Bridgeport CNC Conversion

[@awesomesam27yobrotha]

I just graduated with my engineering degree and am about to start a new job. I can't wait until I have enough money to do stuff like this.

[@JoshuaMcQueen]

Hey dude -- your video is amazing. I've just come into a Bridgeport Series 1 and thinking about doing a CNC conversion. Just wanted to say thanks for putting content like this out into the world. No doubt learning from your experience will save me weeks of work / trial & error

[@RECombatRobotics]

Thanks! I'm glad I could be of help to your project.

[@VanillahGarillah]

I was shook when I saw your setup in the intro and was thinking ‘damn I wish this guy would do a walk-thru of his set-up’. Glad I watched the whole thing! My machine isn’t quite tight enough for CNC, but you’ve given me lots of helpful ideas to improve my mill.

[@hamzanawaz7945]

This is like the boss of all conversions. That big facemill at the start was amazing.

[@makeorbreak5887]

Love seeing z axis knee conversions. Makes me want to get my other machine projects out of the way so I can get on with converting my beaver vbrp.

[@modelsport2858]

You are the type of young man every father wishes for. My congratulations to him, especially. Well done on this video, the work of converting the mill, and good luck with your career in robotic or engineering, or whatever you choose. Good job. Subscribed and thumbs up!

[@metalillness5951]

Well done, thank you for sharing! You give us all hope for the future of American manufacturing, very inspiring.

[@jdrimm]

Super clean conversion, I really like your knee drive solution! (20 years of toolmaking, and several CNC conversions/retrofits)

[@corrod3195]

This is absolutely amazing. This is what I dream and hope to do in the future

[@Freetheworldnow]

This is a fantastic conversion. Well done!

[@ogaugeclockwork4407]

Great solution to the weight of the knee using the gas struts!

[@krazykozey2259]

I have a 70s bridgeport I just bought for valve jobs and am totally copying you, lol. Love it!!!

[@frijoli9579]

Love the Acorn! Good job on this build. I am not thinking hart about that knee control on mine.

[@cnckeith]

Excellent, how did I miss this video!

[@martyscncgarage5275]

Outstanding work on the conversion!

[@johnjohn-pt7pe]

One of the Best conversations out there. For years heard using the knee wouldn't work well. Using the Knee as controlled Z axis makes a lot of sense. What would you do different if did over? Which collet system are you using? Thx for sharing your videos.

[@RECombatRobotics]

Thanks! I'm just using standard R8 collets and primarily rigid tool holders. Definitely not the fastest, but it's also cheap. If I was to do it over again I'd definitely install reliable limit switches on the machine. Right now the servos homing to a hard stop acts as the limit switch, but I don't trust it enough for table related work offsets.

[@LordOfTamarac]

Really cool build!

[@danielminskey]

great build. well thought out.

[@jamesellery6866]

I have probably watched this video about 20 times. As someone trying to do the same thing, there is so much information packed into this video. You have done an amazing job, thankyou for sharing! I could spend an hour talking to you, but will consolidate to the following questions if you have time please: -You sounded a little disappointed with the z-axis ball screw. Are you implying it might be just as good to stick with the acme screw for the Z-axis? -How did you connect the X and Y servo housings to the table. Did you drill your own holes into the aluminium brackets to run your own bolts or did you use the existing bolts for the "bearing plate"? It seems like a huge amount of weight hanging off the end. I'm not saying there is another option, just want to know how you did it so it doesn't fall off for me! -Did you just keep the standard bearings for the all the axis? -My machine has 32mm acme screw, but in order to fit the ball screw nut for x and y axis I think I need to go for 25mm ball screw. Do you think this would be sufficient (appreciate this is just guess work) Once again, thankyou very much! I will e-mail you as well for any drawings you might have. Cheers, James

[@RECombatRobotics]

Hi James! I sent you an email with additional information, but figured I'd post some of it here for others to take a look at as well: •I think the acme screw to ball screw swap on the knee is absolutely required for a conversion like what I did. If you left the acme screw, you’d need a much larger motor or a gearbox to increase torque, which would result in more backlash and decreased speed. •My main disappointment with the knee screw was that it came without seals and non-preloaded, meaning it has a tiny bit of backlash and requires a lot of grease to stay running free. The screw I purchased is part number BRP-47-8258 from Rockford ball screw. They should be able to provide you a drawing for this if you email them. When I purchased the ball screw from Kaman Industrial (one of their suppliers, they don’t sell direct) it was $508.34, but they have since said this was a quoting error and it was supposed to be $1,677.36. For this price, you may be able to find a different manufacturer to make one for less. •The X and Y servo housings were connected to the cast iron bearings blocks with three additional tapped holes. I drilled clearance holes in the aluminum bearing retainer so I could remove the housing without loosening the bearing retainer. •I purchased new angular contact bearings for the X and Y, and a new ball bearing for the Z, but this was only because mine were worn out. If your bearings are in good condition there should be no need to replace them for the ballscrew conversion. •While I’m not sure what exact machine you have, 32mm balls screws are the same size as what something a lot larger that my machine, like a HAAS VF2 uses. I think most knee mills would run into horsepower and rigidity issues before the ball screw size would become an issue. You could also do with a size larger coupling to fit the larger 32mm screw.

[@jamesellery6866]

​@@RECombatRobotics Thankyou very much Owen! The drawings you sent are amazing! It's like you are going into mass production. I'm going through and ordering my items. Do you have a braked servo for the Z-axis? I will send you some footage when mine is (hopefully) finished. Thanks again. James

[@RECombatRobotics]

Glad to hear those worked for you! The servo on the Z axis is the same as the X and Y, no brake. On my system the gas struts balance the knee to the point where one isn't required. Depending on how close you get the balance on yours, you may need a braked or more powerful servo.

[@fnsfabrications6844]

Nice work man

[@agg42]

Nice work on the Z axis. I don't think any production 3 axis knee mills redesigned the z axis as far as you have.

[@RECombatRobotics]

Thanks! I've been very happy with it so far. Now that I know how well the gas struts work, the harest part about this style conversion is probably cutting the keyway in the pulleys.

[@sniperasys]

When it comes to the conversion, I think it's a decent job. The concept of a motor housing that attaches the motors to the original castings through a sparsely welded enclosure made from skinny sheet metal is probably contributing to the subpar finish. I'd also assume that it's quite difficult to access some of the fasteners inside. I'd replace the housing with a short (as short as it can be) milled spacer that bolts onto the casting on one side and makes contact with motor flange itself on the other side. You can take it a step further and make it a motor mount that connects the table casting directly to the motor flange so it functions as a bearing housing for the ball screw support bearings too. Access to the coupling can be provided conveniently from the underside. It's a good idea to cover the motor in order to protect it from damage, but I'd definitely replace the welded structure with a shorter and more rigid part with minimal overhang. It's probably enough to bend or weld an inverted U-shaped cover and bolt it to the motor mount. Some CAD suites offer FEA, so you can model, analyze and compare the alternatives with the current structure. Not to mention that now you have a CNC mill. :) I'm not sure what combination of lack of rigidity and machining parameters led to the poor finish, but there is at least some chatter. Are the ways alright? If you're making a lot of these parts, consider adding another setup for facing the slant side. It's not economical to waste that much time creating such a simple feature.

[@CJ-ty8sv]

With those setting for you end mill, you are way slow, even for a bridgeport as you are running it in a conventional type of strategy with the setting you are running. Dial back your radial step over to about 15% diameter (or about .055") and run the cutter as deep as the pocket will allow or if pocket is deeper than 65% or so of flute length than probably limit your depth to about 65% of the flute length for a less rigid 2 flute endmill (with 3 flutes in aluminum, I run them all day long at 90 to 95% flute depth if the pocket allows). Being a vari-drive head (and even more-so with having the VFD) you should have no problems getting roughly 4200 rpm out of it. Now increase your feed to account for chip thinning because of the small radial stepover and you will be really using one of the advantages of HSM. For example, your settings (.100 step over, .200 step down, 3500rpm and 25ipm) are netting you about .5cu/in MMR. for say a 1/2" deep pocket. Even with keeping your 3500 rpm speed but dropping down to the .055" stepover and running full 1/2" depth, you would want to run roughly 35ipm (or about 40% faster than you are) and are at about a 1cu/in MMR and all the while not increasing the load on the cutter any even with the full 1/2" depth. Max that spindle speed out and adjust the feed to match and you will pick up even more speed. Its key to not that tool manufacture chip load specs for profiling are based on 50% cutter engagement any time you are less than 30% engagement, you will want to increase the feed to adjust for radial chip thinning so you dont rub. Above 30%, the risk of rubbing is very low BUT you are also not running the chip load at what you think you are because of radial chip thinning. Also since this was about aluminum, if surface finish is important to you, then you are better of roughing with carbide and then using a high quality HSS tool for a finishing pass when working with aluminum because HSS is always sharper than carbide and will give a nicer finish because of the sharpness difference.

[@RECombatRobotics]

Thanks for the advice! I'll definitely try those changes on the next aluminum part I work on. With the spindle speed, I need to rebuild the vari-disks first before I start to overdrive the motor with the VFD. Right now it sounds way too loud to push it much beyond 3600 rpm.

[@CJ-ty8sv]

@@RECombatRobotics are you sure what you're hearing isn't clutch chatter? I hear a lot of people complain of varispeed heads having a problem with the discs and such when really its just wear on the drive dogs and just needs to be adjusted. Usually if there is really a problem with the varispeed setup, you will get bad noise with adjusted to max speed in low range / backgear. If its nice and smooth / quiet maxed out while in in low range then its likely that the noise you are getting when up high in high range is clutch chatter and its adjustable through the selector lever. Most clone machines you just have adjust the position of the lever on the shaft. Original bridgeport design has a plate that can be flipped to offset the detent hole that the lever engages in when in high range position.

[@CJ-ty8sv]

@@RECombatRobotics also about the speeds, my original post was based on the assumption that you are using actual HSM type tool paths. If not, those won't help you much (and can hurt you) when you get into corners which is the main reason why HSM type patching was developed. With more conventional type patching, straight line linear moves aren't a problem but when you hit corners, you can suddenly load the cutter to far more engagement than you thought that you where (namely when machining pockets). The other thing is you paid for say 1 inch length of carbide, you might as well use it. By that I mean, why wear only .200" of that 1.00" long flute. If you are going to be matching a lot of shallow depth features, you're better off with a cutter that has flute length just long enough to cover the depth (like stub length mills) and then use the added rigidity to push it harder (I.e., taking max or close to max chip load). On the flip side, not being a "production environment / setting" speed likely isn't as much of a concern for you so if you're happy with the results, then probably dont need to worry as much about speed.

[@RECombatRobotics]

@@CJ-ty8sv I'll have to look into that, I'd just assumed it was the disks. Haven't gotten to take the head apart yet to see the wear on everything.

[@CJ-ty8sv]

@@RECombatRobotics Before you take (possibly waste) the time taking the head apart, you can quickly and easily tell if the noise is clutch rattle / chatter. With the selector in HIGH gear, grab the nose of the spindle (or with a drill chuck or something installed and tightened dawn) and twist back and forth with for see if you get a knocking noise when why change directions. Where should be no play or noticeable loud knocking. You might get a faint dull light knocking noise which is usually the drive splines of the spindle itself which has to be like that because of the running fit between the spindle splines and the drive hub or the quill wouldn't go up and down real easy. What you are looking for is slop / looseness in the drive dog clutch of the HIGH speed drive. The two haves mesh with trapezoidal shaped teeth (very very slight taper on each side of each tooth). Over time, the teeth wear and there becomes slop between them. Bridgeport know this was likely to happen and thus made it adjustable both because of wear over time and also to allow adjustment during assembly for machining tolerance error during manufacturing. Pretty much all the adjustment is doing is changing the location of the locking detent of the selector lever to allow the front bull gear assembly to push up more (tighter mesh of the teeth) into the drive hub that is part of the bottom vari-disc assembly.

[@smerfu8734]

Impressive machine you got there

[@timd1191]

cool

[@HuskyMachining]

well this is super cool! p.s. this is a super cool project!!!

[@kenjohnson6338]

Gotta ask... ,how many hours did do put in before it worked as you wanted and approximately how much did this kit cost , without machine... I have heller ftv-3, and am thinking of this conversion...

[@RECombatRobotics]

I worked on the project on and off for a couple of months, but knowing what I do now it could be done in only a week or so. If your machine already has ballscrews on it this greatly simplifies the process. The cost sheet in the video description should answer your questions on cost. Let me know if you're wondering anything else!

[@kenjohnson6338]

@@RECombatRobotics I'm looking now , to do a ball screw conversion .. here in Central Europe are a bit pricey... but not out of range... Great video dude... Gives hope and motivation... thanks for sharing ...

[@dota2crew226]

What CAM software did you use to get the Gcode ? I tried to run FreeCad on my knee style cnc mill but the CAM software moves the spindle instead of the bed :)

[@RECombatRobotics]

I'm using Fusion 360 with the Centroid or HAAS pre NGC post processor for controlling my mill.

[@mattbentley8958]

I have a very similar milling machine with the z axis running the quill. I am going to do the exact same knee ballscrew and servo drive conversion as you have done to get more Z travel. I am having trouble finding the ballscrew and other parts. Do you have any more documentation on this part of your conversion? Any drawings or schematics along with part numbers and suppliers would be very helpful. Thanks

[@RECombatRobotics]

There is a spreadsheet linked in the video description that includes most of the parts and part numbers for the conversion. The z axis ballscrew is especially difficult to find. It is part number BRP47-8258 from Rockford Ball Screw, which you'll need to purchase from one of there suppliers (I used Kaman Industrial). I belive it has increased in price since I purchased mine. I have some manufacturing drawings for the project I would be happy to share. Easiest way would be to send me an email at [e-mail] and I could send those to you.

[@oliadams444]

@@RECombatRobotics Hi, very impressed and also looking at doing the same. Might I also be able to email you for some of the information on screws etc..? Thanks

[@RECombatRobotics]

@@oliadams444 , of course! I'll look into posting a download for drawings and additional part numbers on a future video on the mill--but for now you're more than welcome to email me. [e-mail]

[@ShopperPlug]

5:26 - I started to see many CNCs builds are using timing belts to move the axis, doesn't timing belts diminishes accuracy when compared to driving the axis directly with the servo? What kind of tolerances are you obtaining?

[@RECombatRobotics]

I'm able to regularly hold 2 tenths repeatability on this machine with the direct drive axis setup. The belt on the Z axis isn't too bad either, because the weight is always loading the system in one direction. Aside from the compact configuration, I'm not sure why so many kits use timing belts for X and Y controll.

[@ShopperPlug]

@@RECombatRobotics Thanks for the reply. Good to know you get 2 thou with your CNC conversion, thats really accurate, close to factory level using the mill manually. So I guess the timing belt should also be something like 2-4 thou.

[@joshuaschwark]

Hey, i have a cnc based on a series 2 that im currently working on. Id love to see some more details on your z axis setup. Any chance youd be into sharing your ideas with me?

[@RECombatRobotics]

I'd be happy to! Shoot me an email at [e-mail] and I'll try to answer your questions and provide some more info on this.

[@DuxburyPrecisionConcepts]

Very clean conversion! I think I need to do the same knee conversion to my Bridgeport. I CNC'd the quill and have always wished for more Z travel. What material is your backsplash made from? Really like how that looks with the bellows as well.

[@RECombatRobotics]

Thanks! The white backsplash is PVC cardboard from Home Depot. It definitely helps with keeping the mess down and to brighten things up.

[@DuxburyPrecisionConcepts]

@@RECombatRobotics Definitely going to have to do that for mine! I built a aluminum extrusion and polycarbonate table enclosure for mine, but could only build it so high in the back because of the ram extending into the work envelope. The backsplash would help out with that a lot!

[@chananielwizman2401]

Why so worried about backlash on Knee? Gravity takes care of that? Just need powerful enough servo and no gas shocks or whatever they are called. Am I wrong?

[@RECombatRobotics]

The servo required would have to be huge, or use a very expensive zero backlash gearbox to increase the torque. The price of the ball screw was less than either of these options when I compared it.

[@chananielwizman2401]

@@RECombatRobotics Have seen 1.5 KW servo set ups going for like $1000. You paid less than that for shocks and ballscrew etc etc?

[@RECombatRobotics]

When I purchased the ballscrew kit for the knee is was only ~$500, plus $50 or so for the gas shocks. I'd definitely be interested in larger servos for future projects, but for this the small motors works pretty well.

[@chananielwizman2401]

So say u got lucky on ali baba for big servo motor. U would then have thought gravity take care of backlash?

[@RECombatRobotics]

Yes, with a big enough servo there would be no backlash issues with keeping the stock ACME screw and nut.

[@eackerw85]

would you drop a link to the ballscrew manufacturer

[@RECombatRobotics]

X and Y ball screw kit is from H & W machine. machinerypartsdepot.com/product/rbs-bp/ Z axis balls screw is from Rockford Ball Screw, but isn't a publicly listed product--It's part number BRP47-8258. Give that part number to one of their suppliers (I used Kaman Industrial) and they can quote and order one. rockfordballscrew.com/ballscrews/bridgeport-kits/

[@hoonyur5973]

Dude.. this is so badass! Do you think this could this be done to theBridgeport series 2 interact 4 for around the same cost..?

[@RECombatRobotics]

With that being a bit bigger of a machine you may want to go one size up on the servo motors. The rest of the Centroid controll would stay the same.

[@marceloarielmarichalar8982]

donde puedo descargar el software intercon y cencomm para bridgeport cnc1?

[@RECombatRobotics]

Software downloads for the Centroid software I'm using can be found in the link below. I use the "Mill Pro" version, which does require a purchased license. Hopefully that answers your question! Link to downloads: www.centroidcnc.com/centroid_diy/centroid_cnc_software_downloads.html

[@jevlars]

Nice build, would You mind listing the servo model and sizes you used?

[@RECombatRobotics]

I'm using three Clearpath SDSK 3234S-ELN servos from Teknic for this project. Link to the product page here: teknic.com/model-info/CPM-SDSK-3432S-RLN/?model_voltage=75VDC

[@lexcrob1]

Amazing video. I just got a 1 1/2 hp series 1. Which vfd did you choose and why. I could not make out your model number but was referred a WJ200-007sf.

[@RECombatRobotics]

I choose a Hitachi WJ200-015SF VFD for this, primarily because it was in stock and had enough power that I felt I wasn't pushing it too hard.

[@privatestaticstringwily4325]

What is the maximum rpm of the spindle

[@RECombatRobotics]

In theory it's 4200 rpm on the serise 1, but mine gets too noisy above 3600 rpm due to the variable disk assembly being worn out. I keep meaning to purchase a rebuild kit from H & W Machine to fix it.

[@ronbown3836]

Got a question about your knee travel hydraulic cylinders. How long are they and the part no.?

[@RECombatRobotics]

Here's the exact part I used on McMaster Carr: Gas Spring, 5/16"-18 Thread Size, 35.43" Extended Length www.mcmaster.com/4138T63

[@ronbown3836]

@@RECombatRobotics okay I ordered that found that on their website. The ball screw is what I'm having a hard time finding

[@ronbown3836]

@@RECombatRobotics I just measured my ways on the base of the Bridgeport and it looks like I got about 11 in of travel.

[@ymenaviator]

What tolerances are you getting on each of X, Y and especially Z. Thanks.

[@RECombatRobotics]

X and Y I can hold 0.0002" or so on bearing bores when I creep up on it, and Z I've never noticed any blending or height misalignment over maybe 1 thou. The weight of the knee means it's always loading the screw in the same direction so there's barley any backlash to contend with.

[@ymenaviator]

@@RECombatRobotics Thanks for that. Did you ever test without the ball screw conversion? If so what was the tolerance then. Thanks.

[@RECombatRobotics]

@ymenaviator I did not, but with the stock ACME screws you could get similar Z results, but significantly worse for X and Y

[@ymenaviator]

@@RECombatRobotics really appreciate responses. Just to clarify, so you are saying that you could just attach a stepper/servo to the existing handle input through to acme screw on the knee and this would probably work just as well providing you have some weight on the screw?

[@RECombatRobotics]

There would be a bit more resistance than the ball screw and gas strut setup on the knee, so you'd need a larger motor or higher reduction, but it should work.

[@johnkennethfernandez61]

What ball screw kit you uses

[@RECombatRobotics]

I used the kit from Rockford Ball Screw for all axies on this machine.

[@satchelsieniewicz5824]

what did this conversion cost??

[@RECombatRobotics]

There is a link to a Google docs spreadsheet in the description that details all costs for this project. The spreadsheet comes close to 11k (including the base mill and tooling), but the conversion itself is more like 7k. Let me know if you can't acess the spreadsheet and I can try to fix it.

[@dsadzaglishvili]

Is there a special adapter from cnc to r8?

[@RECombatRobotics]

The mill still retains the original R8 spindle with the conversion. I use Lyndex Nikken brand solid tool holders to ensure repeatable Z heights from tool to tool.

[@dsadzaglishvili]

thanks

[@dsadzaglishvili]

i got mhp with hydra-path III controls. Do you know where i can get tutorials, owner manual, documentation etc for this machine & control system?

[@RECombatRobotics]

I'm not familiar with that control system, sorry. So far I've just used the Centroid system on this machine for contoll.

[@dsadzaglishvili]

@@RECombatRobotics thanks

[@DedySetiawanMesin-iz6cs]

Keren mesinnya bos.... Berapa Axis mesinnya?

[@RECombatRobotics]

“The engine is cool boss.... How much is the Axis of the engine?” Forgive me if I incorrectly translated your question, but the conversion shown in the video is a full 3 axis. I’m working on a 4th axis rotary for it as well. Maafkan saya jika saya salah menerjemahkan pertanyaan Anda, tetapi konversi yang ditampilkan dalam video adalah 3 sumbu penuh. Saya sedang mengerjakan rotasi sumbu ke-4 untuk itu juga.

[@williamhuang5329]

Hanzhen harmonic drive gear , over 30 years experience , robot arm gear reducer,

[@jagboy69]

wtf happened to my comment? try bidspotter, you could have gotten a 10yr old haas for 5grand.

[@RECombatRobotics]

Not sure, I saw the 1st comment appear on my end then disapear as well. I'll have to take a look at that site, but I didn't go with a VMC maily because of space and for my 1st machine I wanted something that was a bit more adaptable to large/weird parts.

[@Steelcrafted]

I like your z axis design.... I've got a cnc knee mill and the z axis is in the quill movement, kinda stinks, but it does ok...I've been thinking about how to do it with the knee, but without the bevel gears...good job....

[@laprepper]

Bro don’t ever work for someone, just start your own machining And or robotics and automation business. I’m serious.

[@laprepper]

You could make enough in one or two automation projects per month to match the salary of a starting engineer so really at your skill level there’s not much point in getting a job unless they’re paying you a stupid amount of money and you’re working with ridiculously intelligent people