Z and zero plate maybe

[bthoron]

We've noticed that our Z plunge depth isn't quite going as far as it says it should. For instance: a partfile with a step down of .375 appears to be cutting only .365 (roughly .01 less than it should).

We're using the zero plate and a new PRS with the 5/11/07 release of sb3. The unit values match what other posts have indicated. The thickness of the plate is .12", just as indicated in the setup routine.

When we MZ to 0.0 we appear to be accurately touching the material.

Thoughts?
Ben

[kirkkelsey]

I am currently dealing with the same thorny situation, where even after we Zzero the bit, it is cutting .009" thick.

I created a simple test block with a few slots in it to test actually milling depth. We measure the remaining material, and compare it to what was in the code, and while it varies within the .002" tolerance, it is consistently about .009" thick.

I talked with support yesterday, and they agreed that we needed to go into the C:\SbParts\Custom\my_variables.sbc and ADD .009 to the .121" of the Zzero plate value of &my_ZzeroThickness. By saying that the bed was further away from the current stop point of the Zzero plate, it would make the cut that much lower and nearer the bed.

There are posts in the archives that bear out this idea, and what is important is the actual machined material and not a caliper of the Zzero plate: http://www.talkshopbot.com/forum/messages/29/5497.html#POST25008

Unfortunately, when I went to test this today, I found we are moving the ShopBot to a new location this weekend and could not verify this in practice.

On a similar note, using the same milling test with 3 slots of differing widths, I have tried to correlate the actual slot milling width to that in the code. I found that our 1/4" Onsrud end mills caliper out to .246", and that I had to code a mill diameter of .242" in order get the slots to mill the correct width. Again, ShopBot support verified that this is again a norm and to be expected. I am not saying anything bad about the Onsrud bits as they are fantastic, but that any brand of 1/4" bits we have purchased caliper out a bit undersized.

I have been running test blocks as it is important getting accurate dimensions for joinery, and learn what are the milling limitations and quirks. Milling cross grain on the 3ply Bamboo plywood I am commonly working with also results in a narrower slot by .002" than with the surface ply grain orientation. Something more I have to allow for in the design and clearances.

We have a vacuum table, and I am trying to get the onion skin as thin as possible without marring the table, and need to determine the necessary tolerances before I get too close. The operator would not forgive me for milling into his table with my coding.

I can hardly wait for the ShopBot to be operational again so as I can test all this information. Please post your experiences as well.

[bthoron]

Very good- I will enter an adjusted zero plate thickness. I'm confused though, If you add thickness to the zero plate setting won't it cut less deeply? I had also noticed the cutters were undersized, huh, weird. The single flute straight cutter did slot at nearly .25", the spirals were all slightly more like .245". My operator was a little frustrated by my "mistakes with material thickness".

[knight_toolworks]

endmills are a bit undersized just the way it is even in the metal working field.

[richards]

There are a lot of contributing factors that add up to ruin tolerance. Yesterday I cut a bunch of airplane 'rocking horses' out of baltic birch plywood. No matter how I tried, that plywood would not lie flat on the spoil board. I ended up using both screws and vacuum before getting it to the point that I could cut parts. BUT height variation was still greater than 0.01".

When I first got my PRT-Alpha, I blamed the machine, now, after moderate experience with the machine and because of all of the possible variables involved, I make a few test cuts in the actual material that I'm going to use and then adjust the cut file to compensate for the 'sprites' that are active at that particular time.

[jamesgilliam]

Ben and Kirk, Have you checked the actual thickness of the Z-Zero plate? When I got my prt years ago it said it was .131", but the dial calipers measured .135". They do vary a bit sometimes.

[kirkkelsey]

Yes, checked the zero plate thickness. And after about the fourth consistently "wrong" number, I had them check the calipers were zero'd properly as well. I also constructed a test block with enough slots of the same size to be able to verify the dimensions in different places. I can now get enough measurements to average out measurement deviations by the operator. I also do not tell them what dimensions I am looking for during these test so they do not "try to please" me with dimensions forced a bit closer than they would have been normally.

You do want to ADD the distance you are too high above the table when milling. By telling the calibration routine that your zero plate is "thicker" than it really is, you are saying the table is further down than it really is, and to go lower when milling. Took me a while to wrap my mind around that one too.

I like the wood 'sprites' reference. Just as I was getting close with my compensations, we switched from some cheap upspiral bits to an Onsrud compression spiral, and my "deviation" changed. Different type of bit, different deviation. At least the boss saw the improvement in cut quality and is sold on the "expensive" bits. Based on the preferences posted in the forum, I also ordered a few Whiteside compression bits. Hopefully I will not experience deviations between brand of bit as well.

Oh, and to further complicate matters, ShopBot support related an experience where the thickness of the bed changed noticably due to a drastic change in relative humidity. I will have to keep my eye on the weather and hygrometer as well.

At least I have a use for all those small pieces of scrap. I may be running quite a few calibration test blocks, and definitely before any close tolerance joinery. Maybe I should redesign them as slotted trivets and the boss can sell or give them away.