Henrik Olofgors
Thu 22 February 2007, 12:51
Hello all,
First, let me thank you for this truly excellent resource. I?m a novice in most fields involved in the magic that happens here, but I?m eagerly soaking it all up.
Venturing out trying to build a mechmate is far beyond my horizon, but I have some ideas for things involving many of the parts included in the overall concept. I hope I?m not overstepping the bounds of the forum in presenting them, if so please delete this.
To concretize with a specific case, my Scheppach circular saw could do with some mods. Here?s an image (http://members.arstechnica.com/x/hiphink/table-saw-v1.jpg) of what I have silently dreamed of in my chamber.
Background;
Several times a month we do major sawing operations for cabinet-style work, often involving dozens of differently sized parts on many dozen sheets.
We use panel-optimizing software to print cut sheets/lists. While this is a terrific aid, managing a good workflow strategy is a headache because we can rarely fit all pieces of one singular dimension into one cut operation. We try our best but we usually have to switch back and forth between fence settings due to the way parts are nested for minimizing scrap, which is a problem because parts that are supposed to be equal often are not: parts supposedly equal in size but cut in different sequences differ ever so sligthly, enough to cause problems in subsequent stages of production. If we did melamine or even veneered mdf, we could get around most of this by accepting 15-20% scrap instead of our current 2-3%, but for most of our jobs this is not an option due to the panels being high end and often costing more than $100/m2. As such, I have thought a lot about some kind of nc/cnc fence modification to our current saw, nothing fancy but being able to quickly change between settings and being right on the mark every time. A brand new Altendorf would be nice, but I just don?t have that money, and if I did I?d buy a cnc router anyway.
I?m not looking for a complex integrated cnc solution that downloads code and basically tells the operator ?load panel type x, do y cuts, wait for input? etc, just a simple numerical control fence with decent precision and great rigidity.
Preliminary considerations;
The basic idea is to mimic (and possibly even improve) the nc controlled fences on semi-high-end rip saws. The operator inputs the desired distance on a keypad, the fence positions itself, locks, and off you saw. Repeat.
Overall precision need not be better than 0.2mm, though higher precision would of course be nice.
Since different saw blades will be used, thus affecting the X0 position, it must be easy to zero the unit.
The fence must be very stiff when under load. This is, I assume, the biggest problem of the overall design. Our current Scheppach fence is locked in place with a screw, and has less than +/- 0.2mm flex at the point furthest away from the lock. Speed is not a concern. Locking power is.
Due to available distance in the x axis being a precious commodity, it is desirable that the gantry be as limited in size along this axis as possible.
One thing I have not seen on high-end saws with nc fences (though it may well exist and even be common) is the ability to control the Y position of the fence relative to the gantry. Precision is not necessary; a resolution counted in centimeters would be fine.
Discussion;
If undertaken, the resources dedicated in time and money must be reasonable. I am not approaching this as a must-have capability: if so, I?d bite the bullet and go to the bank, cap in hand. Rather, I see all the absolutely amazing systems people here are building, and I can?t help but wonder if something I think of as far simpler and more bland could be possible for even an electronics/etc dunce as me to achieve.
While I?m saving up to buy a shopbot, there are so many things in my current machine park that could benefit from a little nc/cnc love, it makes my head spin. It?s never going to achieve the open ended possibilities of a good cnc router system, that?s for sure, but if it could be done for moderate cost and money, even as that adds up, I can easily see ten to twenty scenarios on the current machines which if nc controlled could --all combined-- achieve the same rationalization grade as a single cnc machine would, across the totality of operations we do.
I am going to get me that shopbot, so help me God. But when I do, if my table saw and my shapers, my line borer, etc, all suck spectacularly, I?m just going to be producing around them. They are good machines, and with the right modifications they could be so much more easy to work with.
Anyway, I?m blathering. If I could hope for anything, please give me your ideas on the general feasibility of this.
First, let me thank you for this truly excellent resource. I?m a novice in most fields involved in the magic that happens here, but I?m eagerly soaking it all up.
Venturing out trying to build a mechmate is far beyond my horizon, but I have some ideas for things involving many of the parts included in the overall concept. I hope I?m not overstepping the bounds of the forum in presenting them, if so please delete this.
To concretize with a specific case, my Scheppach circular saw could do with some mods. Here?s an image (http://members.arstechnica.com/x/hiphink/table-saw-v1.jpg) of what I have silently dreamed of in my chamber.
Background;
Several times a month we do major sawing operations for cabinet-style work, often involving dozens of differently sized parts on many dozen sheets.
We use panel-optimizing software to print cut sheets/lists. While this is a terrific aid, managing a good workflow strategy is a headache because we can rarely fit all pieces of one singular dimension into one cut operation. We try our best but we usually have to switch back and forth between fence settings due to the way parts are nested for minimizing scrap, which is a problem because parts that are supposed to be equal often are not: parts supposedly equal in size but cut in different sequences differ ever so sligthly, enough to cause problems in subsequent stages of production. If we did melamine or even veneered mdf, we could get around most of this by accepting 15-20% scrap instead of our current 2-3%, but for most of our jobs this is not an option due to the panels being high end and often costing more than $100/m2. As such, I have thought a lot about some kind of nc/cnc fence modification to our current saw, nothing fancy but being able to quickly change between settings and being right on the mark every time. A brand new Altendorf would be nice, but I just don?t have that money, and if I did I?d buy a cnc router anyway.
I?m not looking for a complex integrated cnc solution that downloads code and basically tells the operator ?load panel type x, do y cuts, wait for input? etc, just a simple numerical control fence with decent precision and great rigidity.
Preliminary considerations;
The basic idea is to mimic (and possibly even improve) the nc controlled fences on semi-high-end rip saws. The operator inputs the desired distance on a keypad, the fence positions itself, locks, and off you saw. Repeat.
Overall precision need not be better than 0.2mm, though higher precision would of course be nice.
Since different saw blades will be used, thus affecting the X0 position, it must be easy to zero the unit.
The fence must be very stiff when under load. This is, I assume, the biggest problem of the overall design. Our current Scheppach fence is locked in place with a screw, and has less than +/- 0.2mm flex at the point furthest away from the lock. Speed is not a concern. Locking power is.
Due to available distance in the x axis being a precious commodity, it is desirable that the gantry be as limited in size along this axis as possible.
One thing I have not seen on high-end saws with nc fences (though it may well exist and even be common) is the ability to control the Y position of the fence relative to the gantry. Precision is not necessary; a resolution counted in centimeters would be fine.
Discussion;
If undertaken, the resources dedicated in time and money must be reasonable. I am not approaching this as a must-have capability: if so, I?d bite the bullet and go to the bank, cap in hand. Rather, I see all the absolutely amazing systems people here are building, and I can?t help but wonder if something I think of as far simpler and more bland could be possible for even an electronics/etc dunce as me to achieve.
While I?m saving up to buy a shopbot, there are so many things in my current machine park that could benefit from a little nc/cnc love, it makes my head spin. It?s never going to achieve the open ended possibilities of a good cnc router system, that?s for sure, but if it could be done for moderate cost and money, even as that adds up, I can easily see ten to twenty scenarios on the current machines which if nc controlled could --all combined-- achieve the same rationalization grade as a single cnc machine would, across the totality of operations we do.
I am going to get me that shopbot, so help me God. But when I do, if my table saw and my shapers, my line borer, etc, all suck spectacularly, I?m just going to be producing around them. They are good machines, and with the right modifications they could be so much more easy to work with.
Anyway, I?m blathering. If I could hope for anything, please give me your ideas on the general feasibility of this.