Show how to build a CNC machine from the very beginning to the end

[Skyfire]

very good work skyfire, i started reading this post of yours a year or so ago, and then you went quite for a while. im so glad you came back and have now nearly finished this machine, i've thoroughly enjoyed reading its progress.

your attention to quality and detail is superb, and your posts have been very informative, i've learnt a lot from it.

i wish you the best of luck in selling this machine to many customers in the future, im sure you will be successful in doing so. i would love to buy one myself, however i plan to build my own!

how much are you planning on selling these for anyway?

Hi, Thank you for your inputs here. Yes, I had paused this building for a year because of other projects. But I will certainly finish it this time

Thank you for your wishes of my machine business. I think the machine itself has a very good cost/perfomance ratio but still need to keep my way of paying attention to quality and details. I will also need to do much on every details of marketing, delivery, services etc. to make my machine win public praise. I think it's the only way to win the market and set up a new brand here. So... I will just keep hard working and wish myself good luck here too

I want to say somthing here about the price/cost issue because seems it's getting important now especially you guys have many choices now on several brands CNC, or convert/build a CNC on your own, or based on a manual mill.

My first point here is the price item as you asked. We know the economic conditions are really not good now and we should try to save money anytime. I realized from my former experience that if I want to win some market I must make the product most valueable to customers. That means I will try hard to give as many as poosible of "good features" and try to control the price reasonable or even low to customers. It's easy to achieve any one of these two ends, but hard to get them both. I guess many suppliers now are attractive with low price but many defects hiden in the products that you hardly to find until some bad things happen. That's what I will never do on my machines.

But how can I provide a machine with good price? The only way for me is to save as much cost as possible. You guys can see that I designed every part by myself; made molds by myself; getting parts in reasonable price based on my former supplier resources; and many hard works on manufacture optimizing. So these really saved much developing fund to make me be able to control the price. untill now, my basic price target is $3500-$3700 for comleted machine including the wood shipping package. I think I can reach this no problem now. I would like to keep this way for best cost performance if can get you guys maket support.

Another item is the building/conversion options like you would like to do now. I'm a big fun of DIY also. I really would like to build everything I want to have. My career of CNC field actually started from a conversion of SX3 years ago. It's a exciting thing to make something by own hands. Yet I fould two disappointing points after many DIY works.

1. Always short of tools and the result always not that satisfied. That's really funny When I converted the SX3 long time ago, I found too many things I needed beyond the converstion kit. I didn't know the backlash, I didn't know the travel error, so I need a dialgage of 0.01mm. But the dialgage is not enough to measure the spindle error, so I need another dialgage of 0.002mm. Then I found I don't know the perpendicularity and parallelism, So I need a high accuracy referance stone rular. And I realized at last, if I want to build a good CNC, I need a real CNC first! Of course I couldn't purchase so many tools then, so the result is that the conversion was finished, but I knew it's just working but not good enough.

2. Always spend much more money than you expected if you really want do this job right. I understand building is an exciting thing. But because most DIYers are not quite known well about the total conversion plan cost. The initial plans are always rough, and you will find many items need to improve then during your works---but must go on to finish it or just drop it. So, more money invested or original invest wasted. I experienced this a lot and really wasted much money. I have many various parts may never be used any more but just piled up in house.

So.. I generated the plan of building a complete CNC machine and bring to market to earn some money, and save customer's money together. I hope this will work for all goodness. And, I will say that if I don't have a big CNC and many other tools now, I will not be able to build this benchtop CNC satisfied.

[Skyfire]

Thanks for this thread, Skyfire.

I am in the middle of putting together a somewhat similar mill. A bit heavier, 170 pounds base casting and 170 pounds for vertical column. I did go with lost foam casting, so surface is not as nice as yours, still need to machine it. Gray iron casting prices here in Canada are also a bit different... For a total of 480 pounds (a bit over 200 kilos) I had to pay $1100


A couple of questions: How much did you pay for those bellows? Who is the manufacturer? Are they English-speaking?

Also, I remember reading somewhere that it is better to mount air cylinder in reverse direction, but maybe I am wrong.

Thank you,

Alex

Hi Alex. I think you really will spend much on your building.. I guess at least $5000 cost for all works done of machining, parts, assembling etc. ( but I don't know the exact configration you plan to do) The casting cost on your side is about 5 times than in China. I guess the machining cost will be much more expensive also.

I casted the machine body in a realated small casting factory but good quality. They can't speek english and have no website at all. Only the locals know their exist. You know any big factory with website and can speek English will not take such small orders unless you have really volumn orders of tons.

And, You are quite right about the gas struck installation. It's the recommanded way to install it in reverse direction. The reason is that it will be lower the friction and have better buffer features.

I didn't use it that way because the lower end will be very dirty with chips and coolant. This will be much bigger problem here. But thank you for noticing this anyway

[alex.d]

My casting is done already and I already have arrangements to get it machined for a relatively good price.

Currently I am only interested in way covers. You mentioned that you sent your bellows design to some company who made it for you. I could order custom made bellows at gortite.com but I have a feeling that getting it done in China is more affordable.

[diyengineer]

AMAZING! Very impressed. So your thinking of sending a few to the US for sale?

[Skyfire]

My casting is done already and I already have arrangements to get it machined for a relatively good price.

Currently I am only interested in way covers. You mentioned that you sent your bellows design to some company who made it for you. I could order custom made bellows at gortite.com but I have a feeling that getting it done in China is more affordable.

Yes. the way covers are made by my supplier the same way as castings. but the reason they did this for me is that I sent this with my another volumn order of decades sets like the steel one I posted. Hum.. When you have your sketch I might be able to help on this.

[Skyfire]

AMAZING! Very impressed. So your thinking of sending a few to the US for sale?

Hi diyengineer. Yes. I'm considering a batch sales in US now. Just to find the best way to do it now. I'm quite busy on trade mark desgin, website building and many service issues setting up etc. now. You know, once it's put to market, everything must be very formal but not just a building anymore. Frankly, this is very serious thing to make sure of customer and dealer's benifit.

Thank you for your attention and welcome any more discussion. I will need suggestions and I'm very appreciate that.

[handlewanker]

Hi, at the price you mentioned, I think I'll pawn the family jewels, but at my age they're pretty well worn...LOL

BTW, will there be any other spindle tool holder options available??

It would be good if it could come with an ER40 spindle end, or ER32 at least, so that with a 1"or 25mm collet (ER40) or a 19mm or 3/4" (ER32) you can do a QCT method by having 1"or 3/4" plain shank tool holders (like Tormach TTS) with tools already mounted and still not be sticking too far out of the spindle end.

I realise the spindle with an ER end and cutter direct in the spindle is the most ridgid set-up, but as the cutting forces are going to be small, I think a plain shank tool chuck in an ER40 or 32 spindle will be OK.....you only lose approx 25mm of Z travel but gain the enormous versatility of being able to have a QCT set-up.......for me, ATC is not considered to be as important as QCT with this build.

I have an ER32 chuck on my Ajax mill and am contemplating adding some smaller 20mm plain shank chucks to give me a QCT capability.

The ER32 chuck I am using has a hole all the way through the body, so to give a depth stop control to plain spindle mounting I intend to locktite in a hardened steel button behind the collet for the plain shanks to butt up against as the TTS method (against the spindle face) cannot be used against the face of the ER nut.

What kind of shipping charges would you expect to have to Australia at present?
Ian.

[Skyfire]

Hi, at the price you mentioned, I think I'll pawn the family jewels, but at my age they're pretty well worn...LOL

BTW, will there be any other spindle tool holder options available??

It would be good if it could come with an ER40 spindle end, or ER32 at least, so that with a 1"or 25mm collet (ER40) or a 19mm or 3/4" (ER32) you can do a QCT method by having 1"or 3/4" plain shank tool holders (like Tormach TTS) with tools already mounted and still not be sticking too far out of the spindle end.

I realise the spindle with an ER end and cutter direct in the spindle is the most ridgid set-up, but as the cutting forces are going to be small, I think a plain shank tool chuck in an ER40 or 32 spindle will be OK.....you only lose approx 25mm of Z travel but gain the enormous versatility of being able to have a QCT set-up.......for me, ATC is not considered to be as important as QCT with this build.

I have an ER32 chuck on my Ajax mill and am contemplating adding some smaller 20mm plain shank chucks to give me a QCT capability.

The ER32 chuck I am using has a hole all the way through the body, so to give a depth stop control to plain spindle mounting I intend to locktite in a hardened steel button behind the collet for the plain shanks to butt up against as the TTS method (against the spindle face) cannot be used against the face of the ER nut.

What kind of shipping charges would you expect to have to Australia at present?
Ian.

Hi my friend. I see...I will keep my wallet tight too when I'm not yang any more.. Haha Thank you a lot anyway for so many great inputs here!

Yes. The spindle holder can be actually any size of ER series, ER32, ER40 are both avaliable. I think should go with them when formal manufacture.. I will take your suggestion definatelly~

I agree that a QCT is very good for the real appliacations as you mentioned. It's a smart way smiliar to what Tormach did. I think I will try to build up such things next for my machine to give best convenience to users. So that meas at least ER32 and ER40 is the best.

Hum.. As my export experience before, The shipping charges to Au are similiar to US and CA. around $100 for no more than 1m3 wood case. So. totally the shipping charge should be okey.

[Skyfire]

I just measured the spindle errors today. the spindle are pretty smooth after grind in now. the radial runout is 0.012mm and the axial runout is < 0.01mm. I measured them both at the taper side and the tools. Just very similiar error. I think the spindle should be better when some batch production with more standard tolerance control. all should be under 0.01mm.





I also received the last piece of PCB I need--the small panel PCB. I welded one board and connected for test. It contains 4 axis free buttons and LEDs to show machine status.



run MACH software. normal status the green LED will light up.




I trigered a travel limit switch and the green LED went out and the red alarm LED light up, and also a yellow over range LED.



Actually I run all of the 3 axises today and tested all signals except the spindle driving port. I will finish all tests very soon.

And the cooling fan I planed to install:

[handlewanker]

Hi, that's very good news......the ability to have any ER spindle plus the extremely moderate shipping charges.

As the Bishop said to the Actress, "I don't know how you do it for the price", it is now looking like a must have at all cost item.

Just for the record, is there any other machine on the market with similar specs that approaches this price structure?

BTW, Skyfire, are you going to put the first machine on Ebay to see how the interest goes with the bidding?

You'll have to have a name for the mill....not too many alpha/numeric characters please....thinking of the SX3 etc for simplicity.
Ian

[alex.d]

The T-slot table looks pretty dark in latest images. Did you coat it with something? What is it?

Also what is your T-slot configuration (e.g. slot size and distance between slots)?

Do you plan to install an automatic oiler?

[DRock]

Skyfire,

The build looks great! Can you take video? I would love to see this thing in action.

[Skyfire]

Hi, that's very good news......the ability to have any ER spindle plus the extremely moderate shipping charges.

As the Bishop said to the Actress, "I don't know how you do it for the price", it is now looking like a must have at all cost item.

Just for the record, is there any other machine on the market with similar specs that approaches this price structure?

BTW, Skyfire, are you going to put the first machine on Ebay to see how the interest goes with the bidding?

You'll have to have a name for the mill....not too many alpha/numeric characters please....thinking of the SX3 etc for simplicity.
Ian

Hi, I think for the whole machine .. I almost thought over every details a CNC may related. So.. I didn't see a one better than mine for now that covering many detailed features. Except that I have no a famous brand yet But on this machine driving system I'm only using the basic but very excellent stpper driving system and P-port controller to make machine most economic but enough for most appliacations. Servo driving system are also avaliable for update too.

I think I can try Ebay.. I don't have experience on it but can try~ And my own website is under building now..It will be the main window to market I think.

Yes. I have the name of the machine now...I will bring out the name when the machine completely finished. cheers~

[Skyfire]

The T-slot table looks pretty dark in latest images. Did you coat it with something? What is it?

Also what is your T-slot configuration (e.g. slot size and distance between slots)?

Do you plan to install an automatic oiler?

Hi, the worktable looks dark is just the light problem. I took most pictures in middle night and the light is really not good. And the table is coated with dark color thick oil.. so it's another reason.

The T slot is standard 10MM slot. using the M8 or silimar inch size clamping parts. the center distance between is 47mm.

No, the automatic oiler is not considered. Oiling the blocks and ballscrews manually is enough.

[LeeWay]

Excellent. You took just a little too long though. I ordered a torus from Novakon instead. I needed a second machine and the price was right with them.
Who knows though, I may need a third one at some point too.
Excellent work on this one. I can guarantee people will love the ways. Mine is similar with all linear ways. No gib adjusting with those.

[Skyfire]

Skyfire,

The build looks great! Can you take video? I would love to see this thing in action.

Hi. Yes I will take some videos soon when do some real cutting..I will upload to You Tube for watching.. Actually I have run the machine up today with MAX rapid moving speed of 3400mm/min. And the spindle controled by MACH software is tested too. I run a demo G-code today for long time and everything is good. I didn't take a video today because my camera is full.

I have ordered some clamping tools and mill tools today and will do some real job with this machine. I will take the videos then. Please just pay attention to my updates and welcome any inputs.

[Skyfire]

Excellent. You took just a little too long though. I ordered a torus from Novakon instead. I needed a second machine and the price was right with them.
Who knows though, I may need a third one at some point too.
Excellent work on this one. I can guarantee people will love the ways. Mine is similar with all linear ways. No gib adjusting with those.

Hi my friend. Nice to meet you today. It's a pity not catching your order time..haha.. But I have to take every step necessory and make sure the whole machine is no problem. That's the only way to build a succeful machine for the first one and then next ones will just be standard... I really hope can catch time of your third machine!

Mine is so near to finish now. Actually it's full operatable now. just need several metal sheet parts: the control panel and the spindle top cover. I plan to mill the panel on itself. I will do this in 2 days when I receiving the right tools. Then I will do some more real cut tests and final approving the machine value. :cheers~

[Skyfire]

Today I just finished the stainless covers for worktable and way cover brackets. And also the cable tubes.. All of them are just as what I wanted.

worktable stainless cover and front cover bracket.




Head side are also covered with stainless so all linears sealed. And the spindle motor wire tube.


very cleaning now




intial tested with a demo G code. no tool installed.


So now only two main parts left: spindle box cover and the control panel..

[diyengineer]

Just a thought, but maybe make a small spash/ cover for the stepper motors. If you were to drop a tool, vise, or jig, it may very well bend that 2nd shaft protruding out. If you left them exposed, i assume there spash proof, but doesn't hurt longevity wise to keep them clean, atleast on the X and Y axis.

gorgeous machine, can't wait to see it run.

[Skyfire]

Just a thought, but maybe make a small spash/ cover for the stepper motors. If you were to drop a tool, vise, or jig, it may very well bend that 2nd shaft protruding out. If you left them exposed, i assume there spash proof, but doesn't hurt longevity wise to keep them clean, atleast on the X and Y axis.

gorgeous machine, can't wait to see it run.

Hi. Yes it will be best that the motors are best protected by covers. I didn't cover the whole motor bodies because I think the heat dissipation is a much bigger issue. The stepper motors will be hot especially under long time working and maintain MAX current for big torque. The stepper motor will demagnetization under high temperature and lose torque then. So I didn't want to cover them totally for lower temperature. because the end shaft sticks out only about 25mm, so it's not big chance to be damaged. That's my thought. Thank you anyway for this suggestion. I will see if can have some way to do it better. cheers!~

I'm preparing some G codes today and I will do some real works from tomorrow.. The current plan is to mill it's own control panel firstly