G0704 missing thrust washed on left hand x axis feed

[MAlcocer]

And there in lies the problem. If you read what i actually posted you would see the backlash is over the spec that is recommended and actually is very inconsistent. I attempted to "adjust the machine the way it was designed". I just find the engineering reason you mention difficult to grasp considering some of the other " engineering" choices that were made on this machine.

Further more I challenge you to show me a day when the temp here went from 20°-105°. The daily temp swing here is tops 30 degrees even on record days.

But thanks for chiming in, calling me an amateur. And telling me I need to "learn to be a machinist". That was very productive of you and helped to resolve the issues that I mentioned but you didn't bother to read.

 

[Baithog]

And there in lies the problem. If you read what i actually posted you would see the backlash is over the spec that is recommended and actually is very inconsistent. I attempted to "adjust the machine the way it was designed". I just find the engineering reason you mention difficult to grasp considering some of the other " engineering" choices that were made on this machine.

Further more I challenge you to show me a day when the temp here went from 20°-105°. The daily temp swing here is tops 30 degrees even on record days.

But thanks for chiming in, calling me an amateur. And telling me I need to "learn to be a machinist". That was very productive of you and helped to resolve the issues that I mentioned but you didn't bother to read.

Sorry that you took offense. That was not my intention. Yes I did read everything you wrote.

Your original complaint was .006 of backlash. That is in spec for that machine. Later you talk about "a good part of a rotation", but that is after you started messing with the bearings.

If you now have much more than .006 and it is inconsistent, then it is not in the nut, and tightening the nut isn't likely to fix it.

Put it together without the nut and your bearing modifications. Check the axial play in the lead screw. It should be minimal and consistent. My mill has been converted to CNC, so I can't go out and see what a correctly working thrust bearing pair has for axial play. I would think that it would be less than .001. If you have too much axial play in the bearings, then call Grizzly and get replacements. Grizzly has been really good about taking care of problems that crop up with low cost manufacturing.

Fixing both ends of the lead screw is a bad idea. I will restate - It is not the daily temperature swing that gets you. It is adjusting the bearings on a cold day in February and then trying to run the machine on a hot day in August, without having the machine in a tightly temperature controlled environment, or you re-adjusting the bearings on a regular schedule. Steel moves, and a surprising amount. The Grizzly machine isn't the only one with the lead screw fixed at only one end. Even old Iron like Atlas and South Bend don't support both ends, unless the screw is long, and then only with one end axially fixed and the other with radial control only. Go take a look at some of the CNC router designs. You will be hard pressed to find one that has both ends of a screw fixed axially.

When you get the necessary parts and fix your mill so that it runs as designed (.003-.006 backlash), then learn to live with it. I have built prototype fixtures on knee mills that had twice that much backlash. Make things. Its far more rewarding than trying to make a machine do something it wasn't designed to do.

 

[bladehunter]

Larry gives some good advice, put it back together as it was delivered without the nut installed and check for any axial play.

Rather than chase for minimal backlash, which wont last long with the standard ACME screw and nut setup invest in a DRO or mark your stock on the bench, centre punch any holes and sneak up to any lines checking with a caliper or micrometer.

I don't know how the Grizzly hand wheels are marked, metirc or that other weird system, but your leadscrew could be metric and the hand wheel markings a very rough approximation.

I think I heard mention of some of the early 7x mini lathes having metric leadscrews and imperial marking on the hand wheels.

 

[John Hasler]

It does help. Thank you for all of the help. That is the but I adjusted. I could not resolve the backlash to less than .006 with that but which is why I started looking elsewhere for a solution. When I adjusted screws 67 it only made it more difficult to turn the lead screw. I will go back over it today and see if I missed something.

All that aside wouldn't it take a good amount of heat or temp variance to cause the lead screw to stretch or compact? I live in California and our typical temp swing in this area is very minimal. As an example, today we will be going from 49° to 64°

What matters is the differential thermal expansion between the steel leadscrew and the cast iron bed. By my calculations that comes to 1.2 ppm per degree F for a difference of .0024" over an 85 degree F range on a 24" screw. That's assuming grey cast iron. I think that the semisteel used in most of these machines would be much closer to steel.

I found that much of the backlash on my Grizzly table was due to the handwheels shifting slightly on the shaft. They were keyed but could not be tightly locked down because the same nuts that retained them set the bearing preload.

My table has simpler bearings than yours but a similar design. Both handwheel nuts affect the preload: the leadscrew is held in tension between the two thrust bearings.

 

[george wilson]

Backlash is something that everyone has to live with,unless you have a CNC machine(an expensive one) that uses ball screws. Most CNC machines are programmed to adjust for backlash anyway.

Backlash does no harm. Say your dial has .008" of backlash on it,and is set to the number 10 on the dial. You back the dial up,reset the X axis where you want to re start a cut. Then,turn the dial BACK to the number 10. Then,advance the dial from there to take however much you want to cut,then proceed with the cut. No biggie. I do it even on my Hardinge HLVH,which is one of the most expensive small lathes ever made in the USA. Like all machines,it has wear and some backlash. I have a new crossfeed leadscrew for it,but I haven't bothered to install it. That crossfeed leadscrew cost $250.00 years ago. I got 2 of them NOS for $15.00 each,and gave one to my friend,who also has a Hardinge. That's how little I worry about backlash. My Hardinge even has adjustable leadscrew nuts on the compound. I never bother to tighten them. They'd re wear anyway if I did.

Another thing that happens inevitably over time,is the leadscrew gets more worn in the middle than it gets on the ends. Adjusting the nut to be tighter in the middle of the leadscrew just makes it TOO TIGHT as you approach either end of the leadscrew. This rapidly wears out the nut,and you get much more backlash than when you started.

The poster who advised you to learn to be a machinist might have sounded a little harsh,but you admitted you are a beginner. I used to get talked to that way when I was starting out,by an old time machinist who certainly was a friend and mentor. He just had a helpful but gruff way about him. I learned a lot from him,and still use things he taught me 50 years ago. When I asked a stupid question ,he'd say "Then you are a poor machinist". And,I WAS at that time!!

Bottom line is: backlash is always going to be present,no matter how much backlash is not important. Just reset your dial like I said,and proceed from there,no harm done.

Say you want to drill 2 holes a certain distance apart on your mill. There is .016" backlash in the dial. You know this. Remember this. rotate the dial .016" in the direction of the 2nd. hole you want to drill. Then,loosen the dial and set it to zero. Then,go ahead and rotate the dial however much you need to to set the distance between the holes. Proceed to spot drill and drill.

I hope this makes sense. This is what you have to do,period. Even if you eliminate all backlash,it will soon start to return because the leadscrew and the nut will wear,no matter what you try to do.

I advise you to take the advise of experienced machinists here. After all,you ARE trying to learn to be a machinist,aren't you? I wanted to throw my first lathe off a cliff a lot of the time. It was not a good machine to begin with,which made my learning curve even more difficult. But,I always took the advice of experienced machinists,even if it wasn't sugar coated. And,I learned how to be a machinist.

 

[John Hasler]

So let me get this straight.

Despite having found a solution to the issue that gets the machine back into spec and eliminates the excessive and inconsistent backlash (thanks to John Hassler, bladehunter and tmarks, the issue was the right side bearings were not preloaded and the leadscrew was just floating under tension between the too ends which created the inconsistency.

Tensioning the leadscrew between the two ends should work, though, if that was the intent of the original design. It should preload both bearings. I wouldn't worry about the temperature thing. If you set the preload to optimum in the sixties you'll gain maybe a thou when it's damn cold and lose maybe a thou when it's hotter than hell.

BTW if what another member writes irritates you just don't read it, ok? You don't need to defend yourself, much less counterattack.

 

[HMF]

It looks like a solution has been achieved.

Unfortunately, a misunderstanding has also occurred and things have become personal. We can't have that on here.

Therefore, I am closing the thread so that things can cool down.

Thanks for understanding.