Dang, I need to stop by here more often! Lots of new posts!
I’ve been sidetracked since my last post installing a natural gas heater in our garage/shop. Just like everything else, it takes longer than you would expect… Good thing I don’t do HVAC for a living.
David, after having gone through building the control box, even with the huge jump start your bill of materials gave me, the idea of being able to buy a pre-built panel is intriguing. For me, it’s about the learning experience, but for others, that box might actually represent a barrier that separates them from getting started in open source DIY CNC.
However, bear in mind that as with many manufactured products, including software, there can be a lot of post-sale support required. Also with electronics, DOA problems that are real, perceived or consumer induced could be a problem.
When I was about halfway through my box, I wondered why I didn’t do it on DIN rails and maybe separate the power supply or supplies into a separate enclosure so it was more modular/upgradable.
Hopefully, I’m back at a point where I can get started on the mill conversion. I never imagined it would take me this long to get to it. I guess it’s been that kind of year. Now I’m curious to see how the holes on my mill line up with the parts Heavy Metal supplied. Also might have to spring for the nut upgrade, so I don’t have to do that again…
Yah that is my biggest concern with this possible business idea. Most people who are going to buy a manual mill to convert it to CNC are going to be looking to either A) save cash or B) learn about the process. In either scenario, that person is not going to be likely to pay for a cabinet built correctly, which could cost close to $1k for a stepper system and perhaps even >$2k for a servo system. I find building the cabinets to be therapeutic, so I thought I would toss that idea out there.
Yes this could be a huge time suck from my other hobbies. I could test the control box on my mill, but who knows what kind of problems can arise during shipping and user installation.
Thats awesome man! The threaded ports are a great addition.
It’s awesome that this thread has 141 replies and almost 2,000 views! Its a little bit hard to read though with the separate builds, so I created a new category just for posting CNC mill conversions. Feel free to continue to use this thread to ask questions about my mill, but if you need help or want to share your project (please do!) then head on over to that new category and your own thread.
Left you a message on YouTube as well on your latest video. I just finished my mill which is an 833TV as you recommended. I noticed in the video you didn’t want to change out the ballscrews due to the work, however, i thought you might be interested to know it can be done without taking the mill apart. Also, not sure what you had your driver’s tuned to (pulse/rev), but I can get rapids of 300 IPM with my steppers and they are very smooth at low speeds. If you aren’t feeding them enough voltage, this will also have an effect on max speed due to motor impedance.
Hey man! Glad to hear that the conversion is going well. I do see how the ball screws could be finagled to get everything in there with disassembling the mill, but you would at least have to lift the mill up or remove the table to attached the ball nut for the Y axis wouldn’t you?
Also I was using 1/8 uStepping, so 1600 steps per revolution or 320/mm. I was using some cheap drives too, which could be part of my resonance issues. What drives were you using?
I have a small split in my table so I was able to slide the “Y” axis out the bottom, however, you should be able to just tip it back and slide it out from underneath. I would put a small block under just in case (don’t want to lose any digits). The X axis can be removed by removing the plate and sliding it all the way to one side and proving it up with a table/books from underneath and exposing the bolts while giving you enough room to slide the ballscrew out. Z axis can be accessed by moving it almost all the way to the top leaving enough space on the inside to use a long rachet extension between the spindle head and in between the ways on the one side. The other side is exposed via the rear hole in the top. Lock the head using the locking thumb screws and if you are really safety conscious you could put a block of wood under the spindle.
As far as steppers, I have AM882 leadshines running at 80 volts per motor micro-stepping at 1600. I used cheaper ones and 48 Volt supplies on my cnc router table and noticed a huge difference when I spent a little extra and bumped the voltage in both torque and noise.
I would love to put servos on, however, just can’t justify the cost and it moves plenty quick as you pointed out at 200 IPM. So quick in fact I can attest that even at 50% (100 IPM) it is difficult to catch before noticing you forgot to zero after inserting a long drill bit!
Pro Tip: It will rip through 1/2 inch aluminum, and about 3 inches of hardened steel (vice and table) at 100 IPM without breaking an 8 mm bit! Unfortunate but impressive!
That is close to how I was envisioning the sequence of events for installing the ball screws without taking the column and table off. I appreciate you taking the time to write that process down.
I was supplying 72V to my DM860T drivers. I really only had noise and resonance at that super slow probing speeds everything else was smooth.
The movement of the mill feel so much faster in real life. Especially when the only thing between you and that endmill is a couple feet of air.
My old PM vise that was in shambles in my last video has seen a few incidents with the end mill. I’ve come a long way with generating tool paths, so the Kurt vise should be safe for the time being.
If you have pictures of your build and more tips, then it would be awesome if you could be the first person to start a thread over at: CNC Mill Conversions and Builds - Dr. D-Flo
It would be awesome to have a collection of different cnc conversion for other to draw inspiration from
Will do. I have to remove the motor mounts at some point and replace them with metal milled ones which will give me a chance to take a few better pics to go with a write up.
I milled the connections manually for the ballscrew connections and 3d printed the motor mounts in PETG which worked out great, was quick and easy to get up and running. I was shocked at how strong they are and actually didn’t think it would work at first. That being said I will be using the cnc now to mill them properly.
I use my control box for both the mill and my 50 inch by 50 inch router table so, like any good military man, it’s a portable, water tight mob box with aircraft connections. I wanted the option to use it for multiple projects in the future that use stepper motors to drive linear motion… I’m thinking an automated wood mill and (potentially) a concrete 3d printer!
PS - Thank you for your initial video and your responses. It really helped me in my decision to take the plunge and my final decision on which mill to use as my base.
@Rdwelty I apologize for missing your message! I have no complaints with the Hallmark probe. Whenever I flip a part, I pick up the ID of a hole that I machined in the last OP. Hallmark probe + Probe basic makes this task super easy. No problems with false triggering like I had with my last probe.
@TCandee I picked up an inverter duty Black Max WVA145THTN6046BAK motor to finally do my rigid tap conversion. There are a couple details that I need to sort out and any help would be much appreciated.
The spindle timing belt that came with the 833TV is a weird spec. It is a 20mm HTD belt with 5mm pitch. For the sake of my viewers and readers, I don’t like to purchase supplies from temporary retailers (e.g., aliexpress and ebay), but prefer companies like McMaster-Carr. However, I am only able to find a 15mm wide HTD belt with 5mm pitch. With rigid tapping that belt is going to experience higher forces, so what is a sufficient belt width for this 2HP setup? Also, HTD belts have more backlash than newer belt profiles and I was curious how much that matters for rigid tapping and what the best tooth profile would be?
Is there any reason to not have a 1:1 gear ratio between the motor and the spindle?
You have your encoder in line with the spindle. I understand that such a setup would give you the exact orientation of the spindle at all times, but with my powder drawbar, such a configuration is more difficult than just attaching the encoder to the shaft sticking out of the back of the motor. Would backlash and belt stretch make this an undesirable placement of the encoder?
If you are going to have your encoder on the motor side you have to have a 1:1 ratio unless your encoder input has a way to calculate the ratio on the fly. I don’t know of any that can do this accurately and fast enough. Mine is not capable. I had mine on the motor side for awhile and it worked fine with a 1:1 ratio. I redesigned my PDB and switched it when I had my pm-25mv and it was underpowered for some rigid tapping sizes and I had to be able to switch gears.
I used the 20mm belt width. I ran into the same problem with McMaster Carr and the HTD. With no real reason, chose to stay with the 20mm HTD and had a custom pulley made for my motor. The ratio is not 1:1 only to be able to reach 6000 rpm at the motors max safe speed rating. The HTD belt has a very high HP capability. I think the 20mm is about 10hp. So I think going to 15mm would be fine and then you have more options with McMaster. The HTD belt I have is working fine. I bought a polyurethane one and it’s pretty squeaky.
I don’t really have much experience with power transmission capabilities of any other belts. My knowledge of the HTD comes from ordering the custom pulley from B&B Manufacturing. They are a great resource and do custom pulley work for a reasonable price. They will calculate the HP capabilities of your setup and help you achieve your goals. I too like to source from quality suppliers. My McMaster Carr account could fund a small third world country. I don’t think I have ever purchased anything from them that ended up being cheaply made or Chinese garbage.
A few hints about rigid tapping. I’m not sure what vfd you are using but a vector vfd is recommended. These allow you to maintain max torque at low rpms. Not sure how your controller works but mine has settings for how far the spindle turns before it comes to a complete stop, spindle acceleration and direction change delay. I didn’t think much of these settings until it would not work. With my vector vfd all of these setting are set in the VFD and you can just transfer the settings to the controller. I used a lenze ac tech 2hp vfd. I bought this used and reconditioned. I also put the matching brake on it. I wanted the spindle to stop faster from higher rpms and the brake built into the VFD could not do it to my liking without going over-current.
I have not experienced any or at least enough backlash to cause any issues. However, in my setup belt stretch or belt backlash are not an issue as far as positioning is concerned. This is why having the encoder on the spindle shaft is more desirable. Mine is really only capable of 1/2” UNC. This size puts the mill power to the test. I always use a peck cycle with all tap operations. With smaller sizes it prevents tap breakage and at larger sizes it allows larger diameter taps. If you don’t intend on tapping larger than 3/8 then I would say that you would be fine with the encoder on the motor. My controller suggests at least a 1000 line encoder. I use a 2500 line.
I have received some criticism on another forum about my use of round belts for my encoder. They are very skeptical about the belt slipping and the pressure it puts on the encoder shaft. Both are valid concerns that I also had when I designed it. However, I use them on my lathe and mill without any issues at all. The reason I went this way is I can make my own pulleys on my lathe and cut them at the same time to achieve a perfect 1:1 ratio. Making timing belt pulleys is very complicated and not worth the time for me. I was using the stock gear on my lathe spindle to drive another identical gear that turned the encoder. The gears were really loud. It is very smooth and quiet now. The lathe uses the encoder on almost every operation when using constant surface speed feeding. So, I think this is a pretty good test of the belt slippage concerns.
As usual I am extremely long winded but hopefully this helps. Please ask if you have more questions. It’s nice to be able share my successes and failures. I would love to see your PDB in action. I have kept an eye on the Priest Tools website and I don’t think they are offering one yet. Are you using one from them?
One more thing you may want to consider for the spindle belt and pulleys is contacting B&B. They may be able to suggest a different belt style that may be better suited for this setup. The HTD belts and pulleys are really expensive and way over doing it power wise.
@TCandee Extremely helpful information. I will reach out to B&B manufacturing and describe my needs. I still do not have a lathe, so I would have to get the pulleys from McMaster machined at the local shop. It looks like B&B has the correct bore size for what I need.
Regarding the VFD, I am using the DURApulse GS4, which has vectorless control and can save a couple different profiles for the settings you described.
I purchased the same encoder as you: TRD-S2500-VD, so thank you for that previous tip.
What is the “correct” way to tie the encoder into the spindle? I can see why people might say it would slip with the round belt, but wouldn’t all types of belts put a slight pressure on the encoder shaft?
What is the name of the component (pictured below) on top of the spindle’s pulley? I have used the notches in it to tension the drawbar against with a spanner wrench in the past. I need to remove it to get at the pulley, and I just wanted to make sure that I don’t do any damage. I backed out the two set screws and its still on there tight. Should I just pull it off with a gear puller? If so, I would have to yank the pulley off at the same time. Any advice?
It is threaded onto the spindle with left handed threads. I believe there are set screws in the pulley too. There also might be 2 set screws in the same threaded hole for a lock. Once all of this is loose it should just pull right of easily if I remember correctly.
Just so you have an idea of how much my pulleys were, I paid about 250 each for completely custom HTD pulleys. You could probably cut that in half using a different type of pulley and belt. I wish I had done this.
The ideal way to connect the encoder would be to have the belt between the spindle and a self supported pulley on an axle. The encoder could then be connected directly to the axle via a coupler. This generally requires more space than a PDB allows.
The lock nut is left handed thread and the pulley had two set screws on top of each other in each hole. I used a spline wrench to hold the spindle still while I loosened the lock nut.
I am considering making a new PDB on mine similar to a Bridgeport. I already have the design for it but I need to design an air powered spindle lock. I have had some issue with TTS pullout lately in some heavier cuts. This one would utilize R8 tool holders. To be continued.
@TCandee I just purchased the 2" Shear hog and I’m also worried about pullout. Please keep me posted on your new PDB design.
Also, I spoke with B&B manufacturing today, and the lady I spoked with was less than helpful. I had asked her some of my questions that I posted here, and she basically said that B&B does not offer design help. And I should use the design your own pulley configurator, which was only marginally useful.
Do you have any information on the pulleys that you purchased? Catalogue numbers or drawings?
My recollection is that I used the “design your own pulley” then they contacted me with a design review and we adjusted to make it work. I believe I had to remove the hub from the design and a few other small items then they made them. I never really talked to them about different types of pulleys. Disappointing that they won’t help you with at least that. I looked my order up and I used the 5mm pitch 25mm wide. I swirtched to the 5mm pitch by their recommendation because I could use stock pulleys with some bore rework. You can see the cost was substantially lower than I told you. I looked at the wrong invoice.
@TCandee This pulley configuration is headache inducing. The key on the spindle shaft is a non-standard size Its 6mm x 6mm. But for a 40mm shaft it should be 12mm x 8mm.
Its 6mm x 6mm. But for a 40mm shaft it should be 12mm x 8mm.