Reason for sudden Z plunge?

[profexorgeek (Justin)]

I have a Shapeoko 5 Pro and was recently cutting my dice box pattern, which I have cut many times. The mill seemed to suddenly plunge into the work surface, although that may not be what happened. Here’s what I know…

The file looks like this:

The simulation looks like this:

I have cut this pattern many times, usually with a 1/4" downcut but sometimes with a 1/4" upcut. I recently changed magnet vendors and made a very slight change to the magnet pockets for size differences. I also decided to use my 1/4" upcut because my downcut has a slight inward vee and leaves a little bump in the magnet pockets, which makes them hard to glue flush.

I had just cut the file in Walnut within the last hour or so. I was cutting the pattern in Purple Heart with no changes to the file. As far as I can tell, the 1/4" upcut mill plunged suddenly into the stock while making the first pass in a pocket cut. This ripped the work out of the hold downs and I ran to the machine, hit the emergency stop, and turned off the spindle.

This is what the piece looked like:

Notice how the cut ramps downward into the work before plunging all the way through in the ragged hole. This could have been caused by the work surface lifting but… would it do that? Can an upcut mill basically make a piece climb the bit? It seems more likely that the router began to ramp into the material but I don’t know why.

After cutting the walnut piece, only two things changed:

• Switched from downcut to upcut with a flatter milling surface
• Switched from Walnut (1000 janka) to Purple Heart (2500 janka)

Either the machine/controller malfunctioned and did a sudden Z-plunge OR the upcut combined with the hardness of purple heart ripped the work out of the clamps and bored through it. It seems unlikely that the mill drilled all the way through 0.75" of purple heart without a backing surface but there is no spoil board damage. Also, I had just cut the same g-code so I’m reasonably confident there are no g-code errors.

Finally for reference, this is what these boxes look like when finished:

What happened?

[mhotchin (Michael Hotchin)]

Are you certain that the bit was not pulled out of the collet? What method do you use for tightening the collet?

[gdon_2003 (Guy Donham)]

I recently had a bit come loose and plunge down into the work. It did not plunge as far in as yours. I was sitting right in front of the machine looking at my phone. Just prior to starting the job I had removed the 1/4" collet and cleaned it with brake cleaner and used a q tip to clean up inside the router shaft. So I am sure it was not dirty or oily inside. I stopped the job and changed the collet to a spare I had and went on. I did stop every 5 minutes for a while with a pause and checked the tightness of the collet. I think I had a bad collet and have it segregated right now.

luckily the plunge was in a pocket near the middle of an 8.5" x 11" where an 1/8" piece of Baltic birch was to be glued to for me no harm no foul.

So check your collet for dirt or other contaminate. You can put a magic marker at the top of the bit where it meets the collet for a quick visual check if your bit is coming loose.

In my case the wood was South American Rosewood which is very hard and was vibrating the spindle and with my collet maybe defective it was coming loose. My bit was a down cut which would try to dig in and/or pull the wood up due to the high tool engagement.

[quicky06 (Ryan Quick)]

My 5pro did the same. Ended up being a loose terminal at the z motor connection.

Shake the connections and see if you can get the axis to bump when idle.

My pin was visually larger and discolored from the heat.

They sent me a new motor and harness assembly

[profexorgeek (Justin)]

I really appreciate all the replies! I read these as they were coming in but haven’t had a chance to do some objective testing.

@mhotchin and @gdon_2003

I thought I was certain that bit didn’t move significantly out of the collet because at one point the work piece was hanging on the bit and spinning furiously before I hit the kill switches . But, logically this could be a cause.

I have the carbide router (not a VFS) and manually tighten the collet. I regularly blow the sawdust out of it and I tighten it pretty hard with a wrench. That being said, it could have descended some and then pulled the work out of the clamps. That would explain why the spoil board wasn’t damaged in the “plunge” and the work ended up spinning around on the bit.

If I’m thinking about cut direction correctly, an upcut bit would pull itself into the material more deeply, whereas a downcut bit should push itself out of the material. It’s logical that the harder material and change in cut direction caused this.

@quicky06

This is really interesting because there is one other time where this same thing happened and it also looked like a random plunge into the material. But it has only happened twice and it seems like the Z-axis problem would be more frequent if this was the issue. I will check carefully!

Thanks for the tips, I’ll investigate and test as soon as I can and post some photos of anything interesting that I find.

[quicky06 (Ryan Quick)]

The longer the cutting sessions the most likely it will show up and the worse it will be overtime. I’m a tech at a new car dealer. I can tell you how it works. You have a loose connection which causes an increase in resistance which generates heat. The heat increases resistance but also causes the female terminal to expand which causes more resistance. And it will snow ball until it looses connection.

Mine did not happen very often. It only did it twice. First was at the end of s 3 hour session. The other a 3.5 hour session. Luckily I do nested parts so loses were minimal.

It’s an easy check to start shaking wires.

[profexorgeek (Justin)]

Okay, had a chance on lunch to do a little investigation.

Here’s the bit:

First, I don’t think the bit slipped in the collet. If it did, it was a very small amount. You can see the shank in the lock hole on the router shaft. It may have moved 1/32" but not the 1/8" or more reflected in the plunge ramp.

Second, there is a lot of buildup on that bit! This bit was brand new. It was razor sharp and shiny so all of this buildup happened during this cut. I haven’t personally experienced buildup like this cutting purpleheart but this was a new piece of wood and the wood database says:

Working with purpleheart can present some unique challenges: if the wood is heated with dull tools, or if cutter speeds are too high, purpleheart will exude a gummy resin that can clog tools and complicate the machining process.

I also turned the machine on, examined the Z axis stop and stepper wires and moved them around a lot. I saw no electrical activity and no sign of damage, arcing, or poor fit. I don’t see any evidence of a Z plunge for mechanical reasons.

Finally, I also examined this piece again and compared it to another piece that failed some time ago. My other failures have all been in woods that were also incredibly hard. This bloodwood cut also exhibited really strange Z activity and then drifted wildly through the wood (some of this happened after it was ripped out of its work holding).

I have not eliminated the potential for software malfunctions but I don’t really have a way to diagnose and eval. The fact that I have had 4 failures in probably 40 cuts of this pattern with similar variations, and that each of those cuts were in the hardest woods that I cut, makes software malfunction reasonably unlikely.

In conclusion I think the most likely culprit is poor work holding and more settings tuning needed for each wood type. Purpleheart and Bloodwood have both proved very challenging to machine. I thought perhaps I needed higher RPM and lower movement speed but that may just cause more resin buildup.

Thanks for the feedback and suggestions. Let me know your thoughts.

[neilferreri (Neil Ferreri Ω)]

@profexorgeek

That would explain why the spoil board wasn’t damaged in the “plunge” and the work ended up spinning around on the bit.

I think this is your answer. An uncut endmill is designed to pull material up. Make sure work holding is good and your depth of cut is adjusted for the material. My guess is that the purple heart got pulled up slightly, forces were too high, machine lost track and took got buried. Easier to pull the work up than pull your Z axis down.

[dustmite (JN)]

With extremely hard woods, you’ll be safer using a smaller DOC, and perhaps getting a higher quality bit. There’s lots of search results on google with people having problems routing purpleheart.