell the cw to ignore measuring and let me jog the bit to starting point (the upper left corner of my project, inverted and reversed to the lower right corner). then all the cw has to do is find the surface.

You can. Select "jog to start postion" when given the choice of centering or not.

ChrisAlb
03-23-2008, 01:09 PM
Why then don't you just design a nice "manual" system for youself and not worry about the "automatic" one already in place, since it's seem to give "you" so much trouble???

Or better yet, call LHR and give them your words of wisdom. Because all this whining is helping no one in here. Jeezzz

John
03-23-2008, 02:03 PM
Please see my edit to post #8 in this thread.