Gerald,

I am assuming that I have a flexing Spider or do you think I am on the wrong path again?

I have looked at the two thread that I could find with regards to flexing spiders.

http://www.mechmate.com/forums/showt...flexing+spider

Beefing up a thin Spider that Heath posted.

I don't believe I have the right equipment to check the flex in the spider but I believe it was flexing.

I am wondering if the larger bushing that I made, 1" diameter, .75" long, to offset the VWheels so that I could have enough space for the nut/bolt for the quick release plate and offset the Z Slide, are actually creating a more torque on the Spider.

I even replaced my new bushings with the smaller ones I had been using, .375" Diameter, .4375" long to see if it would stop the flex but it did not.

I realize that the suggestion to place bars across the back was a solution to the thinner Spider. I also saw the stay solution were you added an angle piece that went to another bolt hole.

I could easily try either solution. Then I could at least see if this fixed my problem.

Thanks

A random thought . . . . what would be the effect of having too little capacitance at the power supply to the geckos?

Posts relating to capacitance moved to Capacitors

Nils,
this is my last post on this subject,
I will say it was a pleasure to talk to you and try to help you with your problem,


with that said, I cannot believe how some of you trouble shoot your problems this is a simple deduction, and conclusion process.
everyone said that this was a computer problem, so a mutual friend of mine and Nils, went over to his house with his known good computer with no issues and ran Nils machine, found that the same problem existed with new computer,
so the computer was eliminated,

bearings in the router were thought to have been an issue, they were changed and problem seemed to get worse, not expected but understandable.
(since the run-out or slop in the bearings would or could dampen the back lash and when the slop was reduced it made it worse )

also the spider was thought to have flex which could multiply the back lash but I wouldn't consider it a major issue at this time, but I would repair it.
which brings us back to backlash,
this can be determined by two ways, by dropping the geared motors from the rack and feeling the movement in the gears since he is running 30 tooth pinions it will be easy to feel the back lash...( our mutual friend said that he could feel quite a bit of movement here).
which brings me to the second way to check the back lash, and that would be to put 20 tooth pinions on the motors if there is back lash you will see a reduction in the chatter marks, ( this is how I found my problem except I went from 20's to 30's and seen a large increase in backlash)
also I offered to make Nils a set of belt drives for free he supplies the parts and I will supply the labor he chose not to do this.
our mutual friend offered to bring his belt drives and put them on his machine to see if this would correct his problem he also chose not to do this.
all I can say is what I have done to correct the problems that I have had. as stated before you can choose to use this information or not,
but I would be willing to wager that it is back lash not a capacitor ( give me a break )
not the computer, not a program problem it is a plain and simple physical problem,
I guess I deal in reality not, conjecture or make-believe, I also had this issue with my machine and corrected it, my friend had the same problem and corrected his problem, so I speak from experience if you do not want to fix you machine that is your choice, mine works great...
//chopper

Chopper,

I appreciate all the help everyone has offered. I have tried some of the suggestions, others I have not.

I did not take you up on your offer to add belt drives to my machine because I did not want to add another level of complexity with out understanding what is actually wrong with my machine. You are convienced that adding belt drives will fix my problem. I am not convienced that is the solution.

There is a machine is CA that will cut straight cuts at 400 IPM and diagonal cuts at 300 - 400 IPM without a problem. That machine is like mine except for the ground rails. Same geared motors. I can cut a 4" square that has straight cuts a 600 IPM but I can't cut diagonals at 70 IPM.

I think I do have some mechanical issues that I am addressing. I do believe that the Z slide had some play in the V Wheels. I also believe that the C Channel not being level is a mechanical problem which I am addressing.

This is a tough problem to resolve espically when the owner of the machine (ME) isn't mechanically inclined and the people helping me can not see the machine. If I had lots of money to throw at it, I could try every thing but I have limited resources.

Nils,
I understand,
I meant no dig on you at all,
it was directed towards all the BS that was spewed, and I realize that you are limited in your mechanical ability, and mean no disrespect by saying it I hope you work it out and get it up and running soon,.you know how to get a hold of me if you need to ..
//chopper

Hi Chopper,

I've been following this thread with a lot of interest. I am glad to see that at least we know what the problem most likely is. I too am running the same motors Nils is using. Now, the question to you would be why does he have this problem and I don't. I cut 12-18mm Acrylic day in and day out often single pass at that. I am using a metric 1.5 pinion and rack. About the equivilent as Nils. If you can see chatter on anything, nothing is as apparent as acrylic. Now, my cuts are silky smooth with none of problems Nils has. I cut diamond shapes all the time at a wide variety of angles and never once did I ever see the type of patterns that Nils has on his machine. I haven't heard anyone else experiencing the same problem either. What's your take on this? Is it possible he got lemon motors??

David,

It has not been determined what is causing the problem.

I don't understand this comment Nils. You need to try it to dismiss the possibility that it will fix it. If it is the cause of the problem, then no matter what else you try you won't fix it and you'll never fix the problem.

I would take Chopper up on his offer to try his friends belt reduction. Theres no/mininal cost, low risk and a very good chance of fixing the issue.

All the best with it.

Matt,

When will you able to stop by and help trouble shoot?

I, do not know why yours are silky smooth, try cutting aluminum .250 at the diagonal, with a two fluted up spiral bit, and post the pictures, you can say it is so but we will never know if you cannot show pictures.
also if your motors are new (low hours) this seems to happen as the motors break in mine were fine at first also, with time the backlash becomes more apparent
//chopper

Matt

I apologize for my comment.

Nils..
A thought came cross my mind with my 1st morning coffee….
Could this be an effect due to some parts being out of round tolerances (Ie : Gear bore ctr, gear mesh, motor shafts… etc….moving parts ? )
Robert ;)

still waiting for pictures of your cuts, (plastic is easy to cut) in the meantime here are some more I cut today
6061, .250 thick 2 flute up spiral no lube, remember these are close up photos, things are bigger than normal size....//chopper

Nils,

Have you tried this cut in passes ? If it is a flex issue small cut passes would eliminate the flex. Look at each pass, maybe even do seperate lines side by side at different speeds 40,50,60 ipm.

Another thing I may have missed. Swap your Z motor with you X or Y then test then swap X and Y. If it is a single motor backlash issue you may be able to move it out this way.

Also note on the Aluminum cuts they are in .020 inch per pass.

If it is a osillation issue your pass cuts may show the imperfections in the same spot each pass.

With the even spacing on the cut path this looks like a cut path instead of chatter or backlash.

Another culprit to look at is your input power to the system. Like Gerald said the Capacitors or even the power from the wall could be suspects. Do you have a UPS OR is your table and PC on seperate circuits? Check your grounds as well.

Just some different ways to look at the issue. Hope this helps.

Also remember once you get the unit cutting at a decent speed and quality that suits you, don't push the edges again just to see what happens. This will cause frustration if the machine acts up running at speeds and feeds you wont use anyway.

[QUOTE=MetalHead;31841]Nils,



Another thing I may have missed. Swap your Z motor with you X or Y then test then swap X and Y. If it is a single motor backlash issue you may be able to move it out this way.(the backlash will be in all the gear heads so this will do nothing)

Also note on the Aluminum cuts they are in .020 inch per pass.
what are you referring to here????
If it is a oscillation issue your pass cuts may show the imperfections in the same spot each pass.
I agree with this statement
With the even spacing on the cut path this looks like a cut path instead of chatter or backlash.
(this could be true but unlikely but i didn't write the program so i don't know)

Another culprit to look at is your input power to the system. Like Gerald said the Capacitors or even the power from the wall could be suspects. Do you have a UPS OR is your table and PC on separate circuits? Check your grounds as well. ( do not agree with this statement the problem is.......wait for it..........backlash.......oooooohhhhh)
//chopper

I don't have much experience in CNC building, but I have lots of experience in manufacturing especially finding and fixing problems.

This thread reminds me of different problems in a factory. When we have a tough problem that is hard to find the root cause, we have a meeting with several experienced people and discuss possible root causes. We start with the easiest most likely solution and see if that fixes it. If it doesn't we go to the next and the next. During this time we have many tempers flair, but we keep going. We don't give up until we find a solution.

People on this tread is doing the same. Trying different things, coming together with ideas, and eliminating all the variables.

Though the solution is not yet here, we know what it is not.

One thing about using this forum, everything is well documented and can be used in the future.

Who knows what this issue will end up being.

The more steps we document for him to check here will help others in the future.

Nils another test would be to cut the same diagonal line job 4 to 6 times or look back at jobs you have already ran. Put them side by side like chopper has done and others to see if the cut paths look the same (ie the "chatter" marks line up) .

Chopper,
Those are some very fine cuts. I only get that kind of quality when I use my vertical mill. How about posting some photos of your MechMate, especially some closeups of the V-rails? Your MechMate is capable of making cuts that make the rest of us drool. Please share with photos.

there are pics of it in my build, I dont think there are close ups of the rails,
I will take some photos in the next few days, I had plans on doing that anyway
//chopper

Hillbillie,
I totally agree with you here, the problem is there is no logical path of deduction being followed, people are grabbing at straws and throwing out random guesses, and I guess one should expect this from an internet forum,
you have people here who do not even have a machine guessing at what the problem might be, never run a machine, or built one and for them it is almost impossible to relate to the problem, yet they spew out ideas that come to them with there morning coffee, not realizing that they are misleading others who may need the help, there are two classes of problems that can happen to a machine like this, one is electrical the other is physical,
there could be two different problems going on at the same time.
To me the easiest way to chase a problem would be to determine if it was a physical or electrical problem,
I have deduced that this is a physical problem, from the nature of what is happening and because of the known backlash problem in the 7.2's
and also knowing that he is also running the 30 tooth pinions,(which will magnify the back lash by going to 20 tooth pinions he will see smaller chatter marks and prove that it is backlash)
if you go back in this post you will see that I had a similar problem with my machine, and it was corrected by changing out the 7.2's to a belt driven system, if you look at the cuts above you will see that I have very little chatter and I am still using a Milwaukee router, so there is still room for improvement by going to a spindle, how ever slight the improvement may be, so what I am trying to say is fix the known problem of backlash, and if that does not fix the problem (which I sincerely doubt)
then chase the fairies in the electrical system if they still exist at this point...

Backlash can do many funny things that are not expected when cutting,
it will shorten the life of your bits, cause loss of accuracy and repeatability..
once the initial cut in the material is made and you have backlash the bit will follow the path of least resistance and follow the previous cut path, so if the cutter jumped around a lot in the first pass it will do it on the rest of them since the backlash is allowing the bit to do so. and the so called pattern could be generated by how the bit made the first pass and the flutes of the bit cut through the material which could quite possibly leave a pattern like what is seen ...
sorry I know I said my other post would be the last on this subject I guess I lied....
I also want to give a little background on myself so you know where I am coming from, for 25 years I did metal fabrication,( never really quit) from welding to forming metal of all kinds,machining of metals and repairing of machines, from semi trucks, bulldozers and fire trucks etc... then I got into the computer world when it was just getting going at least on the networking side of things, I worked for the local university and did everything from installing telephones and computers to
to designing complete networks in each campus building and then install them, (440 buildings) trouble shoot them etc.... so I know of what I speak, and have
quite a bit of experience in the trouble shooting field, both on the fab side and on the electrical side of things, was I ever wrong when trouble shooting of coarse I was but not very often,...
//chopper

I use EMC, so I have no practical experience with Mach, but has backlash compensation been tried?
http://www.youtube.com/watch?v=A27nOkTFYDY

Also, wouldn't cutting a small circle quickly identify any significant backlash?

on edit:
After more reading about EMC, I need to test (and correct if necessary) my machine for backlash too.

One other thing;
I'm having a hard time reconciling the idea that backlash is to blame for an irregular straight line (diagonal or otherwise). Backlash is normally manifest during changes of direction. Once a stepper begins turning in a given direction, backlash within the mechanism is taken up.

I guess I can envision severe backlash allowing chatter, but elsewhere in the thread (because of the regularity of the pattern) chatter was deemed unlikely.

this is true until forces push against the direction of travel, so as soon as the bit hits the material all bets are off on the back lash being taken up...
I do not know what your experience is with backlash some people do not understand it but I have had this issue and I did extensive research on how to eliminate it it does not just simply go away when the machine is in motion it is always there until removed you cannot wish it gone, and with the problem that I had I put a lot of thought in how to eliminate it, and as you can see by the cuts produced it worked....
//chopper

I do not think that the backlash compensation in Mach works very well I did try it with limited results on my machine....
you cannot fix a physical problem with software....
//chopper

Although I mostly agree with Chopper about backlash, I'm not sure that the "chatter" marks are the fault of the gearbox or of backlash. Here's an image of the edge quality that I got several years ago. The bottom photo shows a great deal of "chatter". It was cut on my PRT-Alpha that had neither gearboxes or belt-drives. The better cuts were all made with 3:1 belt-drives.

Mike,
I agree that backlash may not be the only problem Niles is having, but it would be hard to tell what other underlying problems may exist until the backlash is removed,
also I would like to ask what differences are there on my machine compared to others that would allow me to get the cuts I get?
there are some small changes that I made while I was building the machine, but none that I believe would account for the quality of cut on my machine vrs others,other than the belt drives,
the changes on my machine are as follows:

1)...I welded the rack to my machine instead of bolting and using vhb tape

2)...I welded the Z assembly instead of bolting

3)...I modified the spider by welding 1.25 X .625 thick band iron behind the
V wheels for spacers on the Z assembly and on the left side of the spider
I bolted the top and bottom V wheel to the spider with out using an eccentric adjuster, and put the eccentrics on the other 4 wheels so everything squares to the two wheels that were bolted solid.

4)... 4 to 1 belt drives

I also spent a bit of time making sure that things were level, square and true, other than the above my machine is built directly from the same plans every one else is using,...
my friend that has the belt drives on his machine has the same Z assembly
but none of the other changes and he gets really great cuts with his also,
I do not know if he has cut aluminum with his if he has I have not seen the cut quality to compare but I know what he has cut in wood and etc.. have been chatter free, so I do not think that my machine is some kinda fluke since both machines were built by two different people at different times in different states the only two things in common are the Z assembly and the the drives and both are chatter free...
//chopper

Mike,
I also wanted to ad that as you see with your belt drives vrs the direct drive that resolution of the step plays a big part in the cut quality also, that is why I went to 4 to 1 on the drives I built and I use the 20 tooth pinions, I do not know if that is a significant enough gain over the 3.1 or 3.6 to make that big of difference, and I also know that is why the 7.2 are so desirable, but the backlash negates the resolution in the end,...
but it still makes me wonder why our machines can achieve the cut quality that my friend and I are achieving and no one else?
my friend actually has cut with a 1/16 of an inch bit with out breaking it, he commonly uses 1/8 inch bits also, and if there were any kind of backlash in the machine those bits would snap of quite quickly,
//chopper

Nils - I wanted to add what i experienced,

Till now I was thinking that I was having the chatter marks on my machine due to a single spindle plate used to clamp up the spindle, yesterday i made the second clamp to get it on the Z assy,

upon removal of each part on the Z assy, I found that though the V wheels were holding the plate so tight that even after a considerable effort I could not get a play on the Z plate, but once I had the spring tension removed there was a plenty of play on the top two v bearings, the reasons were the slightly bent bolts on the top two bearings and also that the spider plate has bent and that makes the V wheels not aligned to each other , I did not have any spare bolts nor was willing to remove the whole Z assy to get it right, so currently I will get chatter when I stress the machine more, but on 3d works I did not see any at all.

I am planning for a totally new Z axis redesign -

RDS
Irfan

Irfan,
lets see the pics of your cuts, can you cut aluminum with out chatter marks?
I think the aluminum is the true test, if you can cut it with out chatter, you will not have a problem with any other material, I would like to see them to compare cut quality etc.. what is the ratio of your belt drives?
//chopper

Chopper,

Please spell my name correclty

I havn't looked at this thread for a long time but today I was told that I should check it out.

Quite frankly, I am most disappointed with the tone taken on by some of the members. I am not going to run a forum with these attitudes around and I am sorely tempted to delete this thread and send a couple of guys to the cooler for a week or more.

There's been a lot posted on the forum about belt-drives and gearboxes. I searched on the word "chatter" and was directed to "Driving Mechanisms: Rack/pinion, gears, screws, belts & chains". In that section is the thread, "Reduction Drive by belt - a collection of various designs". In the six pages of that thread, you'll find a lot of different designs that have been tried. On page three of the six-page thread, I saw the photos posted by Alan of some mirrors that he cut from oak. I hope that he doesn't mind that I've reposted one of them. That cut was made on his MechMate with motors that had neither belt-drives non gearboxes.

The chatter on a MechMate is very similar to the chatter on a Shopbot when chatter conditions are ideal. That caveat about chatter is important.

On my Shopbot, I didn't always get chatter. On most of the photos posted by MechMate owners, I don't see chatter. But, under the right conditions, especially with non-geared/non-belted motors, many of us get chatter.

On my Shopbot, with 3:1 belt-drive transmissions, MOST of the chatter was eliminated. Sometimes I still got chatter, although the roughness was greatly reduced. Even now, with 7.2:1 Tapered Hob gearboxes, I still get some chatter when the conditions are just right to produce chatter. It could be the gearboxes. It could be the construction of the Shopbot. It could be the lack of maintenance on my machine. It could be almost anything. I've done everything reasonable to eliminate as much chatter as possible, and I just live with what's left and rub it out with a few strokes of sanding. In the case of aluminum, I use a power sander. In the case of plastic, I run over to a nearby glass shop and borrow some time on their wet sander.

Here is Alan's photo:

Mike,

When I cut those mirrors, it was with two passes at about 4m/min. I have since tried similar cuts at a slower speed and doing rough passes about 1mm away followed by a finishing pass on size. This has given remarkably improved cuts with hardly any chatter. (using spiral / slot drill cutters). I am still going to make reduction drives but it is possible to get very good results if the correct cutting strategy is followed.

sorry it was a typo
//chopper

Alan,
Thanks for posting that information. Chatter is often a function of speed and depth of cut. I ran a series of tests on my Shopbot and saw a remarkable difference in chatter depending on how fast the machine was cutting. Recently, I made some cuts in aluminum with the machine running only 50% of the recommend feed speed and only 0.020" per pass. The cuts were very clean and basically chatter free.

There are a lot of factors that have to be watched to get an ideal cut. Many times, when cutting wood, it is much more cost effective to get a quick cut on the machine and then to spend a few more seconds with a sander - especially if I would have to sand anyway.

There was a fellow on the Shopbot forum who sold his original Shopbot because of chatter problems and then bought the newer PRS-Alpha model. He experienced the same chatter problems. He posted a bunch of very blurry pictures that he said showed how bad the problem was. We had to take his word because of the quality of the photos. Because he was using a program that Bruce Clark and I wrote together, I played with his numbers and saw that he had very unrealistic goals. He assumed that since that machine could cut at 12-inches per second, that it should be able to produce perfect parts at that speed. At 5-ips, on my PRT-Alpha, I got almost perfect parts. Shopbot cut some parts at various speeds and also suggested 5-ips as the ideal speed for that material and that type of cutting. He disagreed and returned the machine for a refund.

I've seen some terrible cuts made on a Thermwood when the operator didn't use the right speeds. I've seen some beautiful cuts on my entry-level Shopbot when all the factors were right. The MechMate is a much sturdier machine than my Shopbot. It has less flex and more mass, which would help in those "picky" situations when everything has to be just right for the cut to be perfect.

Although I agree with almost everything that has been posted about the value of adding a belt-drive transmission or a gearbox with zero backlash to a non-geared motor, I don't believe that gear reduction is always the magic bullet that kills all of the common problems that we experience. I've found that gear reduction minimizes the problems to the point that I consider them to be eliminated, but not always.

Your post pointed out the value of a little experimentation to find the best parameters. Thanks.

Mike,
I couldn't agree with you more, if you are running a reduction to get rid of the chatter and do not use the proper feeds and speeds the reduction will not save you, be it belt or gear.
//chopper

chopper - I don't believe that machining Aluminum is the tru test to test for chatter marks,

while cutting Alu - i reduced the step down and the feed rate of the machine - now what happens is that the Alu is beautifully cut, but when i cut ply / mdf at much higher speeds the chatter reappears

today i observed something while cutting 3d too, if I keep my speeds around 3000mm/min the results are great, any higher speeds and I observe pitting,

but I know what the problem is - its me who did not get my Z assy right , I am probably just going to shift to linear rails soon on the Z axis and I dont have any argument over that in my mind,

Sorry Nils I am cluttering your thread with my problem -

and Mike - I agree fully with you- we have to understand the limits and parameters under which to ascertain that we get good cut quality results.

RGDS
Irfan

I agree with Alan. As I only havea little 500 W Makita router on my MM, I cannot take deep high speed cuts. This I realised very soon after completing my machine. From there I take it slow and with cuts not more than 1.5 mm at a time. Like Mike said, that sometimes you get a little bit of chatter and other times less. My chatter is so little that with a light sanding they are completely removed....or it adds to the overall 'effect' of the piece I have cut. I actually would like it to CHATTER all the time when cutting flowers, leaves and wildlife scenes.;)

I would like to see the cuts you have made in aluminum, yes you can get chatter in any thing you cut I have heard all this before but no one will show their cuts, I would like to compare if you would show them, thanks
//chopper

tomorrow sir - its 1 AM here