Recently was given an old surface grinder from my work place and thought about doing a rebuild on it but the more i looked at the machine the more attractive a conversion into a horizonal mill seemed to me. Grinders are nice but it’s not really something i would have much use for. besides this machine was free so it’s a blank slate to build from. the ways were shot so linear ways and ball screws seemed like the way to go and the machine weighs around 1500 lbs so it has enough mass to dampen vibration.
long story short i started thinking about what i would use to drive the spindle and landed on the rigid tapping video on Youtube so here i am
just started working on this and am starting to gather little by little as money permits. should have a 3 hp motor sometime today
I figured i would add some pictures of this old Doall grinder to help visualize my end game on this project. this particular grinder has a dual column linear guide for the head and has two 3" diameter columns that the head would normally clamp to. the original spindle was a direct drive 1 HP 3 phase motor but is getting an up grade to an inverter duty 3 phase 3 HP motor with a base speed of 3,600 rpm. full torque in the range of 360 rpm to 3,600 which i think should be fine up to 5,400 rpm. the torque is around 67% with an overspeed at 90 Hz with from what i could tell is pretty normal with better motors. the biggest thing is how well the motor is balanced for a top end speed and what speed the bearings are rated for. in my case i will alter the casting and make a new spindle for a BT30 taper tool holder with a heavy die spring rather than spring washers. from what i looked into around 900 lbs to 1,200lbs of force is what you should have on those tapers. the air cylinder i have does about 1060 lbs of force so i will probably only be able to get the 900 lbs of force.
grinders in there stock form don’t have a normal dovetail way system like mills and the tables are only held on by gravity so that part is getting changed with 20mm rails since i have those on hand already. the original screws used on the machine were 1" lead screws on the head and cross travel the other was driven by a rack and hydraulics but the pump was claimed by someone else at work and the QC department took the magnetic chuck so this would be another reason for the conversion as well. the plan is to add the linear rails to the long and cross travels and keep the double columns which use a split brass plain bearing system. they have some advantages over the more efficient ball bearing linear guides. i also have a new set of DMM servo’s with DYN4 drives(750W) so for precision i very well could go with a duel loop feed back system with some glass scales and run the drives in analog in velocity mode and use Linux to handle the rest since linux is one of the options for such a set up,
I was thinking about going with an Hitachi drive but after looking at Automation direct drives it would seem the GS21-23P0 drive is comparable and it’s $125 cheaper so that’s a plus. down side is they are out of stock but should be back in stock this coming week. the motor i ordered from them hit a shipping snag due to weather but i’m not ready for it yet so it’s not a big deal.
i highly recommend automation direct and have bought many things from them in the past but they are not always the cheapest on everything if you budget minded like myself. after looking through the manual for this dive and figuring out the braking resistor size and checking their price it was apparent they could be found much cheaper. i ended up ordering from amazon and if the listing was correct it was for 2 of them for about 25% of the price automation direct wanted for one, i did happen to notice not using their parts can void your warranty so keep this in mind.
picked up some material so i can start making mounts for the linear rails today so that’s the first order of business. once the rails are in place i can start getting a better idea of what’s need to to size the lengths of ball screws needed and see what i need to alter to fit a motor and screw assembly in the spaces i have to work with.
i will say to those that go the route of buying a kit for a conversion for these popular bench mills. It does save a lot of time if you have this option but in my case i have to design my own. it boils down to what is available and how much money and free time you have to get done what you want to accomplish. so those that would tease or fault you for going that route i personally would not pay much attention too those people you have to work within the limits of the tools you have to work with and you are the only one that should be making those decisions since you are the only one with all the facts.
i did spend some time cleaning and re finishing the saddle and table top so i will add some more pictures of those sections of the machine while i’m here the table top area is about 9" x 31" i’m about 99.9 % sure i will get a section of flat bar and make some kind of sub plate to mount over the casting for work holding and make a removeable angle plate that can be bolted down to this. most horizontal cnc mills have a tombstone mounted on a rotary axis for a 4 axis machine. i think if i would ever go that far i would look into a rotary trunnion configuration as a possibility way down the road if i would ever get that far. 5 axis machines are pretty cool to see in action
had some time to start working on attaching the table to the saddle this past weekend. since this machine is being modified i had to pick up a reference from the original work done to the casting. grinders are a lot like lathe beds since the have a flat on flat way working with a V way to keep the axis square to the other ones, so the clearance grove for the V in the saddle was a good place to get a starting point. once i had that established i machined a datum on the outer side of the casting to serve as a master rail to set the other one with an indictaor to make sure they are straight with one another.
the bearing truck brackets i made have steps to keep the bearing trucks straight and the master has a step on the top and bottom to locate the master rail. once you have these set you can insert the rails and get a measurement to determine the centerline distance between the rails.
since the original ways were a flat on flat and a V way combo the planes are offset so a offset height had to be determined as well so things are flat when attached. after i set the table on top of the rails i was able to check with a indicator in 2 directions and things were better than .001 for flatness with things not bolted down so that was a good sign i had things pretty close,
i had to contact automation direct so they could open a claim with Fedex since they seemed to have lost the first motor that was shipped? but if all goes well i should have my motor and a drive later today. i have to do some further reading on a good choice for an encoder still. the drive can get feedback but has limits on speed to process the pulses so if you get a PPR count to high the drive will not be able to keep up. so that aspect needs some further research. i did pick up a braking resistor off of amazon they were a lot cheaper than automation direct. so that was around $25 vs something like $85 from automation direct. encoders are able to be found cheaper as well. one thing to think about with encoders is the max speed they can run at so that is something to think about when shopping around for those. one other thing to think about is a max belt speed for parts that might be used in the system. from what i have seen a belt drive is ok to the 5,500 rpm range for the most part but beyond that could create issues. not sure if anyone has brought that point up here or not? something to consider none the less.
did some painting today so now things are starting to match. used an alkyd enamel which seems to be pretty durable. the small section is one of the table extensions. those extensions act like a solid way cover so once they are attached everything will stay covered up as the table moves around.
had to sort out the ball screw and motor mount clearances and have a plan for what needs done. i did mount the floating support end but left the rest for later until i get a shaft coupling in my hands. the fixed bearing block needs a riser block that will double as a attaching point as well. the motor mount clears everything except a web in the center of the table. when i get to the rails mounting holes i can trim it down so it’s not an issue. the motor mount needed some clearance for the motor to fit where i wanted and i circled that section on the one side but both sides needed it. the ball screw nut clears the casting so no issues there. i also took a picture of the table extension in place sort of just to show how the motor and ball screw will be covered and not need anything like telescoping way covers. simple solution to a big problem so i will thank myself later for not being greedy and trying to get some extra travel and the mounts are made for the box ways just not attached yet.
This is an awesome conversion! And these posts are a wealth of knowledge. Thank you for taking the time to document everything (w/ pictures!)
You can run the encoder off of any gear train. This will let you achieve much higher spindle speeds while still using a cheaper encoded rated for 5,500 rpm. Then if you are using Linux CNC you can set the pulses per rotations:
setp encoder.3.position-scale 100
500 counts per rotation or greater is ideal for rigid tapping.
With this setup you would have to find a different way of getting your index pulse (like a 1 notch index disc that does around the spindle) because index pulse and spindle rotation need to be 1:1.
I’m excited for your next post! That blue alkyd enamel looks great.
by the looks of things everything will be analog. for duel loop feed back with servo’s in velocity and linear scales working together will be going into a 7i97 Mesa board… i will have to add to this for spindle control and MPG since the 7i97 only has 6 spots for encoders. that part is a ways out yet but i do have the 7i97 card already.
right now i be using 3 750W DMM servo’s with the DYN4 drives running on 240V i was told by DMM the 750W servo’s will lift somewhere around 900 lbs coupled to a 5mm lead screw. they said people use these motors to lift the knee on full size Bridgeport’s.
the motor i am using is a 2 pole motor rated @ slightly under 3600 RPM at full torque it has around 24 Nm and at 90 Hz it still has around 16 Nm so a 1 : 1 ratio should work just fine in the 5000 to 6000 RPM range
this seems like a good trade for some loss on the low end torque. this motor is good down to around 360 rpm so i don’t see to much need lower than that. but if i were to heat sink the motor and add some fans that would correct most low rpm short comings sine this motor is a TEFC model.
should have my coupling for the servo’s this week and hope to have the X azis buttoned up so i can figure out the Z azis or cross axis.
the mechanical stuff is not big deal since i am a machinist by trade for long time the electrical stuff is my weak point i know enough to get myself into trouble
some more progress but it’s slow going setting big items and set ups are a bit crazy so you need a heavy duty pucker switch to pull off some of the hail Mary set ups just to get stuff done
one big problem is i don’t have enough travel so i have to establish a datum to work from swing the head on the mill every which way and pick up from a known point. now i have the DRO on my ball screw center so i can edge find the extra set of rails i have in and get a good number so when i put the table on next i can get my drive bracket mounted right where it needs to be.
this was just a rough assembly to make sure things will go together. i was trying to see where the spindle center line would fall once everything is in place it looks like i should be able to get at least an inch below the table top taking a guess. i had the steel plate in the picture it’s 12" 30" x 1.25 thick, will save me some work since it’s flat and drilled and tapped as a sub plate with dowel holes. having some extra width on the table will help out for attaching angle plates.
finally starting to get the X axis and Z axis sorted out. i had to resort to how the original machine was set up to drive the cross feed, the original only had around 6" of travel the best i could have done would be around 14" but ended up at 11" of travel for a Z axis. should be just fine the extra travel is mostly to get out of the way for setting up and the cutting area will be in a more rigid area of the linear bearings. at first i wanted to have the end of the ball screw supported but after looking at the critical speed of a screw this short it’s not needed. if i could support it i would do it anyways but since it’s not doable i think i can live with it
I had some time this past weekend to get into the spindle of the grinder and compare what i had for a prototype BT30 spindle i did before and get an idea if the casting for the spindle could house this spindle and it looks like a go on that so i ordered a piece of 1144SP. the first one i did was 4140HT and it worked out but was looking to see if this material was easier to machine? both are pretty hard as is so this was why i went with it begin with.
the grinder spindle only used 3 bearings but the one i will build will use 4 in total two angular contact on the tool holder side and two deep groove on the belt drive side. the first spindle i did had some pretty large bearings so i will size the OD of the bearings down just a little since the casting is a little smaller than the housing i made for the first one. it should still be a pintsized brute since the first one had some serious over kill on bearing size i had decent luck with the runout on the first one and only had about .0002 in runout. i am fine with that and is probably about the best i could get working with what i have.
i’m going with the 3 finger style clamping system that works with pull studs on the tool holder. there are a few types but this is what i went with. i figured i would share the pictures of the first one and show a sectional view of the spindle shaft i will be making for this machine. no need for a housing on this one since the casting of the grinder will be modified to replace that part of the spindle.
i started working on the roughing of the new spindle shaft like two weeks ago but was waiting for bearings to be in hand before i started to finish anything on the shaft, the upper bearings came this past Monday so i can start finishing the out side this weekend. i went with an alloy steel for my shaft 1144SP to be more exact. the 2 main spindle bearings on the lower are precision angular contact bearings and the two upper bearings are deep groove radial bearings. the angular contact bearings are a matched set and are a bit expensive. on a high end spindle i think you would find angular contact bearings all around but what i have is far better than what you will find in the hobby class mill i’m guessing. the angular contact bearings are ABEC 7 or a P4 class bearing and size variation is documented to the microns. ABEC 7 or better is what you need in a spindle.
spindle shafts are the most critical part of a spindle. the housings are a little more forgiving on sizes but still tight. only a few tenths on most of it. when i go to finish the outside it will be between two centers. i’m not getting into grinding anything if i can help it. i’m not really set up for grinding so for finish i have to polish but that’s fine i can fine tune the shaft to make it better.
i was playing around with friction driving the shaft between centers and worked pretty good. finishing between centers with no drive dog getting in the way will allow me to do the entire shaft outside in one operation and gave me the chance to tune in the tailstock to remove tapper in the turns. i took a test cut this way and the runout is next to nothing .00005 to be a little more exact. so having a perfect outside will make locating the shaft for the internal work much easier. the OD and ID need to be concentric to within .0002 or better so .0001 or .0002 on pretty much everything is the goal
bearing fits are done on the shaft and got the tapper finished plus i got the threads cut for the bearing nuts, not much run out for a machined spindle from what i can see so far, was able to put a holder in the shaft and get an indicator on the holder and as far as i can tell at this point is it’s in with in 1 or 2 tenths, but it’s hard to say with only friction holding it in place. i still have a ton of work that still needs done but it’s moving forward a little bit at a time.
i managed to get the under cuts and internal angles that work the draw bar clamping and unclamping finished but the middle of the shaft internal work is hard to get to with a boring bar on my little lathe so i ordered a drill to finish the rest. this make my life easier
once the back side of the spindle tapper was opened up i went back in to fine tune and check my fit. the way it was the backside of the tool holders could hang up since the ID was not finished at that point. i feel like i have got it as close as i can get it. the fit was so close i actually got the tool holder stuck in the shaft and had to pop it loose gently with a screw driver. the best way to check fit is with some high spot blue or something like it. it’s cheap and effective. the thinner you can make the high spot blue the better.
i ended up taking about 3" off the length of the head casting since for milling it did not need to be as long as it was. i will wait until i get the shaft finished so i can mount the pulley to see how short the head casting needs to be. i’m not in a hurry so working up to a finished size lets me see how things fall together.
just some more pictures since people like pictures
just some more pictures of the spindle shaft. it’s almost done, the only thing left to do is add some drive dogs for the tool holders. started working on a labyrinth seal but nothing picture worthy just yet,
Ok you took this project to the next level with the building of a BT30 spindle from scratch. Would you be interested in a making a second one that fits my PM-833TV? During heavier cuts, I am struggling with pullout. The R8 spindle + power drawbar does not supply enough clamping force. If you are interested and have extra time (no pressure at all) let me know a ball park cost.