[Thomas Utley]

All,

I've searched the forums and found conflicting inputs on the suitability of using Acetal (trade name: Delrin) anti-backlash nuts. I'm designing my first CNC router around a bunch of salvaged hardware. Good quality Thomson linear bearings and round rods, etc., but everything's heavier than I would have preferred.

The moving portion of my Z-axis assembly will weigh in somewhere between 20 and 30 lbs (Porter-Cable 7518 router motor weighs 14 lbs alone, plus four Thomson bearings and pillow blocks, the spindle plate, spindle mounts, fasteners, and mass of the dust collection apparatus). Call it 25 lbs realistically.

To maximize travel I'm looking at DumpsterCNC's Delrin anti-backlash nuts running on a 1/2-10 single start ACME leadscrew. Their flange size options work well with my packaging constraints.

#1 - Am I going to be disappointed with the life of the nut?

#2 - Can I extend the life of the nut (and performance of the entire Z-axis) by adding a lift spring or gas strut to support some of the moving mass?

#3 - Should I be looking at bronze instead? My Z stepper is, admittedly, a bit undersized at 305 oz-in. For this first build, speed is not the design driver, but missed steps are a non-starter for me.

Thanks,

Tom

[WayneHill]

Tom,

I am using DumpsterCNC Delrin anti-backlash on my router. My router has 1/2 - 10 (5) start ACME leadscrews with a 1:2 ratio gear reduction from the DC servo. The router is a Hitachi M12VC 2-1/4 HP. Using three old Hardinge lathe DC motors for X Y Z axis. Yes, this baby can fly

To answer your questions.
1. No.
2. Depends on many factors as motor size and carriage weight. I am not using any gas assised or spring struts to ballance the load. All is well on the Z axis after a summer of abuse.

3. Start with the Delrin and you will forget about the bronze after a few months.

[InventIt]

I used Delrin nuts on my first router and it worked so well I implemented it on my second router build. First router is using 1/2-10 5 start and has been in use for several years. I am just starting to notice a little (.005) backlash. I use oil on my scews, lube them generously every few weeks with way oil. You will be happy with them.

[harryn]

Hi, I hope you don't mind if I ask another delrin related question in your thread, in particular using PTFE fill delrin as a plain bearing material. The application would be for shafts that rotate - nominally 1/2 - 1 in dia.

The other application would be 1in long x 1 in dia x 1/2 bore pieces to use instead of skate bearings in linear motion trucks.

Any comments on suitability or better / worse than plain delrin / acetyl ?

Thanks

Harry

[Thomas Utley]

Harryn,

No worries on the piggyback post--we're all here to learn from each other, and I was actually hoping you'd get some answers to your post as well.

Thanks to all the others on my original post. I've ordered the Dumpster A-B nuts and we'll see how it works out. Worst case is it works well enough and long enough for me to find all the flaws in my system's design, and a little bit longer so I can use it make parts for the next one!

Take care,

Tom

[mrehmus]

Turcite is what they use on machine tool slides. There is also a 2-part expoxy-like product called moglich which can be cast directly onto machine parts for things like slides and drive nuts.

It may not be as good as Delrin in a wood dust environment but for milling machines, lathe rebuilds and new big ticket CNC machines, it performs well.

[turboedgs]

Dumpster cnc’s anti backlash nuts are the way to go with a hobby cnc. I’ve used both cncrouterparts.com and dumpster’s. The CNC Router Parts nuts are...ok. The major issue in the design is they use a rubber o-ring to compress the threads into the rod. In two months the o-rings dried out and cracked causing major backlash. You can of course easily replace the o-ring as I had to do three times. I got sick of taking things apart to replace a broken/dried up o-ring so I bought a few of the Dumpster nuts and have had no issues at all to date, about 6 months run time.
Just my .02 cents

[ger21]

Try a different O ring material. I've got a homemade AB nut using an O ring and it's going on several years now. Not sure what it's made of. It could also be that lubricant is breaking down the O ring. Again, a different material might last a lot longer. McMaster Carr has a lot of different types.

[mike_Kilroy]

Being a Thompson distributor, I wanted to just add a note to this thread that as such a distributor who sells similar material nut material with screws, there are just 2 things we watch really close in an application with them: speed and load. As long as the load forces do not exceed the nut rating, and the speed does not exceed around 500rpm max, life should be similar to ballnuts. But if you exceed these speeds or the rated loads, the friction causes heat that kills the nut fast. So just watch these two ratings and go for it!

[mrehmus]

Does Thompson have a reasonable cost solution for TAIG Mills?

[mike_Kilroy]

Thomson does not have machine specific kits, just sell parts. you can check out Home | Danaher Motion | International or Thomson Home for info on their stuff and see who your local distributor would be to call if you want.

[mrehmus]

Ah, then the next question is whether anyone has modeled the TAIG mill in MCAD. This would make it much faster to be able to design a new set of leadscrews and anti-backlash nuts. From what I can see with a fairly quick glance, outfitting a TAIG with their products will be a bit iffy because the application is at the low end of their size ranges for the truly zero backlash nuts.

Another question would be, is there an engineer out there who has or who will calculate the load the TAIG mill can place on the leadscrew/nut assembly? Clearly, overloading the system will rapidly wear out one of those zero backlash nuts.

[harryn]

Hi, its not that tough to calculate loads for any mill like this.

From the taig web site, the "new" version is 200 oz in of torque. I don't know the thread size, but suppose it is 10 turns / inch, then

200 x 10 = 2,000 oz force max.

Of course, after doing the work to upgrade it, you still have a very low capacity tool, and probably will want a bigger one.

[ArtCAMinski]

Ah, then the next question is whether anyone has modeled the TAIG mill in MCAD. This would make it much faster to be able to design a new set of leadscrews and anti-backlash nuts. From what I can see with a fairly quick glance, outfitting a TAIG with their products will be a bit iffy because the application is at the low end of their size ranges for the truly zero backlash nuts.

Another question would be, is there an engineer out there who has or who will calculate the load the TAIG mill can place on the leadscrew/nut assembly? Clearly, overloading the system will rapidly wear out one of those zero backlash nuts.

Thanks for the questions...I'd certainly like to know as well if there's a precision acme screw/anti-backlash nut that actually fits a Taig mill, while providing better performance and accuracy than the stock parts, at a reasonable cost?

[Bob La Londe]

For you Taig guys... I have one. (Gecko G540 not Taig control) Its a decent little machine for the price, but requires constant adjustment to keep it running with decent amounts of backlash. I would be torn between it and the Max NC except the MaxNC uses their own driver and lighter motors than I like. (I retrofit one to Mach / Gecko control w/ heavier motors).

I have had my Taig for over two years, and I have put hundreds and hundreds of hours on it. (Probably thosuands) Sometimes running as much as 30 hours continuous on a single job at speeds upto 50 IPM for rapids. It has 20 IPM V thread leads screw.

I have had it almost completely apart several times, and the biggest problem it has with the nuts besides needing to be adjusted about every 10-20 hours of run time is the way they are mounted. The have a stub that just seems to press into the machine. ICK!!! On my Z axis it actually started to wear out and was contributing backlash at that point. Double ICK!!! I took it apart and used a shim and pressed it back together to reduce that problem point. The bearings on the lead screw shafts need to be periodically snugged up. Usually about every 100 hours, although it may be more often. I didn't realize that was a problem point right away. Its easy to tell. Adjust the lead nuts, loosen the gibs and grab the motor to push in and out on the lead screw. If you can feel movement its due for another 1/32 to 1/16 of a turn of tightening. Don't over tighten as you will just crush the bearings and it will make a nice grinding noise as it moves.

Anyway... As fas as acetal...

One poster in this thread "as long as you don't exceed 500 RPM." 500 RPM is awfully slow. To slow for this machine IMO. I have not independently verified their number, but here is what I get.

500 RPM / 20 TPI = 25 IPM. 25 IPM is not bad for cutting, but I have my machine set at 50 for rapids on X & Y, and I sometimes cut as fast as 30 or 40 for some stuff. I can run 60 using stepper motors, but that's right at the edge of reliability for control position.

Please note: 50 is a little enthusiastic for the Z axis on my machine. The Z axis is just not a very good design compared to the X & Y and it has the potential for binding. If you are using the stock motor and spindle even 30 is a little optimistic with all that weight. I was upset when I discovered the Z assembly ways were not straight, but Taig told me .01 was acceptable tolerance for them. Designing my own Z-axis for the machine is on my short list. I'm currently running my own mounting bracket and a Bosch router as a spindle. This gives me more power and a lot less weight on that axis. (not good for steel, but awesome for aluminum)

Anyway if 500 RPM is a realistic do not exceed RPM for a 1/2" V groove lead screw then acetal may not be a good choice for this machine. If looking at acetal a change to a 10 TPI acme screw might be in order. That would put my reliable max speed (for steppers) right at the the max RPM reccomended by the OP. Actually repeatability would improve due to the slower RPM of the stepeprs and reduction of lost steps. Obviously with lower pitch you could run faster, but then you would exceed the RPM target suggested.

Yes I do realize this original thread is over a month old.

Just thinking out loud.