Hi John, I haven't read all of the comments so some of this may already have been mentioned. What your doing is checking the tram of the spindle to the table, but if that head isn't bolted down truly square to the column then you still don't have the machine correctly aligned. Setting a 12" GOOD cyclindrical square on the table and a 10ths reading DTI mounted in or on the spindle, then use your Z axis movement to traverse down the square while sweeping across the square every inch or so for the reading will show if that Z axis is at 90 degrees to the table within the slight errors of both the square and the DTI. That has to be done in both X and Y directions to check the column for lean in both directions. Once that's corrected you then use that DTI to sweep the table or vise in both X and Y. At that point you may need to tilt the head for the X axis alignment and possibly shim the head for the Y axis alignment. So there's a bit of tail chasing in this. Even better would be to use no shims and scrape the head into the correct alignment for the best rigidity. Shimming the column for tram removes a lot of rigidity since the foot print is greatly reduced. Surface grinding on both the bed area and the column foot to get it very close and then both surfaces should be clean and smooth. Then some minor hand scraping to bring it into the last bit for a true alignment would be the best method. A LOT of work with all the trial and error checks and fitting. But there's no short cuts if you want it done right. I only use a 1 or 1/2 thou reading indicator to get me in the ball park. Then move to a 10ths DTI so I know what my readings really are. Due to the flex involved with flycutters in the whole machines system including a very minor amount from the spindle bearings themselves you'll always see a very slight back cut even on a perfectly trammed machine. It's impossible for them not to show that effect on the lighter built machines. If you don't see it that means your machine is out in the other direction, or trammed with the flycutters trailing rotation tilted up and away from that direction. And the larger your flycutter then the more pronounced that back cutting effect is. I know nothing about the Tormack machines yet, or about CNC at all. But these lower end machines aren't properly cleaned at the factory or assembled in anything close to clean room conditions. Even on a brand new machine the table should be pulled and feed screws, nuts way surfaces flushed with a spray degreaser. CRC electrical contact cleaner works very well for the ACME nuts on my mill, I'm not sure how well it would work with those ball nuts though. I do know it's not a good idea to remove the ball nut from the screw for those who don't know that yet. Done incorrectly you'll be chasing the balls all over your shop. And given your machine is doing production I'd highly stress that pulling the table every 6 months or so would be a very good idea. You will find all sorts of chips in where they shouldn't be, but the way oil will be deep black from wear just due to the parts wearing together. That black oil contains metal and further accelerates the wear. Clean ALL that out, inspect your lube system so you know all areas are getting properly lubricated, double check the tightness on the bolts for the ball nuts, relube, then reassemble and do your gib and screw adjustments for tension. Yes all this is a lot of work, but you will be surprised at what you find under that table. And your machine will work far smoother, much more accurately, and do so for far longer. I'd also bet you can get your adjustments a little tighter for better accuracy but still have the same amount of drag between the gib and the way surface. My BP Taiwan built clone came from the factory with .020" of backlash and was fairly stiff to move. After the strip down, cleaning, relube and proper adjustments I had .004" with a silky smooth movement. Not too bad for a manual machine with ACME screws and about what a brand new Bridgeport will have. While I didn't find any casting sand I did find just enough metal and grinding dust from the surface grinding I was very glad I took the time to do all this. My apology's for the length of this, but you did ask for comments.
Hi John, I've seen someone jam a gib really badly doing it your way. The Tormach is low enough power that you got away with it, but it took a hydraulic jack to free off a big machine and people have destroyed Bridgeport style machines by doing it on the knee (if the knee is too low you can't get to the screw to push the gib out). The propper way is to use a dial gauge to measure the rock in the slide. Keith Rucker did a good video on the task; Machine Shop Basics: Tramming a Vertical Milling Machine
+Paul Compton +1 to the Keith Rucker video showing the proper way to adjust gibs using an indicator with a torsional force. I was very confused by Tormach's method of adjusting the gibs by measuring backlash. The video I watched was titled: Milling Machine Maintenance: Adjusting Gibs and Ways.
John Did you try and mill a circle after you adjustments? The traditional way to check if you gibs and backlash are correct is to mill a id bore and check it with a dial bore gage. The actual size is not critical but what you are looking for is reversal marks at the 90 deg. quadrants (Backlash)and out of roundness at the 45 deg. quadrants(Gibs) with the dial bore gage. there is a instrument called a Ball Bar Gage witch is used to measure this it will also tell you if your axis are square to each other. But it is a expensive tool. lacking that milling a bore with a end mill in a scrap piece of material is almost as good. A modern CNC machine should be capable of holding less than .001" roundness to the circle. I don't know Torchmach Machines but there should be a backlash adjustment parameter in the control to compensate for backlash in the ball nut. O one other thing the proper way to adjust a gib on a CNC mill as to measure the current draw on that axis servo motor.
Very informative and inspirational. Maintenance often gets overlooked, but you've reminded us of its benefits. Great time lapse at the end. I often record teardowns so I know where that last 1/4 28 came from!
When I make adjustments I understand that it is as much of a science as it is an art. Try using an amp meter on a motor lead to give you an indication of motor load.
I don't go as far as seizing the steppers. I loosen it, then tighten until a 1-thou step does not make the table move a full thou. I back off just barely from there. It requires more maintenance than keeping it tighter but that's my preference.
This is where the Tormach deviates from traditional C frame mill design. The thrust bearing preload should be set with a shim, that's it. The bearings are manufactured with precision shoulders, you put a shim between the inner races and make sure the outer races are touching, that's it. This shim creates a preload of the bearing races and keeps them from having any lost motion. The thrust bearings on the ballscrews are just smaller versions of the bearings in the spindle, they are both angular contact ball bearings. You shouldn't need to fiddle with preload nuts, but Tormach decided to install the bearings in a backwards orientation. They use a shim on the OD of the race and use the preload nuts to establish the preload between the inner races. The way to fix this is to reverse the bearings in the bore and put a thin shim between the inner races, then the "thrust nuts" only retain the screw in the bearings, they don't adjust anything. The use of 2 ball nuts to create a preload is used traditionally with rolled screws and mass produced ball nuts. Looking at the parts breakdown, the Tormach uses single preloaded ballnuts, not 2 ballnuts installed in opposition. So, Tormach's design is *similar* to other vendors, but there are a couple of things that make me scratch my head and ask "why did they do it that way?".
To War Machine: put a indicator holder in the spindle then tram x and y in the same set up by spinning the spindle by hand. THIS IS THE BEST WAY TO TRAM ANY MACHINE. IF THE COLUMN IS NOT PERPENDICULAR TO THE TABLE THEN YOU MUST FIX THAT FIRST BUT ALWAYS END UP TRAMMING THE TABLE WITH AN INDICATOR BEING HELD IN THE SPINDLE WITH A HOLDER.
While I do not (yet) have the Tormach, I can tell you the processes that many mills benefit from. All screw drives need to have the preload nuts tightened, typically to a predefined torque. In addition, the carriage itself will (in most cases) have an adjustment for the part that drives the table. Most ballscrews have a way to increase the preload of the balls inside, reducing backlash. Same with standard ASME screws. A screw or spacer spreads the engagement teeth so that both the front and back face are in contact. Ball screws benefit from an extremely low friction though, and the best ones use a gothic arch thread profile so each ball has at minimum 2 points of contact. Properly loaded, they have 4 points due to opposing gothic arches. Overall though... When you said you werent sure ho much you were missing... Less that 1 tenth. ;-) Started at .0003, ended around .0003. Well done!
John your right ... gibs need to by tight , but not too much , every part of machine has some flex, if gibs are too tight ball screw, and nut moutn will start flex and backlash start show on.
I wonder what the Tormach tram spec is. When I set up a machine, I like to get under .0005 TIR on a 10" circle. I can live with .001. Interesting that you could adjust the nut on the ball screw and reduce the backlash. On a "real" CNC machine, the thrust bearings would be angular contact bearings and the pre-load is not adjustable. I did work on a CNC knee mill one time that used tapered roller bearings. It must have been adjustable.
John, well done, informative, insightful video as always. Happy that you are happier with your machine now compared to a month ago prior to the maintenance process. I agree that the stepper losing steps shouldn't happen in this mode. But watch out for wear in the middle of the travel range vs travel at the ends. At certain gib tightnesses, you might be OK in the middle of the travel but could lose steps or seize altogether at the ends because of less wear at the extremes of travel.
Hey John, Great video. I went through all this same painstaking tramming myself. It took days to get it dialed in just right but I could have saved myself a lot of time with this tip. Tram the head to the column FIRST, then tram the head to the table. There is a big difference. When I initially measured my table tram it was off in the Y so I shimmed the column until it was dead on. But when I cut or drilled I still had long drills binding and squealing, holes not round, etc. How could this be? I trammed it. Well turns out the spindle was not square to the column it was essentially pointing backwards toward the base so I was trying to make round holes with a crooked stick. So I shimmed the spindle cartridge (you could probably also shim the head) until the Z travel was parallel with the spindle axis in Y. After that I put the Edge tramming tool back in to check the table tram and it was off by the same amount that the spindle was off. My table tram had been perfect and I screwed it up, it was the spindle that was out of square not the table. After removing the shim from the column and playing around with smaller shims and bolt tightness I got the table back in tram. So now the spindle is trammed to the column and the column to the table. You can check the tram of the spindle to the column by putting a long piece of drill rod into the collet. Attach your mag base to the column and put the DTI on the top of the rod right as it comes out of the collet. Rotate the spindle by hand. You will notice some runout, that's OK just center the zero between the needle swing so it is equidistant on each side of zero. Then run the head straight up in Z until the DTI is on the very end of the drill rod and rotate it by hand again. The magnitude of the runout will increase but the amount the needle is off to one side (subtracting runout) is the amount out of tram it is. You can do this for both X and Y. The idea is to get the needle balanced around the zero equally both at the top and the bottom of the drill rod. Then you can proceed to tram the spindle to the table as you did with the Edge tram tool.
+War Machine LLC This is a more direct method of measuring the parallelness of the spindle to the z axis. I like it a lot. One comment. You don't have to mount the mag base to the column. It can be place on the table as well. That way. you can use the table for precision alignment of the DTI. All that is required is that the DTI be parallel to the x axis and to the y axis for the x and y measurements respectively. I used a similar method for measuring squareness of the z axis to the xy plane by placing the rod in a chuck on a horizontal RT and rotating the table to find min and max.
+War Machine LLC Thanks for this. I took the time to tram head to column and then column to table. The head must be in tram with column and needs checked first. Thanks for the heads up and all is well now with the deep drilling.Dennis
Absolutely AWESOME!!! I still think Servo Motors would minimize a lot of the adjustments... Under High Torque loads, Steppers can loose steps, Servos will not... But that's my observation... Servos are a lot more expensive, so there's always a trade-off... Sorrta 6 of one - a half dozen of the other... Fantastic video... It is really important for calibrating a machine. I'm so happy to be a subscriber... I wish I could support you financially, but still can't at this point in time... I have to make due with what you post and for the most part it's all been an awesome learning experience :-) When I am better situated, I will definitely support your work :-) Thanks again :-)
+DStray Cat, Well if I were making a DIY mill for metal like the Tormach. Servos like the ClearPath by Teknic would be a good option -- David Burrell at UA-cam channel Neo7CNC really likes his for his router. But the benefit of something like the Tormach is that you have a system designed to work together with support. Also steppers are almost always oversized precisely in order to not lose steps or stall.
+DStray Cat The upgraded motor package for Tormach machines, which I think John has fitted, uses Leadshine hybrid steppers which have an encoder and a closed loop to their matching controller. They will flag a 'following error' like a servo drive if they skip to many steps and can't correct.
Hey mate. Quick question. For milling steel if you were building a milling machine what spindle would you go for. ER32 collet I prefer over R8. Would you go for a spindle with built in motor or a spindle with pulley and external motor?
I was able to shim in the X tram, no problem. What is the best way to adjust the Y tram? By the way, I zero the Edge gage in the Y axis so that I don't need a mirror. It does not matter where you put the magnet as long as you don't move it during the zero process.
Seems to me that's it would be a bit better practice to mount your tram tool directly in the spindle with an R8 collet.Holding a tram tool in a setscrew holder isn't really that bad, but for optimal results its not best.The inherent nature of a setscrew holder to rock the tool to the side abit when the setscrew is tightened down is not what you want when your inspecting your machine like this. Your thoughts?
Great video. I'm just getting started on a Tormach, and I cannot figure out how to lock the spiundle. My PCNC-1100 doesn't have the operator's panel either.
John I'm a little puzzled as to why both you and Tormach are interchanging leadscrew backlash and gib adjustment. I'm not at all familiar with the Tormach, but I believe the two should be completely independent (well to a point the table is jus flopping around). Were the gibs far out of adjustment? About 6 months ago I wrote in a comment that I felt they were loose and should be adjusted. I'm now curious to know if that was correct. To adjust the gibs on my machines I normally try to completely disconnect the leadscrews so I can move the tables etc by hand. I then make the gib adjustment by moving the table through the full range of travel and feeling the tightness at the gibs by the drag. Generally you want to be careful out at the extremes of travel as the machine doesn't normally get used as much out there, and the gibs can finish up tight if you adjust only for 0,0 and don't check the extremes. I get much better feel for the adjustment by moving things by hand. I have no idea if that's possible on your machine. Finally, does the Tormach have a backlash compensation function within the firmware? Once you've exhausted the mechanical backlash, and if the remaining backlash is consistent and predictable, I think you should be able to take out most of the remaining in software. I have absolutely no idea however how, or even if, it's possible on your machine, but the Tormach people should be able to advise. It certainly is however available on other CNC platforms. Great to hear it's all working out.
+NYC CNC Sometimes it's not necessary to actually remove the screw, and you can just remove the hardware at the ends that supports and controls its backlash. It does feel like you're going backwards to move forward, however in my experience that sweet spot is very narrow, especially as the machine wears more and you will battle getting things "just right" (or at least as good as they can get). Done in the correct sequence (ie first) adjusting the gibs this way shouldn't add that much time to the project. Just be grateful you have a tapered gibb and aren't trying the same with a series of gibb screws!!! Anyway, it's done now and all is right again. Good to hear that the machine is running sweetly once more. I do think 5 years is pushing it a little for a machine like that in heavy use, and a once year check and adjustment should keep a smile on your dial.
+Pete F To adjust the bearing preload on my 770, loosened the gibs to reduce any friction there,pulled the motor, and replaced it with a torque gage capable of measuring inch-ounces. I loosened the bearing preload and measured the torque required to turn the ball screw at 21 in.-oz. which would be due to way friction and ball screw preload. I then tighten the bearing preload nut to a torque of 38 in.-oz. This gave me a backlash/lost motion of .0004". This is also a good way to adjust the gibs substituting a torque reading for the "feeling of tightness".
+Pete F The torque gage is rather simple in design. It is a flat pulley 2" in diameter with a length of fine cord wrapped around. The loose end of the cord is attached to a spring scale reading 70 oz. max. A steady pull on the scale will unwind the cord providing torque to rotate the shaft. Because of the 1" radius, the scale reads torque in in. oz's. directly.
Hey NYC CNC I have a question, because land cost is so high in NYC, there traditionally isn't much manufacturing or prototyping jobs out there for machinists. What kind of jobs do you do usually do in NYC as a machinist?
can you tell me the size of the ballscrews in the X Y & Z? It would be interesting to know what size they use in relation to the amount of weight the lift / Move on the axes, this info isn't on the tormach website.
You are a very smart man. If I am not already I am going to subscribe to your channel. I have a robo drill / technician related question. I am merely a machine operator so I don't really know what I'm talking about, but I would like your help! If I make a video where I explain to you everything that I do know do you think that you could make sense of the situation and maybe say something to me that I could say to one of my technicians at work that would make everything click and then they would implement the proper programming to fix this error that has occurred???
Hi John,
I haven't read all of the comments so some of this may already have been mentioned. What your doing is checking the tram of the spindle to the table, but if that head isn't bolted down truly square to the column then you still don't have the machine correctly aligned. Setting a 12" GOOD cyclindrical square on the table and a 10ths reading DTI mounted in or on the spindle, then use your Z axis movement to traverse down the square while sweeping across the square every inch or so for the reading will show if that Z axis is at 90 degrees to the table within the slight errors of both the square and the DTI. That has to be done in both X and Y directions to check the column for lean in both directions. Once that's corrected you then use that DTI to sweep the table or vise in both X and Y. At that point you may need to tilt the head for the X axis alignment and possibly shim the head for the Y axis alignment. So there's a bit of tail chasing in this. Even better would be to use no shims and scrape the head into the correct alignment for the best rigidity.
Shimming the column for tram removes a lot of rigidity since the foot print is greatly reduced. Surface grinding on both the bed area and the column foot to get it very close and then both surfaces should be clean and smooth. Then some minor hand scraping to bring it into the last bit for a true alignment would be the best method. A LOT of work with all the trial and error checks and fitting. But there's no short cuts if you want it done right. I only use a 1 or 1/2 thou reading indicator to get me in the ball park. Then move to a 10ths DTI so I know what my readings really are.
Due to the flex involved with flycutters in the whole machines system including a very minor amount from the spindle bearings themselves you'll always see a very slight back cut even on a perfectly trammed machine. It's impossible for them not to show that effect on the lighter built machines. If you don't see it that means your machine is out in the other direction, or trammed with the flycutters trailing rotation tilted up and away from that direction. And the larger your flycutter then the more pronounced that back cutting effect is.
I know nothing about the Tormack machines yet, or about CNC at all. But these lower end machines aren't properly cleaned at the factory or assembled in anything close to clean room conditions. Even on a brand new machine the table should be pulled and feed screws, nuts way surfaces flushed with a spray degreaser. CRC electrical contact cleaner works very well for the ACME nuts on my mill, I'm not sure how well it would work with those ball nuts though. I do know it's not a good idea to remove the ball nut from the screw for those who don't know that yet. Done incorrectly you'll be chasing the balls all over your shop.
And given your machine is doing production I'd highly stress that pulling the table every 6 months or so would be a very good idea. You will find all sorts of chips in where they shouldn't be, but the way oil will be deep black from wear just due to the parts wearing together. That black oil contains metal and further accelerates the wear. Clean ALL that out, inspect your lube system so you know all areas are getting properly lubricated, double check the tightness on the bolts for the ball nuts, relube, then reassemble and do your gib and screw adjustments for tension. Yes all this is a lot of work, but you will be surprised at what you find under that table. And your machine will work far smoother, much more accurately, and do so for far longer. I'd also bet you can get your adjustments a little tighter for better accuracy but still have the same amount of drag between the gib and the way surface. My BP Taiwan built clone came from the factory with .020" of backlash and was fairly stiff to move. After the strip down, cleaning, relube and proper adjustments I had .004" with a silky smooth movement. Not too bad for a manual machine with ACME screws and about what a brand new Bridgeport will have. While I didn't find any casting sand I did find just enough metal and grinding dust from the surface grinding I was very glad I took the time to do all this.
My apology's for the length of this, but you did ask for comments.
Hi John,
I've seen someone jam a gib really badly doing it your way. The Tormach is low enough power that you got away with it, but it took a hydraulic jack to free off a big machine and people have destroyed Bridgeport style machines by doing it on the knee (if the knee is too low you can't get to the screw to push the gib out).
The propper way is to use a dial gauge to measure the rock in the slide.
Keith Rucker did a good video on the task; Machine Shop Basics: Tramming a Vertical Milling Machine
+Paul Compton +1 to the Keith Rucker video showing the proper way to adjust gibs using an indicator with a torsional force. I was very confused by Tormach's method of adjusting the gibs by measuring backlash. The video I watched was titled: Milling Machine Maintenance: Adjusting Gibs and Ways.
Paul Compton I
John Did you try and mill a circle after you adjustments? The traditional way to check if you gibs and backlash are correct is to mill a id bore and check it with a dial bore gage. The actual size is not critical but what you are looking for is reversal marks at the 90 deg. quadrants (Backlash)and out of roundness at the 45 deg. quadrants(Gibs) with the dial bore gage. there is a instrument called a Ball Bar Gage witch is used to measure this it will also tell you if your axis are square to each other. But it is a expensive tool. lacking that milling a bore with a end mill in a scrap piece of material is almost as good. A modern CNC machine should be capable of holding less than .001" roundness to the circle. I don't know Torchmach Machines but there should be a backlash adjustment parameter in the control to compensate for backlash in the ball nut. O one other thing the proper way to adjust a gib on a CNC mill as to measure the current draw on that axis servo motor.
It's always nice when you get to open things and learn new things.
Very informative and inspirational. Maintenance often gets overlooked, but you've reminded us of its benefits. Great time lapse at the end. I often record teardowns so I know where that last 1/4 28 came from!
When I make adjustments I understand that it is as much of a science as it is an art. Try using an amp meter on a motor lead to give you an indication of motor load.
a clean and maintaind machine is a happy machine
+Alexander Tauriainen If clean counts too, then John's machine is a very sad one ;)
Always feels good to make improvements to your accuracy and it is always worth the time invested. Cheers, Gary
I don't go as far as seizing the steppers. I loosen it, then tighten until a 1-thou step does not make the table move a full thou. I back off just barely from there. It requires more maintenance than keeping it tighter but that's my preference.
This is where the Tormach deviates from traditional C frame mill design. The thrust bearing preload should be set with a shim, that's it. The bearings are manufactured with precision shoulders, you put a shim between the inner races and make sure the outer races are touching, that's it. This shim creates a preload of the bearing races and keeps them from having any lost motion. The thrust bearings on the ballscrews are just smaller versions of the bearings in the spindle, they are both angular contact ball bearings. You shouldn't need to fiddle with preload nuts, but Tormach decided to install the bearings in a backwards orientation. They use a shim on the OD of the race and use the preload nuts to establish the preload between the inner races. The way to fix this is to reverse the bearings in the bore and put a thin shim between the inner races, then the "thrust nuts" only retain the screw in the bearings, they don't adjust anything.
The use of 2 ball nuts to create a preload is used traditionally with rolled screws and mass produced ball nuts. Looking at the parts breakdown, the Tormach uses single preloaded ballnuts, not 2 ballnuts installed in opposition.
So, Tormach's design is *similar* to other vendors, but there are a couple of things that make me scratch my head and ask "why did they do it that way?".
To War Machine: put a indicator holder in the spindle then tram x and y in the same set up by spinning the spindle by hand. THIS IS THE BEST WAY TO TRAM ANY MACHINE. IF THE COLUMN IS NOT PERPENDICULAR TO THE TABLE THEN YOU MUST FIX THAT FIRST BUT ALWAYS END UP TRAMMING THE TABLE WITH AN INDICATOR BEING HELD IN THE SPINDLE WITH A HOLDER.
While I do not (yet) have the Tormach, I can tell you the processes that many mills benefit from. All screw drives need to have the preload nuts tightened, typically to a predefined torque. In addition, the carriage itself will (in most cases) have an adjustment for the part that drives the table. Most ballscrews have a way to increase the preload of the balls inside, reducing backlash. Same with standard ASME screws. A screw or spacer spreads the engagement teeth so that both the front and back face are in contact. Ball screws benefit from an extremely low friction though, and the best ones use a gothic arch thread profile so each ball has at minimum 2 points of contact. Properly loaded, they have 4 points due to opposing gothic arches.
Overall though... When you said you werent sure ho much you were missing... Less that 1 tenth. ;-) Started at .0003, ended around .0003. Well done!
I love my Pro tram best thing ive bought so far for tramming ... soooo easy
John your right ... gibs need to by tight , but not too much , every part of machine has some flex, if gibs are too tight ball screw, and nut moutn will start flex and backlash start show on.
I wonder what the Tormach tram spec is. When I set up a machine, I like to get under .0005 TIR on a 10" circle. I can live with .001.
Interesting that you could adjust the nut on the ball screw and reduce the backlash. On a "real" CNC machine, the thrust bearings would be angular contact bearings and the pre-load is not adjustable. I did work on a CNC knee mill one time that used tapered roller bearings. It must have been adjustable.
John, well done, informative, insightful video as always. Happy that you are happier with your machine now compared to a month ago prior to the maintenance process. I agree that the stepper losing steps shouldn't happen in this mode. But watch out for wear in the middle of the travel range vs travel at the ends. At certain gib tightnesses, you might be OK in the middle of the travel but could lose steps or seize altogether at the ends because of less wear at the extremes of travel.
Thanks for the concepts you have provided here.
Hey John,
Great video. I went through all this same painstaking tramming myself. It took days to get it dialed in just right but I could have saved myself a lot of time with this tip. Tram the head to the column FIRST, then tram the head to the table. There is a big difference. When I initially measured my table tram it was off in the Y so I shimmed the column until it was dead on. But when I cut or drilled I still had long drills binding and squealing, holes not round, etc. How could this be? I trammed it. Well turns out the spindle was not square to the column it was essentially pointing backwards toward the base so I was trying to make round holes with a crooked stick. So I shimmed the spindle cartridge (you could probably also shim the head) until the Z travel was parallel with the spindle axis in Y. After that I put the Edge tramming tool back in to check the table tram and it was off by the same amount that the spindle was off. My table tram had been perfect and I screwed it up, it was the spindle that was out of square not the table. After removing the shim from the column and playing around with smaller shims and bolt tightness I got the table back in tram. So now the spindle is trammed to the column and the column to the table.
You can check the tram of the spindle to the column by putting a long piece of drill rod into the collet. Attach your mag base to the column and put the DTI on the top of the rod right as it comes out of the collet. Rotate the spindle by hand. You will notice some runout, that's OK just center the zero between the needle swing so it is equidistant on each side of zero. Then run the head straight up in Z until the DTI is on the very end of the drill rod and rotate it by hand again. The magnitude of the runout will increase but the amount the needle is off to one side (subtracting runout) is the amount out of tram it is. You can do this for both X and Y. The idea is to get the needle balanced around the zero equally both at the top and the bottom of the drill rod. Then you can proceed to tram the spindle to the table as you did with the Edge tram tool.
+War Machine LLC This is a more direct method of measuring the parallelness of the spindle to the z axis. I like it a lot.
One comment. You don't have to mount the mag base to the column. It can be place on the table as well. That way. you can use the table for precision alignment of the DTI. All that is required is that the DTI be parallel to the x axis and to the y axis for the x and y measurements respectively.
I used a similar method for measuring squareness of the z axis to the xy plane by placing the rod in a chuck on a horizontal RT and rotating the table to find min and max.
+War Machine LLC Thanks for this. I took the time to tram head to column and then column to table. The head must be in tram with column and needs checked first. Thanks for the heads up and all is well now with the deep drilling.Dennis
+BeveledGlass You bet! Glad that fixed your issues!
+NYC CNC Besides mill just working better after this type tram, the ATC now works flawless.
Absolutely AWESOME!!! I still think Servo Motors would minimize a lot of the adjustments... Under High Torque loads, Steppers can loose steps, Servos will not... But that's my observation... Servos are a lot more expensive, so there's always a trade-off... Sorrta 6 of one - a half dozen of the other... Fantastic video... It is really important for calibrating a machine. I'm so happy to be a subscriber...
I wish I could support you financially, but still can't at this point in time... I have to make due with what you post and for the most part it's all been an awesome learning experience :-) When I am better situated, I will definitely support your work :-) Thanks again :-)
+DStray Cat, Well if I were making a DIY mill for metal like the Tormach. Servos like the ClearPath by Teknic would be a good option -- David Burrell at UA-cam channel Neo7CNC really likes his for his router. But the benefit of something like the Tormach is that you have a system designed to work together with support. Also steppers are almost always oversized precisely in order to not lose steps or stall.
+DStray Cat The upgraded motor package for Tormach machines, which I think John has fitted, uses Leadshine hybrid steppers which have an encoder and a closed loop to their matching controller. They will flag a 'following error' like a servo drive if they skip to many steps and can't correct.
Hey mate. Quick question. For milling steel if you were building a milling machine what spindle would you go for. ER32 collet I prefer over R8. Would you go for a spindle with built in motor or a spindle with pulley and external motor?
I was able to shim in the X tram, no problem. What is the best way to adjust the Y tram? By the way, I zero the Edge gage in the Y axis so that I don't need a mirror. It does not matter where you put the magnet as long as you don't move it during the zero process.
Seems to me that's it would be a bit better practice to mount your tram tool directly in the spindle with an R8 collet.Holding a tram tool in a setscrew holder isn't really that bad, but for optimal results its not best.The inherent nature of a setscrew holder to rock the tool to the side abit when the setscrew is tightened down is not what you want when your inspecting your machine like this. Your thoughts?
Great video. I'm just getting started on a Tormach, and I cannot figure out how to lock the spiundle. My PCNC-1100 doesn't have the operator's panel either.
John I'm a little puzzled as to why both you and Tormach are interchanging leadscrew backlash and gib adjustment. I'm not at all familiar with the Tormach, but I believe the two should be completely independent (well to a point the table is jus flopping around).
Were the gibs far out of adjustment? About 6 months ago I wrote in a comment that I felt they were loose and should be adjusted. I'm now curious to know if that was correct.
To adjust the gibs on my machines I normally try to completely disconnect the leadscrews so I can move the tables etc by hand. I then make the gib adjustment by moving the table through the full range of travel and feeling the tightness at the gibs by the drag. Generally you want to be careful out at the extremes of travel as the machine doesn't normally get used as much out there, and the gibs can finish up tight if you adjust only for 0,0 and don't check the extremes. I get much better feel for the adjustment by moving things by hand. I have no idea if that's possible on your machine.
Finally, does the Tormach have a backlash compensation function within the firmware? Once you've exhausted the mechanical backlash, and if the remaining backlash is consistent and predictable, I think you should be able to take out most of the remaining in software. I have absolutely no idea however how, or even if, it's possible on your machine, but the Tormach people should be able to advise. It certainly is however available on other CNC platforms.
Great to hear it's all working out.
+NYC CNC Sometimes it's not necessary to actually remove the screw, and you can just remove the hardware at the ends that supports and controls its backlash. It does feel like you're going backwards to move forward, however in my experience that sweet spot is very narrow, especially as the machine wears more and you will battle getting things "just right" (or at least as good as they can get). Done in the correct sequence (ie first) adjusting the gibs this way shouldn't add that much time to the project. Just be grateful you have a tapered gibb and aren't trying the same with a series of gibb screws!!!
Anyway, it's done now and all is right again. Good to hear that the machine is running sweetly once more. I do think 5 years is pushing it a little for a machine like that in heavy use, and a once year check and adjustment should keep a smile on your dial.
+Pete F To adjust the bearing preload on my 770, loosened the gibs to reduce any friction there,pulled the motor, and replaced it with a torque gage capable of measuring inch-ounces. I loosened the bearing preload and measured the torque required to turn the ball screw at 21 in.-oz. which would be due to way friction and ball screw preload. I then tighten the bearing preload nut to a torque of 38 in.-oz. This gave me a backlash/lost motion of .0004". This is also a good way to adjust the gibs substituting a torque reading for the "feeling of tightness".
+Robert Sakowski Nice, that sounds like a good way to go about it.
+Pete F The torque gage is rather simple in design. It is a flat pulley 2" in diameter with a length of fine cord wrapped around. The loose end of the cord is attached to a spring scale reading 70 oz. max. A steady pull on the scale will unwind the cord providing torque to rotate the shaft. Because of the 1" radius, the scale reads torque in in. oz's. directly.
just curious....why 70oz max?
Hey NYC CNC I have a question, because land cost is so high in NYC, there traditionally isn't much manufacturing or prototyping jobs out there for machinists. What kind of jobs do you do usually do in NYC as a machinist?
Darn, now I feel guilty for not tranning my mill. Thamks for the great vid..
can you tell me the size of the ballscrews in the X Y & Z? It would be interesting to know what size they use in relation to the amount of weight the lift / Move on the axes, this info isn't on the tormach website.
when are you going to work on the 440?
#roots! love this video!
Need links for times
woo! first!! great video as always!!!
I think it has to do with the soul.
my cnc machine cannot start let alone do what u have done. I am not knowledgeable about it can u be of help
You are a very smart man. If I am not already I am going to subscribe to your channel. I have a robo drill / technician related question. I am merely a machine operator so I don't really know what I'm talking about, but I would like your help! If I make a video where I explain to you everything that I do know do you think that you could make sense of the situation and maybe say something to me that I could say to one of my technicians at work that would make everything click and then they would implement the proper programming to fix this error that has occurred???