Spindle problems . What is the runout on your spindle?

[eltejano]

Just curious what everyone else's spindle runout was. I measure mine at the taper and got about .0008, almost a whole thou. Does this seem a bit excessive? My gear had mill drill was almost completely still when I took a reading with the indicator at the taper, and it was from Harbor Freight. My Bridgeport is almost as good as it.

Seems to get worse the further out. I was using an 1100 at Techshop Austin and when I try to run the exact same parts toolpaths, material, everything, I get a poorer finish and its a bit louder. Especially on a 1.5 by 3/8 keyseat cutter run on aluminum, I can hear it hitting one side and not the other. At the techshop one, the runout sound was there but barely and I got a great finish.

Here is a pair of parts that I made on either one. Both of these were made with the same exact program. On the rough one, I actually had to slow down that top pass because it was getting real loud, I thought my 3/16 bit would break.




Pretty much has been a struggle to make parts, having to modify all my programs cause I have broken 2 of my 3/16 high helix Destiny tool cutters when I tried to run them at the same speed as I did on the other 1100.

Anyone know a good spindle grinding shop in case I get no help from Tormach?

BTW, this machine is a few weeks old. and it reads almost perfect on the outside of the spindle nose, just off inside on the taper.

Thanks

Similar Threads:

• Spindle Runout.
• How much spindle runout is acceptable?
• Newbie- spindle runout?
• X3 Spindle runout
• Spindle Runout

[Jim Dawson]

Kind of looks like the spindle bearings might need to be snugged up a bit.

[GLCarlson]

Kind of looks like the spindle bearings might need to be snugged up a bit.

What temperature is your spindle showing after 10-15 minutes? (Shoot the inside of the spindle bore with a remote reading IR gun- they only cost 15 bucks or so on eBay).

Tormach says normal spindle temp is 150F ( 180F has also been used). If your spindle is dead cold -room temp- after a warm-up, your bearing preload may be low. Easy adjustment, there's a Tormach app note on replacing the bearings that covers it.

FWIW, my original bearings ran barely warm- maybe 10 over RT. Replacements run about 130F at 5K rpm, much warmer but within spec. Runout is negligible on my Last Word measured inside the bore on the R8 taper.

If your machine is a few weeks old, it's under warranty. Call Tormach with the details; if it's really a misground spindle, that's a very easy swap, and once you have the data showing that's the problem, I'd expect Tormach to just ship you a new spindle. I certainly wouldn't get in the weeds regrinding before talking to Tormach, and based on previous experience, probably wouldn't go that way at all. A new spindle is $545 for an 1100. If you can get a quality regrind done at a spindle shop for that I'd be surprised.

[eltejano]

What temperature is your spindle showing after 10-15 minutes? (Shoot the inside of the spindle bore with a remote reading IR gun- they only cost 15 bucks or so on eBay).

Tormach says normal spindle temp is 150F ( 180F has also been used). If your spindle is dead cold -room temp- after a warm-up, your bearing preload may be low. Easy adjustment, there's a Tormach app note on replacing the bearings that covers it.

FWIW, my original bearings ran barely warm- maybe 10 over RT. Replacements run about 130F at 5K rpm, much warmer but within spec. Runout is negligible on my Last Word measured inside the bore on the R8 taper.

If your machine is a few weeks old, it's under warranty. Call Tormach with the details; if it's really a misground spindle, that's a very easy swap, and once you have the data showing that's the problem, I'd expect Tormach to just ship you a new spindle. I certainly wouldn't get in the weeds regrinding before talking to Tormach, and based on previous experience, probably wouldn't go that way at all. A new spindle is $545 for an 1100. If you can get a quality regrind done at a spindle shop for that I'd be surprised.

Its the spindle. I have it marked on the high side. I got all my cutters marked on the opposite, low side. If I dont line them up just right, the runout can get crazy. In the morning, I come out, turn the machine on and let the spindle just run while I get my coffee and breakfast every morning. This finish was after I ran it all day already.

I can mark all my cutters and match the runout opposite to the spindle now, but when I get the tool changer, this will not be possible. It does not stop at the same location every time so I cannot control this. I have gone back and forth with tech support about this. The tech person told me something to the effect that 1 thu was not too bad considering that it is only really off .0005 since the thou is cumulative and thats pretty good. I called and left a message, I need this thing swapped asap.

I actually emailed a good grind shop and explained the situation. They just need to know the spindle bearing size since they will not do it without replacing the bearings with ones of their standard. So, at least, they seem willing to work with me. I need to find the receipts of another shop I have sent several Bridgeport spindles for repair some years ago.

You are lucky to have gotten such a concentric spindle.

[GLCarlson]

The lower bearing is WuXi Sanli Bearing Co 7008C/P5 DT. Don't have a number on the upper, but several forum members have done bearing swaps so someone will know. Presumably with a number it'll be possible to get a spec that will cross to a size match from a US or other supplier. A set of bearings from Tormach (upper and lower) is about 200 bucks. Somebody recently commented thata set from another supplier was a lot more expensive.

There was also a published procedure (from Precise Bits, I think) about how to tweak in the internal bore of a router spindle. I've never tried it, but it looked easy (and rather crude, too). Reportedly the results were good enough that high precision chucks (2 tenths) could be used with very small -.032"- router bits. If you're going to regrind, it might be worth trying that first. Don't know that the same procedure would work with an R8 bore; precision router collets look like an ER collet and don't have a second locating surface in the spindle.

I'd be interested to know what the cost finally ends up if you go the regrind rather than replace route.

[SomeWhatLost]

try emailing Tormach your pictures/description of the problem? maybe also provide a like to this thread?
I don't know...
but what you have is obviously not right, you just need to find a way to properly convey that lack of rightness and how wrong it actually is to them...

[SCzEngrgGroup]

Is spindle runout on the Inspection Certificate, or published specs for the machine? I would not be terribly surprised if +/-0.0005" was the spec for machines in this class... But if you can find a spec, you at least have a club with which to hit Tormach over the head and demand a new spindle (assuming the machine is fairly new...).

Regards,
Ray L.

[RedneckPhysics]

On my shop's 770, I see a bit under 0.0002" TIR at the spindle taper, BUT the TTS system ER-20's give me anywhere from 0.0003" to 0.0025" TIR at the cutting plane, depending on the assembly and insertion angles of the collets and holders. I've had parts turn out like that when I forget to map or match the cumulative runout values on a new cutter.

That's a bummer that you're already starting out with such a high runout at the taper... definitely not acceptable in my opinion. (Kindly) request a new spindle cartridge under warranty!

I would send them those part pictures, along with a video of you measuring the TIR of your spindle taper and tool. I find videos have the highest impact value with Tormach CS, as they're easy to understand, quick to watch, and provide a lot of information, such as making sure you did everything correctly in measurement, for example. I find those always get the best responses!

Hope that Helps,

Mike


PS: The Runout tolerance on the Certificate of Inspection for the 1100 is 20um "near" the spindle, and 30um at 100um away, as measured on a precision R8 calibration bar, I assume.
Same specs for the 770's spindle as well. In my case, the spindle was factory checked to be 10um and 30um respectively.

So, you might be able to interpret a 0.0008" runout on the taper face as out of spec, as I couldn't imagine a scenario where that much runout (if it's 100% angular and radial, and not caused by an error of the bore's shape itself) wouldn't cause an issue when measured anywhere outside of the spindle.

Have you measured at multiple depths on the taper itself, and compared the runout measurements with respect to angle at different heights? Might not be able to tease out much due to uncertainty in the tenths range though.
Also, sorry if I missed it, but what values did the inspector state on your machine's certificate?

Final thought: The short R8 collets that Tormach sources for TTS are mostly crap in my opinion. I recently re-ordered after springing the original (fantastic collet, by a different manufacturer), and ended up having to order 4 in total, before I found one that worked to my satisfaction.

[GLCarlson]

Is spindle runout on the Inspection Certificate, or published specs for the machine? I would not be terribly surprised if +/-0.0005" was the spec for machines in this class... But if you can find a spec, you at least have a club with which to hit Tormach over the head and demand a new spindle (assuming the machine is fairly new...).

Regards,
Ray L.

As usual, Ray makes a very good point. I looked at my own inspection report. The "radial run-out on the axis of the spindle conical hole", quoting the inspection report, is spec'd at 0.02 mm "at the place near the spindle end" and 0.03 mm "at the place 100 mm away from the spindle end". Essentially, if I read the diagram correctly, Tormach chucks up a test bar -perhaps with an R8 collet, perhaps a test bar with an integral R8 fitting- and then measures close to the end of the spindle and then 100 mm away.

0.02 mm is 7.87 E-4 inches.

That's not how the OP measured spindle run-out if I understood correctly. Measuring runout per the inspection document will certainly nail down whether the spindle is in spec or not.

Tormach's inspection report on my 1100 said mine was 0.015mm at both locations. Since I don't have a test bar, I can't replicate what I think Tormach did exactly.

[eltejano]

As usual, Ray makes a very good point. I looked at my own inspection report. The "radial run-out on the axis of the spindle conical hole", quoting the inspection report, is spec'd at 0.02 mm "at the place near the spindle end" and 0.03 mm "at the place 100 mm away from the spindle end". Essentially, if I read the diagram correctly, Tormach chucks up a test bar -perhaps with an R8 collet, perhaps a test bar with an integral R8 fitting- and then measures close to the end of the spindle and then 100 mm away.

0.02 mm is 7.87 E-4 inches.

That's not how the OP measured spindle run-out if I understood correctly. Measuring runout per the inspection document will certainly nail down whether the spindle is in spec or not.

Tormach's inspection report on my 1100 said mine was 0.015mm at both locations. Since I don't have a test bar, I can't replicate what I think Tormach did exactly.

I understand that you need an spindle runout test bar to really check the runout. Runout check at the taper only tells a small bit of the story. At best, it shows that there is a minimum of the measured runout. At worst, it could be indicative of a bigger problem like a bent spindle. All the grinding in the world will not fix that because the top of the spindle where the top of the r8 collet rests is at an angle.

But, I have taken my tooling, checked the runout of the end of my cutters with the same collet on 2 machines, one with almost no noticeable indicator movement at the taper and mine. Same everything except machines. Everything on mine measured just worse.

Cut the same parts I have on that one as on this one, same speeds and feeds, same everything. Mine still sucks. Especially with my wide keyseat cutters, you can really hear it and see the crappy finish.

Anyway, I just talked to someone at Tormach and they are sending another spindle tomorrow.

They should offer a precision ground spindle with a lot tighter spec. I would gladly pay extra, especially with better bearings.

[eltejano]

On my shop's 770, I see a bit under 0.0002" TIR at the spindle taper, BUT the TTS system ER-20's give me anywhere from 0.0003" to 0.0025" TIR at the cutting plane, depending on the assembly and insertion angles of the collets and holders. I've had parts turn out like that when I forget to map or match the cumulative runout values on a new cutter.

That's a bummer that you're already starting out with such a high runout at the taper... definitely not acceptable in my opinion. (Kindly) request a new spindle cartridge under warranty!

I would send them those part pictures, along with a video of you measuring the TIR of your spindle taper and tool. I find videos have the highest impact value with Tormach CS, as they're easy to understand, quick to watch, and provide a lot of information, such as making sure you did everything correctly in measurement, for example. I find those always get the best responses!

Hope that Helps,

Mike


PS: The Runout tolerance on the Certificate of Inspection for the 1100 is 20um "near" the spindle, and 30um at 100um away, as measured on a precision R8 calibration bar, I assume.
Same specs for the 770's spindle as well. In my case, the spindle was factory checked to be 10um and 30um respectively.

So, you might be able to interpret a 0.0008" runout on the taper face as out of spec, as I couldn't imagine a scenario where that much runout (if it's 100% angular and radial, and not caused by an error of the bore's shape itself) wouldn't cause an issue when measured anywhere outside of the spindle.

Have you measured at multiple depths on the taper itself, and compared the runout measurements with respect to angle at different heights? Might not be able to tease out much due to uncertainty in the tenths range though.
Also, sorry if I missed it, but what values did the inspector state on your machine's certificate?

Final thought: The short R8 collets that Tormach sources for TTS are mostly crap in my opinion. I recently re-ordered after springing the original (fantastic collet, by a different manufacturer), and ended up having to order 4 in total, before I found one that worked to my satisfaction.

I sent video and pictures. They said they would send me a spindle tomorrow.

I have enough headache dealing with runout at the tool, dont need runout at the spindle to give me a bigger headache.

Right now I have all my cutters and spindle marked so I can match the high spot one the tools with the low spot of the cutters. It gets ugly when I they are both lined up at the high spot. I broke a few 3/16 cutter that I can normally push fairly hard because the runout was on the same side on both. This is when I began to investigate.

[eltejano]

As usual, Ray makes a very good point. I looked at my own inspection report. The "radial run-out on the axis of the spindle conical hole", quoting the inspection report, is spec'd at 0.02 mm "at the place near the spindle end" and 0.03 mm "at the place 100 mm away from the spindle end". Essentially, if I read the diagram correctly, Tormach chucks up a test bar -perhaps with an R8 collet, perhaps a test bar with an integral R8 fitting- and then measures close to the end of the spindle and then 100 mm away.

0.02 mm is 7.87 E-4 inches.

That's not how the OP measured spindle run-out if I understood correctly. Measuring runout per the inspection document will certainly nail down whether the spindle is in spec or not.

Tormach's inspection report on my 1100 said mine was 0.015mm at both locations. Since I don't have a test bar, I can't replicate what I think Tormach did exactly.

I dont have a runout test bar. But runout at the taper with an indicator is a pretty good indicator that runout exists.

[eltejano]

Thank you, everyone, for your input.

[GLCarlson]

I dont have a runout test bar. But runout at the taper with an indicator is a pretty good indicator that runout exists.

Absolutely.

I was pleased to hear that Tormach is shipping you a new spindle. Hope that solves it. It's also a data point that suggests Tormach hasn't changed its past support practices, as that's exactly what I would have expected them to do.

[CountrySmith]

MY 770 got a replacement spindle cartridge under warranty due to a bearing failure. I had experienced severe run out with the replacement. After some detective work and a session at Tormach, we determined that the cause of the runout was burrs left on the spindle face from the holes drilled to balance the rotor. Stoning the surface greatly reduced the runout but some remains. As a test, I mounted a 3/4" bar in the collet and checked runout at about 15mm and 90mm from the face of the spindle. I interpret low runout near the spindle face but higher runout some distance away as having an angular component. Sweeping the tapered surface will not necessarily detect angular runout.

I purchased a 3/4" collet made by Crawford with a specified .0002" TIR, with the intent of eliminating the Tormach collet as the source of the runout. Measured runout was not significantly improved. I also checked the runout on my Taiwan mill/drill and found it to be less than half that of the 770. I blued my collet to check for contact points and found the contact on the taper was primarily at six points, one on either side of the three slits and nearest the spindle face. IMO, the socket taper is not ground correctly. In the course of meeting with Tormach, the service tech stated that they used an R8 test bar to measure runout. Since my spindle was a replacement, it had never gone through the final QC tests so I have no idea as to what the final inspection results were, if any.

While some information can be obtained by mounting a test bar in the Tormach collet, it is not possible to completely parse out where the problem occurs. It could be at the spindle-R8 interface, the interface between the collet and the test bar, or the test bar itself could be bent. At that point, I decided that I needed to make a single piece test bar. I designed one with the intent of making my own but other things have come up since and I have just lived with the runout.

However, I have realized that a runout test bar could be made fairly easily and accurately. I would use an R8 end mill holder, 5/8" or 3/4" in which I had cut a second groove opposite the original and in the R8 TTS collet. I would insert an appropriately sized bar, fix it with Loctite and lock it down. I would then mount it in the Tormach spindle and use the Tormach and a single point lathe tool to turn the bar concentric to the spindle axis. With care, the runout should be near zero. I would then remove the R8 tool and rotate it in the spindle. If there is any runout in the R8 socket, it will now be apparent and will be twice the actual runout. It will of course be specific to my machine but that's OK. It will allow me to finally determine whether my spindle socket is at fault.

[Steve Seebold]

When I got my 1100 5 1/2 years ago I ran some parts at 5,000 RPM and the finish looked almost like I did them with a hammer and a cold chisel.

I called Tormach to ask what was going on and Greg Jackson told me at 5,000 RPM the machine sets up a harmonic and there is a vibration resulting in a poor finish. His suggestion was to ruff at 5,000 RPM and finish at 4,000 or less.

I did that for a while, the I got a wild hair in my butt and I changed the belt. I got rid of the belt that came on the machine and I replaced it with an equivalent Gates belt. I think the belt was about $20.00, but the spindle vibrations went away and now I get a beautiful finish at any speed. And I do a LOT of work with end mills .010 in diameter.

[AsterMachining]

I did that for a while, the I got a wild hair in my butt and I changed the belt. I got rid of the belt that came on the machine and I replaced it with an equivalent Gates belt. I think the belt was about $20.00, but the spindle vibrations went away and now I get a beautiful finish at any speed.

Was that the initial finish out of the box, and a new quality belt fixed the issue, or was the belt a recent fix, i.e. your belt was worn and due for replacing?

[Steve Seebold]

MY 770 got a replacement spindle cartridge under warranty due to a bearing failure. I had experienced severe run out with the replacement. After some detective work and a session at Tormach, we determined that the cause of the runout was burrs left on the spindle face from the holes drilled to balance the rotor. Stoning the surface greatly reduced the runout but some remains. As a test, I mounted a 3/4" bar in the collet and checked runout at about 15mm and 90mm from the face of the spindle. I interpret low runout near the spindle face but higher runout some distance away as having an angular component. Sweeping the tapered surface will not necessarily detect angular runout.

I purchased a 3/4" collet made by Crawford with a specified .0002" TIR, with the intent of eliminating the Tormach collet as the source of the runout. Measured runout was not significantly improved. I also checked the runout on my Taiwan mill/drill and found it to be less than half that of the 770. I blued my collet to check for contact points and found the contact on the taper was primarily at six points, one on either side of the three slits and nearest the spindle face. IMO, the socket taper is not ground correctly. In the course of meeting with Tormach, the service tech stated that they used an R8 test bar to measure runout. Since my spindle was a replacement, it had never gone through the final QC tests so I have no idea as to what the final inspection results were, if any.

While some information can be obtained by mounting a test bar in the Tormach collet, it is not possible to completely parse out where the problem occurs. It could be at the spindle-R8 interface, the interface between the collet and the test bar, or the test bar itself could be bent. At that point, I decided that I needed to make a single piece test bar. I designed one with the intent of making my own but other things have come up since and I have just lived with the runout.

However, I have realized that a runout test bar could be made fairly easily and accurately. I would use an R8 end mill holder, 5/8" or 3/4" in which I had cut a second groove opposite the original and in the R8 TTS collet. I would insert an appropriately sized bar, fix it with Loctite and lock it down. I would then mount it in the Tormach spindle and use the Tormach and a single point lathe tool to turn the bar concentric to the spindle axis. With care, the runout should be near zero. I would then remove the R8 tool and rotate it in the spindle. If there is any runout in the R8 socket, it will now be apparent and will be twice the actual runout. It will of course be specific to my machine but that's OK. It will allow me to finally determine whether my spindle socket is at fault.

If you insist on replacing the collet, I ONLY use Hardinge collets no matter what I need. Yes Hardinge collets cost considerably more than Chinese imports, but they are well worth the difference.

[CountrySmith]

If you insist on replacing the collet, I ONLY use Hardinge collets no matter what I need. Yes Hardinge collets cost considerably more than Chinese imports, but they are well worth the difference.

When I looked at specs for a replacement R8 collet, Crawford had the lowest specified TIR. Hardinge may be better but I couldn't find a TIR spec for their collets.

I actually rechecked the Crawford collet this morning and TIR is less than .00005" at 5mm and 20mm from the collet face.

BTW, Crawford is a UK product.

[RCaffin]

Runout test bars are $$.
Try a name-brand round bit of (ground) tool steel - say 10.0 mm or 12.0 mm. Very very close!

Cheers
Roger