CNC Router Accuracy

[arthur_ross]

Since nobody's posted lately, it's time to stir the pot.... :-)

How accurate are these systems, really? (In this I mean a general discussion of all CNC routers, not just ShopBot), And do we really care? I mainly sell the Signs On Tap CNC router system to signmakers, where "If the result looks good, it is good". So I can't say I'm worried about hyper-accuracy.

It's easy enough to test for repeatability, coming back to a fixed point on the bed, but without a truck full of laser measuring equipment, are there any easy ways to measure accuracy in all 3 axes?

What kind of accuracy and repeatability do folks have (or think they have) with their machines, and what accuracy is required for the work they do with their machines?

Always thinking-

Arthur Ross
Deltamation Inc./www.signsontap.com

[scottncs]

Arthur - I may be the exception rather than the rule, but I want as much accuracy as possible. I am making small placques with line drawings and text, customized picture frames, and nameplates. I sometimes need to go down to 10 points for the text. I have a Porter Cable 1.5 HP router and have attached a Dremel as a sidecar for really small lines (down to 1/32").
I have the PR32 with the screw drive and it works fine for me. After cutting a couple of patterns that can take up to 30 minutes each, I recheck the 0,0 position. It is sometimes off by 0.04", which I attribute to all of the abrupt starts and stops of the motors as they cut the complicated paths.
As far as the accuracy question - it reminds me of questions I get all of the time regarding accuracy and precision, and the analogy to shooting at a bulls-eye. If your pattern is tight, you have good precision; if the center of your pattern is the center of the bulls-eye, you have good accuracy; but you can have one without the other.
Maybe you could do a similar test- set your "target" some distance away from 0,0 and have the SB cut a small dimple in a test board there. Perform this test 10 times. Also measure the target location manually. See how tight your pattern is (precision; also called repeatability) and how close the center of the pattern is to its true location (accuracy). You could even measure Z by using a v-bit and checking variation in the diameter of the dimple.
What do you think?

Scott Smith
Smith LetterWorks

[bruce_clark]

Arthur:

Scott is right. Accuracy and repeatability are
two separate issues. The best way that I know how
to check repeatability would be to drill holes
with a center cutting endmill in something like Al
plate (something that will not warp/swell or other
things that wood can do).

Now, tell the machine to drill say every .5" or
so. By using an end mill, you a) take away drill
flexing/wandering and b) endmills are chucked
up REAL close to the spindle bearings, making the
end mill more rigid.

Now, drill your pattern of holes (similar to what
Scott suggested) but now put in dowl pins (the
size of the endmill)--it should look like a peg
board) and then measure from dowl to dowl. (dowls
are made to VERY high precision--usually .0001"
tolerances). Just make sure the dowls sit in the
holes tight (to negate any inconsistences on the
holes drilled with the endmill).

Measure from one hole to the next and from 1st
hole to say one 12" away or so (yes, you need a
12" caliper). Now, this only gives you positional
accuracy. Run the test one a few other plates and
compare results and this will give you positional
repeatability.

Now, Mr. Smith, you are probably getting .04"
error from the backlash in the leadscrews. These
screws are usually right on (+/-.002 pitch error
per foot) but if the machine does not compensate
for the backlash, after a 100 direction changes,
you can theoretically be up to .2" off. Now, SB
has backlash compensation in the software, your
settings might need to be tweeked a little.

The other possibility is that somewhere, you are
getting some kickback and the kickback is more
than the motors can hold against. This is less
likely, but still possible.

There are other possibilities too.

Bruce Clark
bwclark@centuryinter.net (mailto:bwclark@centuryinter.net)

[and-how@uniserve.com]

Thanks for this analysis. This is the kind of thing that a new, potential shopbotter needs to know but perhaps doesn't know enough to ask. So please keep talking and let me listen in! What other issues may be important?

[bruce_clark]

Mr Anderson:

There are a lot of other important issues. First,
ask yourself: "What do I want to do with my/a CNC
router?" If you are mostly doing signage or
general roughing of material, them accuracy may
not really be an issue, per say.

For a first time user, I would be more interested
in things like does this machine work with my
drawing/sign/cad program? It the machine easy to
use and easy to learn? Will it machine the kind
of materials I want to use? If I depend on it to
it to put food on the table AND it breaks down,
how long to get it working again?

I think these are VERY relavent questions and
should carry as much weight as "what is the
accuracy of the machine".

Now, personally, I feel that ShopBot has addressed
ALL the issues above. The machine is EASY to use
(maybe a little time consuming to assemble, but I
did get it together) and because I put it
together, I will not only know WHY something has
broken/fails but I will know HOW to fix it. Parts
are available from locale hardware stores or from
Shopbot.

The software is easy to use and with the recent
slate of file converters, now works with _almost_
ANY program, sign, cad or otherwise.

Bruce Clark
bwclark@centuryinter.net (mailto:bwclark@centuryinter.net)

[rickb]

Bruce,

The dowel & endmill method is good, but re: 100 direction changes in a system w/ .002". Backlash in a geared system does not accumululate to 0.2". Backlash is non-cumulative. Slippage is if it is allowed in the design. Backlash is the "slop in the gearing". Are not precision leadscrew tolerances specified as +- .00n" per foot (local positional certainty) and +- .00n" per foot overall (cumulative but fixed)? Right? (Thermal expansion is another issue if you want to get picky or have a very long x-axis with large temperature swings.)

[arthur_ross]

I agree with Bruce on having a good understanding of what and how you want to use the machine. Collect samples of the materials you want to cut and get test cuts made on the various CNC machines that you're looking at. Make test files on the software you are using or planning to use, and make sure the machine and file formats are compatible (don't rely on specs alone).

I'd strongly agree with Bruce on the "food on the table" issue, but with a slightly different conclusion. If the output of the CNC machine provides all or a large part of your income, I'd buy or rent the best you can afford, with the least down time and the best/fastest support/repair. Granted, it's cheaper to build and repair it yourself, but do a real calculation of what your production time is worth, and what cost/volume of product you are losing while you do your own construction or repair work; and the kinds of tools and skills you might need to have to do that maintenance. This is not a slam on ShopBot, but the sort of calculation you should do before you aquire any automation. For instance, if you are making about $10/hr with the parts you make, it would be crazy to pay a tech $50/hr to repair your system, (unless he could do in hours what would take you days). On the other hand, if you are making $100/hr on fine cabinetry, then it pays to buy an expensive machine with instant field service support, because any downtime is quite "expensive".

However, there's no need to buy the system that will "take care of your needs forever". In this market, it's easy to start with a unit for your current needs, and then trade up. People are always looking for used machines, and I'd guess that a built & adjusted ShopBot would sell quite well.

There is, of course, the danger of buying too much machine without knowing that you will have the volume/pricing to support it. I've heard several tales of woe from sign shops who bought another CNC router system (think baby food) and the payments drove them out of business, because they were not set up to generate the volume of sales required to support the cost of ownership.

To end this rant, "the machine is supposed to support you, not the other way around."


Arthur
Deltamation Inc./www.signsontap.com

[rgbrown@itexas.net]

I came up with a different way to check the X-Y accuracy that requires no drilling or accurate measuring, assuming one has a CADD program they are familiar with. Build an irregular polygon of stretched music wire, hook this to your Z-plate; Run plot; move polygon to another area of the table; do same. Transport file to CADD. Compare plots and points; the truth will be evident.

Ron Brown - rgbrown@itexas.net (mailto:rgbrown@itexas.net)

[and-how@uniserve.com]

Many thanks for additional info. I appreciate it!

[WOODPUTZ@AOL.COM]

I'm in the process of starting my own business, and am in the market for a CNC router system. The ShopBot system has gotten my interest piqued. I've got a couple of questions though. Let's say I want to cut 3/4in. Ash or Oak or some other similar hardwood. Can I cut through in 1 pass? I plan on using a Porter Cable 3.25hp router. The machine I currently use at work can handle it just fine, but then again it's a $35,000 system.

[BoscoToys@thebest.net]

I've cut 3/4 ash in one pass and it seemed to work fine. When I did I had the router going at its top speed, which I now know is not ideal. So the cutting tool life might be shorter, since you can't feed the tool through the material at the ideal rate.

[WOODPUTZ@AOL.COM]

David,thanks for the reply.I've got another question for you. When you cut the 3/4 Ash,did you get noticeable chatter marks, or was it fairly clean? I plan on cutting some fairly intricate parts like gears,pistons and other similar parts. Also, did you use HSS bits or carbide? I know when we use 1/4" carbide bits at work, if the feed rate is too fast they tend to snap after a fairly short time. O.k., I lied, 1 more question. Did you use 1/4" shanked bits or 1/2" shank.

[BoscoToys@thebest.net]

Larry, I was cutting some complicated polylines that had to line up with detailed engravings previously done. I used Onsrud's 3 edge finisher. It was a 1/4 shank bit. I don't remember exactly, but I think I was running it at .5 or .75 iches per second. On Onsrud's speed chart it says that more bits are broken and dulled from going too slow rather then going too fast. I'm fasinated by this, but I'm a little skeptical. I don't know that I want to test the theory with $40.00-$70.00 cutters. It seems like spindle speed, and the number of cutting edges effects the feed rate. They claim that with a single edge cutter you have to increase the feed rate, I don't understand this. They also give a bunch of formula's for calculating chip load's feed rates and RPM's. If I was laying out my parts for the first time I would set up for using a 1/2 diameter bit. I think the breakage must be much less with this. It would sure be nice if ShopBot had some sort of charts for feed rates and spindle speeds. It would take a lot of guess work out of the process. Onrud has a Router Laboratory where they will test your layout with there machines and recomend a bit and feed rate. I havn't asked but I dought that they would do a ShopBot setup. There flyer says the service is free. I hope this helps you, as you can see I'm still learning. Oh one last thing the 3 edge finisher had an excelent cut w/ almost no sanding needed. It did leave a lot of hair on the top edge though. My next test will be with a compresion bit which cuts down on top and up on the bottom. It's supose to eliminate the hair.

[WOODPUTZ@AOL.COM]

David, again thanks for the feedback.I'm with you,I'm skeptical of Onsruds speed chart. In my experience, if your feed rate is too fast and your cuts are too intricate, very soon you will break a 1/4 shank full carbide bit. I say this because I have about 5 broken $25.00 bits by going too fast.HSS bits on the other hand handle a little excess speed fairly well.If you are going too fast, the quality of your cut will tell you.Also, your router will tell you by the sounds of it laboring.Once you have burnt out a $300.00 router by not listening to it that should be a lesson learned.Also, it seems to me that if you only have one cutter on your bit, common sense says that you have to slow down your feed rate as the one flute can only remove so much material at a time.As to your problem with hair on the top edge of your material, I found that using a spiral upcut works fairly well. You might also want to check your hold downs. The material might be riding up on the bit.You didn't say if you use a vacumn system or mechanical hold downs.

[BoscoToys@thebest.net]

I've been working on a vacuum system, but all that I've cut has been held with screws. I've used spiral upcut and I get the hair. It would seem like the down cutting action would drive the hair into the cutter to be cliped off. That is what the compression bit is supose to do. Maybe if I just used a down cutting spiral I could save the money since the compression bit is 4x the cost.

[larry]

David, what length bit are you using when you cut?
We use a 2" bit and the only time we get fuzz is when the bit either goes too deep so that the material is being ripped and not cut by the top of the flute were it flows into the shaft. Also, sometimes the screws are not holding down the material flat enough so that the material rides up the bit. When everything is held down snug and the bit is set to the right depth, we get a perfect fuzz free cut providing the bit is still fairly sharp. As to the down cut spiral, we found that it doesn't remove the chips and dust enough and that sometimes it can force the material up because debris gets underneath the material. Also, if you use a HSS down cutting bit it can get hotter because the sawdust is not being removed from around the bit. That can shorten the life of the bit. This is just some of the things that we discovered on our system. My boss also said that Onsrud is right about their statements concerning feed rate. He also has their bible on his shelf. I have to admit that sometimes he is right considering that he has about 10 years more experience on a CNC than I do. That's 1 of the reasons why I'm starting my own business, so I can be right all the time.

[BoscoToys@thebest.net]

Thanks for the info. I guess those down cutters have their own disavantages. The cutting lenght was 7/8" and I was cutting into 1/2 birch plywood.
Maybe the compression bit would be better since the bottom edge is bringing the chip up. I'm guessing it wouldn't go under the part. Unfortunatly they are $40.00 a piece.

[arnold@koyote.com]

want to buy 2 or 3 used shopbots
dave at tanyard springs
arnold@koyote.con (mailto:arnold@koyote.con)

[arnold@koyote.com]

dose any one use the shopbot to cut gunstocks???

[BoscoToys@thebest.net]

I called Onsrud again and it would seem that the straight bit might be the best option for my problems. Easy to resharpen, much less likely to produce hair longer lasting, and much less costly then the compression bit. I'm still concerned about what others have said about spirals being better b/c of constant contact with the cutting edge. Anyone out their using straight cutters?

Hurray for ShopBot Hurray for all their great science and emotion and Help!!!!!!!!!!!!!
Best company I've ever ever delt with, some times I feel guilty because they have been so generous with this shopbotter. Way above and beyond their warranty.

[parlee@island.net]

I've got half a dozen each of 1/2" spirals (up cut only) and 1/2" straight cutters.

The spirals will sometimes leave a slight "fuzz" on the top edge of the cut. Chip clearance and cutter longevity is excellent and they plunge well. Be very careful to have these sharpened at a quality sharpening service shop (I learned this the hard way).

The straights are cheaper, easier to sharpen, also leave a good edge (top and bottom) when sharp. I prefer the "straights" with a slight angle of around 5 degrees which improves chip clearance. Generally, most of these bits don't plunge as well as spirals if at all. They also don't seem to hold an edge as long. I've tried both cheap and "expensive" straight bits. I definitely recommend the "expensive" ones as these work out less expensive in the long run. There was a fantastic article in Fine Woodworking magazine last summer reviewing 1/2" straight bits in CNC tables with surprising results. Well worth finding a copy. The number one bit in the review is available from Routerbits.com item no. 1069 for $13.46 (these don't have the 5 degree angle).

It has been my experience that when cutting solid wood especially, chip clearance is critical to bit longevity. If you don't have effective dust collection, you might try hooking up a compressed airline directed at your bit (cools bit and clears chips). There seems to be quite a science behind rotation speed and feedrate. I generally try to go as fast as I can while maintaining accuracy. If you are getting burn marks, you are going too slow or your bit is dull.

Most of the above I've learned from a recent job of 748 ornate western maple table legs. The straight bits couldn't clear the chips which took me to the spirals.

[sheldon@dingwallguitars.com]

Is there a way of knowing when you are driving the steppers too hard and are at risk of missing steps? Would a load or amp meter help in determining maximum feed rate including a little headroom?

[bwclark@centurytel.net]

Sheldon,

Unfortuantely, there is not a way to tell when a
stepper motor is about to miss a step. The
standard way to set up stepper drives is to have
the machine move as fast as it can RELIABLY with
out missing a single step. Then set the machine's
top speed to 70% of that setting. This way you
have a 30% "reserve" when loading varies on the
stepper motor.

The other route is to go with a closed loop
control like ShopBot used on the cable drive. An
encoder tells you when the machine missed a step
and the controller/software can "add" that missed
step back to get everything back on track again.

Hope this helps,

Bruce Clark
bwclark@centurytel.net (mailto:bwclark@centurytel.net)

[CONOR HAUGH]

We have been considering buying
a custom size shopbot 100in x 60in. All discussion
about accuracy / performance etc on the furum have related to prt96 Anyone out there with a shopbot around this size.We have a welding shop
We could build a heavy duty table for the machine
If this helped accuracy/ repeatability.

[mechtron@iafrica.com]

Connor, send me your e-mail address and I think that I can give you the type of info you need. My address mechtron@iafrica.com (mailto:mechtron@iafrica.com)

[ron_cleaver]

I've been doing a variety of tests after assembling my R&P tool.

After doing some calibration, I started with a simple cut. I cut a piece of paneling using the following (with z=0 on the surface of the paneling):

mz 2
my 0
jx 90
mz -.25
my 12

I measured the length of the cut part and it was 90" at y=0 and at y=12 (of course within the limits of a measuring tape and the human eye). I then tried wider cuts and all was well, except...

One problem I've been having is a vibration for a few inches at two places along the x-axis (i.e., only x-axis moving at the default move or jog speed).

To isolate the vibration I wrote the following to move the tool back and forth along the x axis:

mz 4
my 0
loop:
jx 0
jx 96
goto loop

I found some loose bolts as a result of this, but the vibration remains, both at move and jog speeds. I made sure every bolt is tight.

But I got distracted. After about 10 minutes, the tool got hung up at the hard stop at maximum x. I stopped it and looked at the screen, which indicated x=~82. How can I get this much error in such a short time.

I entered the configuration parameters for the motors when I assembled the tool, so that's not the problem. The x axes are parallel with equal diagonals.

Any ideas?

[Ted Hall, ShopBot Tools]

Hi Ron,

Can't be a Unit Value issue because errors in these would produce the same error going up and back the X; and thus you wouldn't go further and further out of position.

So what it sounds like is that where you are getting the vibration you are also losing steps and thus accumulating an error. The error must be worse in one direction than the other ... probably when you are jogging.

That would suggest that 'vibration' is more than vibration ... and maybe we are being misled about this being a mechanical issue ...

What are your speed and ramp values??

Better yet, send me your PROBLEM.LOG file ... or post it here.

[jkforney]

Ron & Ted
I had a similar problem which directly related to the fact that I had my turn buckles were way to tight on an ungreased x and y axis. When I loosened the turn buckles and greased the axis (all of them) my machine stopped the random stutter (It was approximately the same direction and about the same place but not exactly.) What I think was happening was the axis was building up a bind and then it would lose steps. On other machines I have worked with they all seem to have the motors too tight.

Can't say this is your problem but it's worth a try.

[ron_cleaver]

John,

I'll try loosening the turn buckles. But I already greased the two racks and pinions on the x-axis (as well the the ones on the y- and z-axis too).

I'm using default speeds for all tests so far. I don't have a similar problem on the y-axis, and I believe the tension on all turnbuckles (both x- and y-axes)is about the same.

Because the x-axis has two motors, I'm wondering if the tool requires the turnbuckle tension to be equal on both sides within a small tolerance.

Ted,

I'll send a problem log as soon as I get a chance.

[Mayo]

I had a similar problem with the chattering/stuttering of the movement only in the -x direction.

On investigating, I discovered one of the turnbuckles on the X axis was too loose.
Pressing down on the stepper motor, I could actually move it downwards fairly easily. The other X axis side was not as loose. I turned the turnbuckle one-half turn tighter and that's all it needed.

[ron_cleaver]

The problem for me seems to be the newer motors. But I changed the default move and jog speeds after talking to Ted, and the problem went away.

BTW, the problem with the tool accumulating error was simple to fix. Ted suggested removing the CD-ROM driver from the boot files. No more problem!

[Gerald D]

The spring/turnbuckle system is not ideal. We really need a softer spring that can use more of the turnbuckle's range of adjustment. If anyone can give me the load (lbs force) required at the pinion, I offer to design/source a more suitable spring system.

What is the maximum side load allowed on the motor neck bearing?

[tony jenkinski]

remove the cd-rom driver from the boot files what is that and how do you do it. what will it do?? thanks

[ron_cleaver]

Tony,

CD-ROM drives, when used with DOS or Windows 95, required (usudally) a line in the config.sys file to load the driver, and a line in the autoexec.bat, usually
mscdex
to map the CD-ROM drive to a logical disk drive.

Apparently many CD-ROM drivers interfere with the ShopBot motor operation. I'm using an old 486 laptop with DOS to run the ShopBot. When I removed a line of code from the two files, I stopped having problems with the ShopBot (see my post above).

[choose01@msn.com]

Ron,

We have been having some problem with a stuttering as well, become really accentuated with a radius cut. What did you do to eliminate the stutter? We are running in DOS with a CD-Rom, should that be eliminated? You mentioned a change in the default speeds, did you slow them down, to what? Most of our files we are trying to cut at 2IPS, the router is not under strain, and for the most part we will not have a problem with the stutter. Any speeds over that and the problem occurs. The stutter is always there when cutting or routing a radius cut. Any thoughts?

Chuck

[ron_cleaver]

Chuck,

Removing the CD-ROM drom the boot files and reducing the default x and y move speeds (to 1.5 ips if I remember correctly) eliminated all my problems. Even a circle cut works now. (I think I also reduced the default x and y jog speeds but I don't remember the value)

You might want to create a boot disk without the CD-ROM, so you can still use the CD-ROM drive when not using the ShopBot. That's what I did.

I got my ShopBot last March (2001). That was about the time they started using newer, more accurate motors that require changing the default speeds. I think there was a post from someone at ShopBot about that a few months ago.

Dave or Ted - anything to add?

[sea_nc@bellsouth.net]

Ron and Chuck,

Removing a CD-Rom driver from autoexec and config.sys files can solve connection problems and memory problems, that's for sure. As for solving stuttering problems, my guess is that the change in speeds was responsible for minimized the stuttering problem. How are you generating the files that are stuttering? Depending on the cause of the stuttering, the stuttering problem may or may not be solved. For example, rapidly cutting a file such as sloppy clipart with lots of overlapping/reversing lines and arcs can cause the machine to shake and stutter as the machine moves where it is told to move. Slowing down this same file would minimize, but not eliminate, the problem. Similarly, an axis that is binding will balk at moving rapidly but will move fine at lower speed.

Dave

[Chuck Hoose]

Dave,

What is the best way to determine if an axis is binding? With out the motors engaged the x and y axis slide just fine and do not seem to be having any problems. Is there a way to diagnois if the motors are operating properly, if for example on of the motors on the x-axis was weaker or had a problem, could this cause the stuttering? Ron mentioned that you have upgraded the motors to newer and more accurate ones, any info on this, whats new and better about them, are ours upgradeable?

Chuck

[sea_nc@bellsouth.net]

Chuck,

If the Y car is binding, tug on the car with the motor engaged and SB software running. Since the motor is locked into position with software on, any play that you detect in Y car may be due to improperly adjusted holddowns. Readjust holddowns to eliminate binding.

If the car moves smoothly without binding, then axis binding is likely due a combo of turnbuckle tension and how tight the motor mounting bolt is tightened.

Motors are very reliable and not likely to be "weak" unless driver is bad. You can swap motors or drivers to troubleshoot this, but I'd thoroughly check out turnbuckle/motor bolt tension first because it is a common cause of binding.

Stepper motors have more power at lower speeds than at higher speeds, so a "binding" that can be overcome at low speeds may stop tool at higher speeds. Over time, motors have been used with different gear ratios and pinion sizes resulting in tradeoffs in speed and power, but I wouldn't say that any one combination is better than another.

Dave

[gerald_d]

David, what is the latest factory instruction for adjusting the turnbuckle tension? Can you quantify the load on the pinion gear as **lbs force? If so, I would throw out the turnbuckle and fit the correct value tension spring.

[flyboy]

I seem to be having the same problems. When I first set ti up it was working fine, I broke some bits so I changed the speed values and now it screws up every file I try to run. How do I make a boot up disk without the cd-rom? Forgive me but I've never had to deal with that sort of thing. I was running the old software - I bout mine in March.

[ron_cleaver]

Tom,

To set up a boot disk without the CD-ROM driver:

1) Create a system disk (i.e., one formatted with the /s option), or use one you alreay have. If they are not on the disk, copy config.sys and autoexec.bat from the root directory of your hard drive (usually c
.

2) In config.sys look for a line that starts with 'device ='. If there's only one line like that, chances are it's the CD-ROM driver. Delete the line and save the file. If there's more than one line like that, then you need to know the name of the CD-ROM driver. It's a file name that ends with '.sys'. Remove the appropriate line from config.sys and save the changed file on the disk.

3) In the autoexec.bat file, remove the line with 'mscdex' in it and save the changed file on the disk. That enables the Microsoft CD extensions.

4) Reboot.

[Mayo]

Tom
If things are screwed up after changing speed and ramp settings, make sure all your ramp speeds are about 60% of your move speeds. I experienced a lot of chattering and general havoc when I changed speeds without changing ramp values properly.

[Ted Hall, ShopBot]

This thread sort of slipped up on me since it was added to an old one. But let me try and help out here.

First, a couple of thoughts about trouble shooting. Usually, when you're having a problem that is not solved easily, there are probably a couple of things going on ... not just one. That's why it can be sometimes difficult ... expect to have to tweak a couple of things.

==>On the CD-ROM driver issuse:
As Ron points out, DOS may be starting this driver when your Win95/98 computer boots up. The driver is started in by the CONFIG.SYS and AUTOEXEC.BAT files which are automatically run when the computer starts. This driver does not need to be started because Windows has its own CD-ROM driver. It is only there in case you are running DOS games that require a CD. It -does- interfere with ShopBot by competing for what called an interrupt line. This driver as well as mouse drivers (also not needed by Windows), DOS network drivers, and other DOS drivers can cause connection problems, uneven running, and memomory problems (as Dave pointed out).

Dealing with this problem just means cleaning out of your AUTOEXEC.BAT and CONFIG.SYS the start-up of any offending driver. Open the file with an editor and put a REM at the beginning of the line. This will disable it. Anything with MSCDEX in it is related to the CD-ROM. (See Ron's notes above ... and for more detail, see the "Working with Your Computer" section of the 'ShopBot Users Guide'.) I don't think you will really need a boot disk because for most purposes (except DOS games) your computer will run fine without these drivers and you can just remove them from the primary boot files on the C: drive.

==>Chuck's Problem with Fast Arcs:
Chuck problem is probably not related to the driver issue. What he is getting is uneven running when executing circles and arcs at fast speeds (above 1.75 in sec). Here is what causes the problem. When ShopBot is cutting anykind of curve, the vector speed of the motors is constantly changing. Because the speed is actually managed by the control box (to keep Windows from interfering with the timing). This means that there is a large amount of information being passed to the control box about speed, as well as the usual step information ... sometimes almost twice as much. While the speed of serial communication does not come into play for a straight jog until about 5 in sec (at standard gearing), in a curve the speed of the serial communication and the computers speed becomes limiting at a much lower move speed. Thus, the fastest smooth speed you can usually get with curves is about 2 in sec (with standard gearing). This can be a little lower or higher, depending on computer processor speed, and can be modified by changing the gearing ratio (with other trade offs). The speed limit imposed on curved paths by serial communications is one reason we list a lower cut/move speed than jog speed. The trick is to do a few circles and see what the limit is for your computer. If it is much below 2, then it is probably computer speed that is limiting and not communication speed. In this case, running in DOS may get you more curve speed.

==>Tom:
I think your problem sounds a little different. Maybe you could describe it with a bit more detail and we can get it figured out. If you think the problem may be related to speeds or ramps. Then, I'd just: 1) make sure you have your Unit Values at hand; 2) delete the SHOPBOT.INI file from your SB200 folder; 3) restart the ShopBot software; 4) enter your Unit Values with [VU] if needed. This will give you a fresh start with all the speed and ramp values.

==>In general for PRT's purchased last year:
We did not include grease with the earliest R&P tools that we sent out. But the pinions do run smoother if you grease the rack with any good bearing grease (or clear greases like SuperLube). Debris accumulation does not seem to be a problem except for certain types of platic chips.

==>Ramps ...
(FOR PRT's) I've become pretty happy with just setting the X & Y Move and Jog ramps to .9 and not worrying much more about it. Also, if you are moving below 1 in sec, you don't need to be bothered with ramps. Of course, you can do a lot of further tweaking with ramps for your own purposes ...