Dialing In a Buddy

[acb]

Over six months with my 32 I'd noticed that my through-cuts seemed consistently a bit shallow... I almost always had some "skin" at the bottom. So this weekend I did a few quick tests.

Sure enough, it turned out that on cuts from 1/4 to 3/4 deep I was about 0.5 percent shallow (e.g., a cut that should have been 0.5 deep calipered at 0.496). Same factor seemed to apply, roughly, to X and Y... although I didn't mic my 1/4" mill, so that might be just a tooling variance.

Has anyone else seen consistent small errors on this order? And if so, is it worth adjusting the Bot's setup factors, or should I just apply a correction factor in the fill-in sheet each time I go to cut?

[erik_f]

You are well within spec. I would try playing around with they way you are machining. Also your spoil board even if surfaced will change hour to hour. The less stress you put on the machine and bit when cutting the better off you'll be. I know the buddy can cut pretty fast, but that doesn't mean you'll be happy with the results.

[gerryv]

Sounds like it could be slightly under-sized pinion gears but I'm not sure how this could be verified.

You might get more feedback on this question on the general forum board.

[billp]

Art,
You didn't mention if you had verified the actual thickness of your Z Zero plate yet. If the cuts are consistently shallow you may have the wrong number plugged into your Z Zeroing program. Mic the plate and see if that's the number your machine is using...

[ljdm]

Even though the z,x,and y all seem to be the same amount off, they might not be related. The z could be the z zero plate being a different thickness than what SB thinks it is, while it's not uncommon for 1/4" end mills to be off by .004. Just a thought or two....

[acb]

Yes, my Z-zero plate thickness is correct at 0.121", and in any event I made my depth measurements relative to a .01" pocket on top just to make sure I wasn't thrown off by material or spoilboard variations.

And my cut rates are 1 ips in Z and 3 ips in X and Y, which I don't think are excessive for cutting MDF with a quarter inch mill. (And yes, the tool is sharp!)

So aside from blaming the operator, anyone have any constructive suggestions as to how best to compensate for a reproducible error?

[erik_f]

If you are using the same feeds and speeds as well as the same depth of pass you will find your error will be reproducible. Try ramping your plunges and slowing your z down to .5". Also on the x and y try 1 ips and .125" per pass. Cut some test circles and squares with these more mild setting to rule out machine flex. I hope this is constructive enough for you.

[acb]

To recap, my questions were: 1) Has anyone else here seen anything similar? (apparently not), and 2) What's the best strategy for compensating? (nobody seems to have one, except for trying to talk the problem away.)

I'm assuming that slowing down to half an inch per second in MDF was meant as a fix...

[erik_f]

Art if you are frustrated don't take it out on the people that are trying to help you for free and out of kindness. No, slowing down was not meant as a fix it was meant as a means to investigate. You act as if you already know that everyone is wrong when you are the one asking the question. When I first started using my Buddy I was coming from a slower PRT machine and was excited about the new found speed and power. I was getting repeatable errors. I asked questions on the forum politely and took suggestions. I can't say for sure the problem I had is the problem you are having, but it sounds similar. Maybe you are too smart to ever have a operator error, but maybe, just maybe you are just like the rest of us.

[gerryv]

Thanks Erik

[acb]

Sorry, Erik, but I missed the part where you described your own experience. All I heard were suggestions of how my observations might be wrong. I certainly appreciate your good-heartedness in appointing yourself my personal Buddy (and now manners) guru, but if you can't bring yourself to take my questions seriously... or even if you just don't find me appreciative enough... you really aren't obliged to comment at all. I'll muddle through somehow.

That said... I have no reason to believe this is a malfunction, and slowing down that far isn't a solution. Like I said, I'm really just looking for thoughts on the best strategy for making a fine calibration adjustment.

Gerald may have had it right... this may be something I need to take to a larger pool than just this Buddy sub-forum.

[ljdm]

We all take everybody's questions seriously, otherwise why would we take the time to come up with possible solutions?
You are right - you aren't appreciative enough, so guess you'll either have to put up with helpful suggestions, or have to muddle along alone.

[ljdm]

Add to last post - not trying to flame anyone, just making the point, we all try to help each other. Frustration sometimes makes it hard to appreciate helpful hints, especially when the suggestions don't solve the problem. But at least you are getting input from others who might have similar problems, which might help solve your problem.

[erik_f]

Gerald, Lou and even you Art!

[acb]

OK, I get it... I'm incapable of telling what's helpful to me. Everything that gets posted here (by anyone other than myself) is perfectly correct and precisely on-point. And I'm an ingrate for imagining otherwise for a single second. What WAS I thinking?

Whatever... life is short and the tribe has spoken. And who can argue with logic like that?

[acb]

Lou and Erik both slipped in ahead of my previous post and cooled the tone a bit, which is a good thing.

Nonetheless, I think we all may want to check our agendas a bit if we assume that we're entitled to have everything we say here automatically treated as if it were purest gold. Realty exists: the iron doesn't care how sincere we are. I assume that a community of craftsfolk appreciates that.

Somewhere along this thread I got characterized as "frustrated," which turned out to be pretty much self-fulfilling. Actually I hadn't felt frustrated up to that point... I was just trying to correct what I saw as a degree of topic drift.

In fact, being labeled that way ticked me off precisely because I had the distinct sense, rightly or wrongly, that I was the object of a degree of projection by folks who were themselves frustrated that they weren't getting the degree of flattery they assumed their inputs deserved.

But I'm certainly prepared to be wrong about that.

[rhfurniture]

Art, to get back to your original question..
What I have done on mine with the X and Y is to drop 2 holes along the axis at a large convenient distance apart, measure accurately, and tweak the factory set unit values to compensate. They are VERY SLIGHTLY different, but my cuts are all now spot on. One could no doubt do the same with the Z, but it has never bothered me, and the 4 thou you reported in post 1 is probably as much to do with the vagaries of Z zeroing, measurement and material thickness.

[erik_f]

Art is the perfect name for you, because you are a piece of work my friend! Good luck I hope you work it out.

[ljdm]

Ditto Erik.....
Last post for me on the subject.

Try playing with the unit values a little. I had to tweak mine, my x znd y cuts were off by about 1/32 over 48" and x was off by about 1/16 over 96". Same percentage error on x and y. Call Shopbot, they can calculate for you which values to enter. Example - 1833.4455 might need to be 1833.3455, etc.... Or just calculate it yourself. Take the unit value for x and y, and try either adding or subtracting .05 percent from it.
There. A suggestion. Worked for me. MIGHT work for you. It will definitely change your cuts by the percentage you change it by. Still don't know why mine was off like that, but at the suggestion of SB support, I tried that to compensate for the error.

[acb]

OK, so... a couple of folks HAVE observed the same sort of thing, and their strategy was to tweak their unit values globally.

Seems so simple when we all stay on point, doesn't it? Thanks.

(As for anyone who's still nursing a personal grievance... I'd encourage you to reflect on your own motivations. Were you really posting out of a desire to help someone else, or is it possible your primary concern was for your own ego? Just asking...)

[Forum Admin]

Reflections on motivation will be considered off topic and should go private.

[myxpykalix]

Art,
Have you tried taking a piece of wood or mdf and making a long cut in the X and then in the Y to see if the consistency of depth is the same in both directions all along this path? That might rule out whether or not your table might be out of plumb.
Have you called tech support, they would be the best source of help on this.

[acb]

Didn't do it precisely that way, Jack, but I started with a fresh surfacing run on my whole spoilboard. Then I made my depth measurements relative to a shallow "surfacing" pocket on top of the test material to level out any material variations. That's why I'm fairly confident I was measuring actual cutter Z-movement.

(As mentioned above, I was less rigorous in the X and Y, as I measured those pretty much as an afterthought... it was my depths consistently coming up short that had me really curious.)

Anyway, being off half a percent with the generic unit values didn't surprise me all that much. What I was curious about was how much variation other folks had observed, as an informal sample of variance over a larger population of machines... and whether folks had dialed-in their global unit values without (or with) any non-obvious side effects.

[ljdm]

I only noticed my x and y were off when I did some full sheet projects for the first time. 1/16" over 96" wasn't noticeable on smaller jobs like 10" x 12" plaques, etc. Had me scratching my head for days until I talked to SB support, when they had me try to adjust the unit values. Still don't know why or what was off to need to adjust the value, just know it fixed it.
Hate to get you going again, but, if the z is consistently off the same percentage, try adjusting the z value. Now don't go beserk at the suggestion, mic the actual z travel vs the SB reported distance, if its off by a certain pct, try adjusting the z value accordingly. There are no side effects that I know of, except that it might cause more accurate cuts.

[acb]

Well, that was more or less my original question, Lou... whether it was better to adjust globally using the unit values or cut-by-cut. Because I did compare the measured-versus-SB values; that's where I started. Was I unclear about that?

Anyway, now I'm confused: if SB recommended you change your unit value and you did and it worked for you, why do you recommend I do it the other way? (Not going "berserk," just asking. Hope that's ok.)

[ljdm]

Not sure what you mean by "the other way". I thought I said to change the values. My recommendation is - if the actual travel of the z,x,or y is off by a certain percentage, .1 at 1',.4 at 4', .8 at 8', adjust unit value accordingly. Best bet,if unsure about changing values, call support, tell them how much it is off at certain distances, they can figure the calculations for you. Measure at longest distances of travel possible, for best accuracy in measurement of error. As far as adjustin globally or cut by cut, global is best. If the unit value is correct at 1" travel, it will be correct at 96" travel. Email your phone number if a call will help.

[acb]

OK, so when you said z-value you meant the unit value. My mistake, I though you were saying to adjust the cut. Got it now.

[cubicdissection]

I just have to say I'm amazed more and more by this forum every time I visit it. This guy comes in with an attitude and you continue to help.

Shopbot forum....you make me want to be a better man *sniff*

[kevin_morin]

Art, when we began testing a new Buddie48 PRS for repeatability we used a dial indicator inverted under a centering point bit chucked in the spindle.

This confirmed that we could develop a few thou hysteresis if we used an XY move then Z up and down followed by an XY return and a Z up down. After a few cycled moves we found about 0.002-0.004" hysteresis in repeatability of the Z axis movements.

I thought that was just the motors and drivers loosing a fraction of a step of resolution over repeated operations?

What did you do, if you did finally resolve this matter, to solve your problem/question that lead to this thread?

cheers,
Kevin Morin