Pros and Cons of home shop CNC

[Jasonb]

The point was mentioned about how much a CNC might stand idle once a "casting" had been created on it until the next part came along. I've just looked through the drawings and photos of teh James Coombes that I have just started to write up.

There are 52 different parts, that does not allow for multiples eg the columns are counted as one not four.

Out of those 10 had some machining done on the CNC, run times ranging from about 2hrs down to 2minutes.

Of those 10 parts only 3 would have been supplied as castings, other work as will be seen in the build used the CNC for shaping parts that would have been done on a rotary table which I seldom use now as the CNC is the best tool I have where one curve blends into another.

So with the first photos of progress being early November I've been working on that model for 3 months and it has just run, there was some other model work going on too so not my only project but a good indication. over that time the CNC got fired up 8 times as I did a couple of items at the same time, the last item shows it being machined on 1st January so idle for a month.

Now lets take as an example someone with a high quality lathe that they do most of their work on and a larger capacity ex industrial machine that sits unused for  a lot of the time but when a large part needs machining it is used as it is the best tool for the job. Same if you have a small lathe for delicate work. I see this as being the same as having a CNC you may not use it all the time but when a particular job comes along it is the tool of choice.

[vtsteam]

Just to illustrate how on the fence I am:

Since I don't particularly enjoy manual milling, I'd probably convert my round column mill to CNC, but that seems like a long project, (even though I have the controllers and steppers to do it already) if ball screw conversions etc are needed. Also I wonder how good this mill would really be as a base for that. I did add a DRO last year (built the control -- it's a bluetooth to phone DRO) and that was a huge step forward in usability.

My present lathe is my second homemade lathe, after the Gingery, and was designed to be massive for a small footprint. It is a 9" swing by 15", but not much bigger than a 7x12  Mini lathe on the bench. MUCH heavier, and stiffer with a good sized boring table.

For that, I've 95% finished my own-design electronic lead screw using a stepper, a 600 line rotary encoder and a FORTH programed Arduino. Yet I have not electrically connected it, even though the stepper is already mechanically installed, and the electronics and programming have all tested out. No idea why I've procrastinated on that. I guess it was just a flagging of interest toward the end. Maybe I'll get inspired to finish by talking about it with you all here. I've been doing without power feeds or the ability to  single point thread for 5 years, now, for pretty much no reason. 

[Mike R]

I forgot to add another reason I got into CNC machines - at the time I was travelling ALOT for work (200+days a year) and I had more PC time than shop time so I figured a way to leverage my PC time into something beneficial to my hobby endeavors was to play with CNC - CAD and CAM so I could "in theory" pop into the shop and run a program when I was home.  It didn't quite work out that way, but I still enjoy the path CNC has taken me on.

[Jasonb]

Agree that the CAD CAM part of it can be done at any time you have spare such as that odd half hour when you don't fancy going out to a cold workshop, while half watching something on TV, etc.

The job can also be left "set up" on the computer so you can come back and do a bit at any time then once the code has been spat out by the post processor it's just a case of mounting the work and pressing Go. You can always break the paths into shorted run times if you don't want to leave the CNC running for a long time and then just come back to that and run the next bit(s)

[Hugh Currin]

Here are a few random thoughts:

I find a CNC mill far more useful than a CNC lathe. If it was one of the other I'd take the mill hands down. I mainly got into a lathe conversion for the ability to cut inch and metric threads easily without change gears.

The cost of CNC is less if you convert a manual machine to CNC.  One advantage of doing this is you know that machine inside out. This, however, quickly leads into another hobby/rabbit hole. It is a lot of work to figure out the details and make the parts for a conversion. But it's much easier to track down problems and fix them if you have intimate knowledge of the machine. Alternately, purchasing a CNC machine costs considerably more but is obtainable now. Since they are purpose built they will likely be more accurate and better optimized. There are two in-between paths. There are now conversion kits for some popular mills, such as my PM728. This leaves you responsible for the electronics and controller. Putting that together is a bit of work but a lot quicker than starting a design from scratch. The resulting machine is on par with a economical ground up conversion (i.e. inexpensive rolled Chinese ballscrews etc). The second is to buy a used CNC machine with an outdated or dead controller. Then replace the controller with a new PC based one. This isn't done commercially because of the time involved, so outdated/dead CNC machine is quite reasonably priced. But it requires your time to do the conversion. Easier to do if the axis drive motors are usable, even better if the motors and drivers are both usable. (How good are you at digging though electronic diagrams?) This will result in a, better than hobby, commercial quality CNC. One problem here is they'll be large and likely require 3 phase power.

I'm sold on CNC machine in my shop. I likely use then more in "manual" mode that running programs. Using the jog option you can move any axis under control at a specified speed. Push the button it moves, release the button it stops. And in MDI mode which accepts single line G-code commands. For example move to X=3.250 at 4 in/min feed. Likewise gives digital readout on each axis. I have not gotten used to using "conversational" mode with has built in routines for common tasks, like a bolt circle. My controllers, LinuxCNC, doesn't have such built in. For these I use a CAM package. I use CamBam and Fusion 360. Tilting more towards Fusion but each has its place.

Most model plans are designed to not use CNC. Most parts are simple and tend towards rectangular or round. The complex 3D shapes where CNC shines are avoided because they're very difficult to make on a manual machine. Model designs, and their plans, are made for home shops which, by a large margin, don't have CNC.

You may not need CNC to do fantastically complex parts. George Brittnel and Ramon, plus others, carve marvelous complex engines from solid on manual machines. I can't do this but can get, for me, complex parts using CNC. I like doing this.

On castings. I believe castings came about for large parts. At scale the waste in material and time/effort to cut from a large chunk is excessive. For models this isn't the case. The reason to start with castings is the look of a casting in the resulting part and the exercise of finishing cast parts. Cast parts start as a mold plug which is "easy" to make with curved edges. Thus castings tend to have a more flowing shape than manually cut parts. Gives a nice look to a model engine. To get this when cutting from solid requires a lot of work and/or CNC. Castings for models are, to me, very expensive. A lot of work to get a small casting compared to a chuck of stock.

CNC is a different mind set than manual machining. For most of us it required a CAM package to interface between the electronic model of a part and the G-code program the CNC needs. You lose some feel for the machine using CNC, but there is still feedback in sight and sound. But for me, the "manual" modes a CNC offers is a good alternate to a manual machine. A professor where I used to work was happy to teach CNC classes without machines, just a CAM package. All the rest of us thought this was foolish. So many things can go wrong between the CAM and a finished part that you must teach the whole process. A CAM model is the start not the final product. To have success with CNC almost requires some manual machining background, and certainly machining time on a CNC. There is a great deal of skills transfer between manual machining and CNC, and vise versa.

I'm happy with my CNCed shop and just wish I had more time/energy to use it.

You all have a great day. It looks like a great day here. Thanks.

[FKreider]

The cost of CNC is less if you convert a manual machine to CNC. 

I think this can be very dependent on where you are located in the world geographically.

Where I live in the U.S. there used to be a very large amount of manufacturing/industry, a lot of this industry has since moved to other parts of the country and more commonly outsourced to the far east. As a result of this you can go on Facebook marketplace any day of the week and find an old Bridgeport mill for $800.

The same can be true for older CNC model machines, you can find very good deals if you are patient and jump on them when they come up. I couldn't buy the controller and servo or stepper motors for the amount that I paid for my milling machine with working and installed 2-axis CNC capabilities. I also have a Hardinge chucker lathe with Omniturn CNC controls that I purchased at an auction for less than you can buy a typical used manual lathe for.

That said I know in other parts of the world like Canada or Australia that same old worn-out Bridgeport that is only worth $800 in my area would easily get $3000 or more.

Based on this I would suggest that the cost of CNC machines MIGHT BE less if you convert a manual machine to CNC. As you suggested conversions/retrofits can be a bit of a rabbit hole.

[Djangodog]

I learned to program an NC lathe many years ago, (paper tapes, no internal memory or crt screen).  I moved on to cnc turning and machining centers and manually programmed four axis simultaneous machines.  I was a journeyman tool and die maker at the time, but I didn’t want to get left behind as the industry evolved. 

As capable as these machines were, they could never compete with a good jig bore.  Circular interpolation is not accurate enough for a bearing pocket or anything that needs to be perfectly round.  Angles and tapers are more perfect when done manually as well. 

A cnc has good value, but if the programmer does not understand how to properly cut the part, then there is little chance that it will turn out well.  Surface speed, chip load, chip evacuation, rigidity of setup and an order of operations to insure accuracy are still necessary.

My only cnc equipment is a 2 axis Proto Trak on my Bridgeport and a cnc servo feed threading unit on my Hardinge HLV.  There are times when circular interpolation would be helpful on the lathe, but not necessary, (I have a ball turning attachment that does a nice job). 

If I was in a great hurry or needed to make money on production parts, then I would take the cnc.  For the pleasure and satisfaction of making something, I prefer my manual machines.

[Jasonb]

I see most of the tooling manufacturers do "Finish boring heads" that are set to a specific size and after using interpolation or insert drill to remove most of the waste the finish head is loaded from the tool changer and used to do a final sizing pass. Not really any different to reaming a hole after drilling or boring and always possible to do some test cuts first to set the head to the exact size you need for the job.

typical heads

One in action about 3.30 into this video. I like the parting off at the end too

<a href="https://www.youtube.com/watch?v=I--_63_wysk" target="_blank">http://www.youtube.com/watch?v=I--_63_wysk</a>

Not really practical to have a range of these in the home CNC shop but nothing to stop you using a boring head for holes that matter just adjusting like you would on a manual machine and the CNC provides a nice power vertical feed be it for a one off hole or the same hole in a small batch of parts that you may find in a multi cylinder model engine.

5th axis would also make the angles and tapers as good as setting the work at an angle on a jig borer as the work can then move in a straight line rather than a series of minute steps. I know not many have 5 axis in their home shop but many have a 4th axis capability which would do on most occasions if the setup was right.