Houston we have a problem! (intermittent spindle)

@Earwigger I want to defer judgement until they can do a teardown of the spindles. We had some fail before they were turned on the first time so in that case I don’t think the bearings could be the problem. Also these are used in another application that has very high vibration, much higher than this application and they do not have this high of a failure rate. I’d really like to give them the opportunity to do an analysis before we make any snap decisions. I promise to give a report as to what the tear down surfaced and the corrective action made.

A Dremel is a nice tool to use as well.

I have also started reaching out to some German spindle makers to see what that side of the market has to offer.

If anyone wants to chime in with alternatives styles I’m all ears. We are interested in a low cost entry level solution for people just getting started. This approach seemed like an attractive solution that was a step up from the generic rotary tools we used to use. We’re also interested in some more advanced models like the water or air cooled VFD and even a really nice German model.

I bolted this guy on yesterday using the mount it came with bolted to Inventables’ Standard Wheel Carriage Plate mounted vertically (I trimmed off the bottom). Using a 48 volt power supply it is whisper quiet and uses the same collet system as the default spindle. It works like a dream.

(It may be the same as Inventables’ 600 watt Quiet Cut - is sure looks similar).

1 Like

“low cost entry level solution” to me means a small router. If you’re tied to a 24V power supply, your options are very limited. I would rather have you research weight/size limitations and post that data - what’s the heaviest spindle I can mount before I get deflection? How much weight can the NEMA-17 motors push? or the -23’s (both the 140 oz and the stronger ones you’re selling.) What’s the widest/tallest possible before it starts to interfere with something else?
Give us all that data up front at the store so it can be seen before we buy.

That spindle is rated up to 24-110VDC. It has a speed rating of 3-13K. The 13,000 will be there if you provide it with 110VDC. What speed is it running at with 48V?

1 Like

I do not yet have a means of measuring the speed. I have lots of Arduino’s laying around and will eventually rig up a some kind of tach or PID. Here is a video:


Most of the noise is my super loud Shop Vac. Sorry about that!

I would love to have a 110V water cooled VFD and mount that I could purchase from Inventables


@JoeMeyer there are theoretical values we can publish but also data from testing.

How do you suggest we measure the deflection in a real world test? Repeatability of a large square, circle, and triangle over a long span?

I think if I had my machine all set up I would slide the carrage off to one side and rig a dial indicator in the middle of the gantry. then I would start adding weights of know amount in the middle of the gantry and see what my dial indicator says the sag is. If I had weighed the carrage before it was installed I could compare the numbers and see if I could go with a heavier spindle.
You could also rig a dial indicator on an arm that is at the center distance of the spindle from the center of the gantry and and rig a way to add weights at that distance and see what kind of twist you get.

I was looking at the various drawings and 3D models of the machines on GrabCad and I think there may be enough room under the gantry between the gantry and the bottom of the carrage to put a T brace that would stiffen the gantry.

I expect that a good portion of your customers are wood workers and may or may not have dial indicators in there arsenal of tools. They may or may not have the equipment to build a T brace for under the gantry. That may be something that you and your team could get an extrusion made that could be a optional add on.


In junior high the science students build bridges out of toothpics and glue, and then hang a weight on it until it breaks. Highest weight wins. I think that with just a little spitballing the engineers on your staff could come up with something just as effective. :smile:

Do that simple calibration test, dial it in tight. Then put a heavy spindle on it, throw some SCUBA weights on top, and put it to work with all movements at top speed - rapid transits, cutting shapes, lots of sudden stop/go/change direction/change depth. And do it for a few hours nonstop. If you don’t fry a motor or driver and it doesn’t just go off and start cutting pentagrams and 666 where you’ve programmed circles and triangles and squares, then slow it down a bit and run another calibration test. Does it hold?
And all the time you’re doing that - measure how hot everything gets, how much current each motor is pulling etc. If you can abuse it for 8 hours with a 10 pound water-cooled spindle and four more pounds of weight sitting on top, then you would probably be comfortable recommending a 10 pound spindle for normal work.

That’s also about how I’d have tested all the spindles you already sell - max speed/feed/depth/distance for hours on end. Use them, abuse them, see where the threshold is for failure. Better to learn it in the testing lab than to hear it in a forum 30 days after launch.

Hi @JoeMeyer your suggestion is similar to the testing we performed prior to launch. Unfortunately, this method is not precise enough for your requested data above.

The dial indicator approach proposed by @DavidSohlstrom will provide some data. We did this in a video for Carvey but there are quite a few concerns that this data is not valid as it is not the same as if the machine was under load.

One observation from the video. repeatability were OK tests. The run out test should of been run on a ground blank in the collet. The OD of the collet adapter could not be concentric with the ID with a ground blank in it.
The other test that can be done is clamp a 123 block down to the table. run your test dial indicator up to the 123 block and zero the dial. Now tell the machine to move say 1 inch. Using a test block that you know is 1.000" tall slide it between the 123 block and the test dial indicator. If you indicator zeros then you know the machine moved 1". Now move back the 1" to the 123 block and see if the indicator again zeros. Do this for all three axis several times. This will give you a better idea of repeatability and if you have your machine set correctly for steps per unit.


1 Like

FYI - finally got a laser tach. The spindle speed at 48v is 6,600 rpm. It works but can be jittery when the gantry is moving away from the front (Plus on the Y axis). More flutes on the end mill seems to improve this situation. I have a 110v supply coming on a slow boat from China. I would like to get this dialed in and set so I can get to work.

Second spindle died this morning. I’m beginning to think its a heat related issue. Machining a 3D part that was suppose to take 18 hours, make it to the 16 hour mark :stuck_out_tongue:

I’m thinking the stock spindle is only good for very very very very very very very light duty applications where you engrave or cut out a thin plate or something.

I’m thinking I need a 110V VFD water cooled spindle. hint hint :smile:

speaking of VFD, how’s the mount for the VFD coming along?

Quick update. We got 4 returned spindles in today. Thanks everyone. We shipped them back to the manufacturer and are now waiting on the analysis from the QA lab.

@JerryWoods we are expecting the mount for the Dewalt 611 in next week. That mount can also be used for the VFD with a shim.

Will you let us know when the mount is in the store. Im waiting to order one of those mounts please.

Yea we’ll make an announcement.

Very cool cant wait. Do you know what the price will be?

I think we are announcing that next week.

It would be spectacular if we could order the mount + the VFD spindle as a combo thing.