[JohnDeere630]

@ Handlewanker...before I forget, you will note that I have the bit completely penetrating the work-piece, and doing 99.9% of the cutting on the side of the bit. This may or may not be making the work easier. I have not tried to cut pockets, and have no idea how things would go if most of the work was done by the end of the bit. Everything I have needed so far has involved cutting shapes and holes out of 1' x 2' sheets of material. Nothing 3D as of yet.

[handlewanker]

Hi, in my opinion....... and for what it's worth as I'm a complete nooby with CNC, even though I spent the last 50 years with conventional machinery,....CNC is taking many small cuts at a shallow depth which means you can move faster than with a deeper cut, which also means the machine can handle the cutter loading without putting a strain not only on the frame but on the expensive ball screws and linear rails.

I was always of the thought that you let the machine do the talking and you do the walking to go get some tea while the "death by many cuts" takes place....LOL.

The old way with manual machines, even with an autoi feed, means you are standing there waiting for the cut to finish and this entails taking the most with the least number of depth re-sets....time being money etc.

With CNC, I see some people are talking about heavy cuts and extreme high feed rates etc.....has the CNC plot been lost in translation?

This to my mind is counterproductive to cutter life and also means you stress the machine beyond it's design capability, that is unless you have a heavy machine costing in the region of $60 to $100K due to industry needs for competitiveness....if the other guy is doing it then so must I etc.

I think that as CNC is a computer controlled affair with the Human element out of the picture machining wise, then the program can handle all the cutter resets and you just pick up the end product without so much as a finger laid on during the complete cycle.......time being as it occurs.

I may be totally wrong in that reflection, but as it's kind to cutters and the machine, unless you have a high throughput and seconds on a job mean less bucks in the bank, taking smaller cuts and more of them would be the way to go, especially with machines made from aluminium extrusions and not altogether that rigid.
Ian.

[JohnDeere630]

@Handlewanker, what you say makes sense. I am somewhere in the middle, with relatively deep cuts, but at slow feed speeds. I don't see much for side-load on the cutter, hence the load on the relatively small ball-screws should be fairly light. You can't really put much load on that tiny 1/8" bit, anyway, as it will just snap off. I am also no expert, just a guy trying to get from A to B without spending a fortune in the process. Were I a millionaire, I'd have the finest 5 or 6 axis CNC machine on the market, along with a real engineer to program it. Alas, like most, I cannot afford champagne on a beer budget. I am happy with this machine, as it does everything I need it to do at a reasonable cost. I don't need it to be real fast, but since I have to stand by to extract the cut shapes and circles from the work-piece as they are cut free, lest they flip up and jam the cutter head, I don't really want to wait for it to make 10 passes either. Also, at the slow feed speeds, I don't have to worry too much about dragging that spinning bit across the back of my hand. If this machine is totally worn out in a couple of weeks, then I guess I'll have to suck it up and admit I was wrong. I do wish there was an easy way to inject grease directly into the linear bearings and ball-screw nuts......

[JohnDeere630]

@Handlewanker...Ha! I reread your post, and saw where, like me, most of your prior experience was on conventional machining tools. I was never an expert on them, either....more of a dilettante. My experience with the smaller CNC mill at New Balance was brief, hardly enough to overcome old habits of heavy cuts with big bits at low, hand-wheel feed speeds. The modern CNC mindset does indeed seem to be about very shallow cuts, but blinding fast feeds. It certainly looks impressive to watch, but for a boring old fuddy-duddy like me, plodding from A to B fits the bill. It's the end results that count for me, not the dazzle factor.....

When I get caught up, and have more time, I'll try it the modern, CNC way & see if it works out any better or not.

[blinkenlight]

My understanding (and I'm the first to admit I could easily be wrong) is that feeds and speeds do not depend on your depth of cut: what limits it instead is your spindle's power to keep spinning somewhere close to the nominal RPM. The forces induced on the machine might be a factor though...

[JohnDeere630]

My understanding (and I'm the first to admit I could easily be wrong) is that feeds and speeds do not depend on your depth of cut: what limits it instead is your spindle's power to keep spinning somewhere close to the nominal RPM. The forces induced on the machine might be a factor though...

That is my take as well, which is why I opted for the 1500 watt spindle. As long as the spindle has the power to drive the bit edge into, and remove material fast enough, the lateral stresses are negligible, with the tiny bit being, IMO, the weakest link. Keeping the bit cool and lubricated also becomes a major priority at these extremely high contact speeds. As Handlewanker says, CNC is a whole 'nother kettle of fish from conventional machining, it seems.

[handlewanker]

Hi, I have so much to learn, even G code is something I haven't fathomed out yet.

I would assume that with a moving gantry router, and a lightly built one too, the feeds and speeds would not be the same as for a heavier CNC machine, be it a router or a mill......that is why I suggested that a shallow depth of cut and a relatively fast feed rate would be the way to go........I have to try that out when I get into the CNC World with both feet.

By all accounts the 6040 seems to be a versatile machine and best bang for the buck anywhere around.

I'm currently in the middle of making a 3025, or to be more precise a 300 mm X 250mm X 100mm router, now gone to the model stage prior to a costing and material sourcing analysis, but that will have to be on another thread.....it's small compared to the 6040, but that is the size I want .....EBAY 3020's just aren't much to write home about.

I intend to get my feet wet CNC wise one way or the other, without going to the extreme of buying one....once the bug bites it's all go to the end.
Ian.

[JohnDeere630]

Hi, I have so much to learn, even G code is something I haven't fathomed out yet.

The g-code itself is pretty simple; it is getting to the point of compiling it that has the steep learning curve. There are 3 steps; first is to actually do the drawing. Real engineers use AutoCAD, but for a hobbyist or small business whose needs are fairly simple, AutoCAD is way too much money. I use Inkscape. The second step is the CAM part of CAD/CAM, where the drawing is converted to cutting paths, and the third part is compiling the tool data and feed speeds, as well as the dimensional coordinates into the form of g-code. I do the second part also in Inkscape with the Paths menu, and use the gcodetools extension to actually compile the code. My needs are very simple 2D cutouts of relatively thin sheets of metal, so this system works very well for me, and all the software is free. Were I milling out complex 3D shapes, this system may not work out so well. Drilling the end-mill straight down, completely penetrating the work-piece, and then moving the bit laterally and quite slowly to cut out the pieces is a far cry from cutting deep pockets in complex shapes, and that may be why some people here don't want to believe my results. Folks can believe whatever they wish; I'm not here to fight or argue with anyone, especially since most here know a lot more than I ever will about CNC in general. I just know what works for me, and I couldn't be happier with the results. Maybe I'm doing it all wrong, and my machine will detonate and spray aluminum shrapnel all over the garage at some point, but I have a whole box full of parts already milled out, and the machine is very close to paying for itself already, so a few more months and I will be able to afford another one anyway.

[handlewanker]

LOL....thanks John.....my first learing curve would be to decide when to switch the machine on and what will happens when it fires up.

At the moment I'm in the waiting lane waiting for a mill sourced from China.....long story, but in the meantime I'm starting to chafe at the bit and have taken the plunge to design a 3025 CNC moving table router initially to get into the CNC race, but mainly to get acquainted with the mysterious world of computer generated movements as opposed to inserting all the axes movements manually.

It's going to be a long haul building from scratch, but that's how I want to do it.

I see you mention 2D......I would think that as soon as you start drilling operations and cutting to depths etc, then 3 D will need to be addressed as the next step.

BTW, it has been mentioned that the SBR linear rails are not all that good for CNC purposes, cheap yes, but not when compared to HIWIN types.

As I will be tossing up to decide whether or not to use HIWIN 15mm slides or 12mm SBR round rail types under the table, it would be interesting once the "shine" has gone off your 6040 to see how the SBR slides (you have) handle wear, and if they exhibit any deterioration due to the balls in the bearing block bearing directly with point contact onto the round rail surface.

At the same time, have you detected any backlash in the single ball nuts, as this is a worry I want to eliminate from the outset, not later

I've designed an infinitely adjustable double ball nut using two single ball nuts, but in a different configuration to a dedicated double ball nut as sold for the purpose purely for backlash elimination.
Ian.

[JohnDeere630]

LOL....thanks John.....my first learing curve would be to decide when to switch the machine on and what will happens when it fires up.

At the moment I'm in the waiting lane waiting for a mill sourced from China.....long story, but in the meantime I'm starting to chafe at the bit and have taken the plunge to design a 3025 CNC moving table router initially to get into the CNC race, but mainly to get acquainted with the mysterious world of computer generated movements as opposed to inserting all the axes movements manually.

It's going to be a long haul building from scratch, but that's how I want to do it.

I see you mention 2D......I would think that as soon as you start drilling operations and cutting to depths etc, then 3 D will need to be addressed as the next step.

BTW, it has been mentioned that the SBR linear rails are not all that good for CNC purposes, cheap yes, but not when compared to HIWIN types.

As I will be tossing up to decide whether or not to use HIWIN 15mm slides or 12mm SBR round rail types under the table, it would be interesting once the "shine" has gone off your 6040 to see how the SBR slides (you have) handle wear, and if they exhibit any deterioration due to the balls in the bearing block bearing directly with point contact onto the round rail surface.

At the same time, have you detected any backlash in the single ball nuts, as this is a worry I want to eliminate from the outset, not later

I've designed an infinitely adjustable double ball nut using two single ball nuts, but in a different configuration to a dedicated double ball nut as sold for the purpose purely for backlash elimination.
Ian.

You make a good point about the recirculating bearings in the hiwins vs. the simpler system of the 6040. I wouldn't expect this mill to be terribly long lasting. I commend your effort to build your own. I may try that myself, now that I am more familiar with how computer signals get translated into stepper movements. The motors and command/control circuitry is dirt cheap, and using my machine in 3D to make some of the parts, it should be possible to make my own, more sturdy machine in the future.

I have not yet, after about 30 hours of use, seen any slop in the ballscrews. Everything is still tight.

Did I have a choice, I'd go with bigger, adjustable ballscrews and hiwins any day and twice on Sunday. Just my 2 cents..... When you get to building your machine, I'd appreciate it if you'd post your progress.....I love stuff like that.

[handlewanker]

Hi, since I started taking an interest in CNC goings on and seeing the videos on UTUBE I got the urge to participate more actually than armchair musing.

To that end the design stage got into gear and now I must start a new thread to "come out of the closet" so to say and start actually translating the ideas I've collected into something that will do the job instead of hypothesising about the merits of various CNC devices that other people have taken the trouble to make.
Ian.

[zoltan87]

First off all hello everyone here!
I have just registered and just started to get my feet wet in the cnc world, so I want to apologise in advance for the noob questions.

I just bought a chinese 6040 cnc machine with 800 w spindle and blue box controller unit (yes I know it's pretty ****ty after reading this thread, but it was a lot cheaper than other versions and also I bought it before I read this thread ). Hopefully it will be ok, it has a different control board in it, not the one that is completely useless. Here is a picture about it:

Attachment 267468

I hooked it up with my pc, it's a modern pc, but I made it dual boot, it has windows 8.1 and win xp on it on two different partitions, so when I want to control my cnc machine I boot my pc in in windows xp. I also had to buy a pcie card with parallel port on it as my motherboard doesn't have it. Luckily it seems like everything is working, BUT:

When I turn on the spindle my PC monitor goes completely crazy, displays a message about some timing issue, then goes completely black. When I turn the spindle off, then monitor off and on, it comes back fine. Seems like the monitor somehow interferes with the spindle. I know most of you guys have changed the vfd / spindle cable to shielded cables. I haven't modified anything so far, and if I can get away then I wouldn't want to spend an other 40 GBP just for a piece of cable. I must add that my pc, my router and spindle are all plugged in to a single extension cord, so if I use two different sockets for the pc and cnc would that help eliminating interference? Probably stupid question, but I have zero experience in electicity, also have no idea what should I ground and how. Any kind off advice / suggestion would be extremely helpful!

Zoltan

[Adverse Effects]

everything you need to know to fic the crappy low end troublesome 6040's (i have one to) it cost me 1/3 the cost of the original unit to fix it and get it running
all i have not changed in the main table unit and the steppers

[JohnDeere630]

zoltan87...not to be a negative nancy, but, I think you are going to have some problems with that blue control box. The pic of the stepper driver board (if that is indeed what it is) has no heatsinks on the driver chips. Heat could very well be a problem, although not your immediate one it seems. I had zero luck going the route you chose with the PCIe parallel port card on a relatively new computer. (IIRC, I had monitor issues as well) I had to build a machine around a motherboard with a PPort built in. Those PCIe chips run different signal standards than legacy PPorts. I had to set the motherboard BIOS to EPP/ECP, if I recall correctly. I also had no luck with the Mach software, so I installed LinuxCNC from here: Download Information (I can help with the stepconf file settings should you choose that route)

I have a similar machine with the black control box and the 1.5K spindle, and am very happy with it. I have about 50 hours on mine, cutting 2D shaps out of 1/16", 1/8", and 1/4" aluminum with a 1/8" and a 1/4" single-flute carbide end mill. I also use kool-mist as a lubricant.

Understand, I am not telling you what to do, just relaying my experiences with a similar machine......

Gary

[zoltan87]

Yeah I was afraid you will say this . My plan is to try to use it as it is, if it won't work out then I will start to change things. JohnDeere630 you said you couldn't get to work your unit through a pcie card, do you mean you couldn't move the machine at all in Mach3 with the keyboard or you could do that but couldn't actually cut out any complex shape? I can move around fine when I push the cursor buttons, but I want to actually try to send translated g code to it to try to cut out something. That's when I will find out if it works. Those heatsink things kind of bother me, although I will cut only polyurethane foam with it for now, hopefully won't overheat. Maybe open the windows, it's like 5-10 degrees outside in the UK now. By the way here is a picture about the inside of the box.
Attachment 267492

The monitor interference is weird, you said you had the same issue? Is it possible it goes through the parallel port to the pc or something? Tomorrow I will try to use different sockets for pc and cnc router, also try to run some code to see how it behaves.

edit: Just checked it again, under that green controller board there are heatsinks, although I am not sure what's the point as air can't really move around there...

[JohnDeere630]

zoltan, using the Mach software, the computer would "talk" to the mill, but it was erratic and jerky. I expect it was probably latency issues with the PCIe PPort card. I ditched the card at the same time I switched to LinuxCNC (went to a different computer), so can't honestly say what the real issues were. LinuxCNC was so smooth and easy (for me) to use, I never bothered trying Mach3 again. If it ain't broke, don't fix it, as it were. I also have no idea what was going on with the monitor (an LCD one, BTW), but got heavy artifacting and distorted screens. That also disappeared with the computer switch, so may have been the PCIe card, the Mach software, because XP is ****ty, or some combination of the 3. I use my mill in a semi-production environment, so was disinclined to do a lot of back-tracking and sleuthing once I got it running; I just put it to work.

I know a lot of people have had trouble with the blue box controllers; there seem to be a lot of issues with electrical noise and/or cross-talk on the servomotor cables, as well as random instances of arc&spark lightshows inside the controller box. I wish I could be more hopeful/helpful, but I managed to avoid that particular bit of trouble (so far) with the black box controller.

How handy are you around circuit boards? I bought enough spare parts to rebuild my control box about 3 times, except for the VFD for the spindle, on eBay. It consists of a parallel port controller card, 3 stepper drivers and a power supply, and wasn't really a whole lot of money..... just a thought.

[zoltan87]

I haven't had time today to actually run some gcode on my machine, will try it tomorrow. If it won't work then I will try to buy a new controllerboard, the G540 seems very popular here. But probably first I will try to buy some shielded cable from somewhere cheap. The place where I saw it sells them for 40 GBP which is just ridiculous.

[tgdavies]

That is ridiculous -- I'm using this cable with my 800W water cooled spindle: Heavy Duty Shielded VFD Cable [CB-22] - US$4.60 : Homann Designs!, The preferred CNC Component Supplier at USD4.60 per metre (not useful for you perhaps, as this seller in in Australia, but shows what specs to look for)

Tom

[hive8]

If you're going to do serious aluminum milling get a mister or some type of coolant set up. At least compressed air to blow the chips away. I've used alcohol in a spray bottle and although it helps with light cuts you'll have issues when slotting in anything 1/4" or more if you can't clear the chips and keep the bit cool

Do you know of a good cheap mister, I need to get one of these for my 6040 since I am doing mostly alu work.

[hive8]

I just finished up the escutcheons from the 3/16" 6061, and everything worked fine. I was using 9 mm/sec plunge speed and 8 mm/sec feed speed. I wound up running the spindle pretty close to full speed, and I had to increase the airflow a bit to clear the chips vs. the 3003 alloy.

I think I made a mistake earlier...I was cutting the 1/4" plate at 6 mm/sec, not 8; pretty slow, with plenty of air and kool-mist.

Was reading over the what you doing really great, I also cut aluminium 6061 without a problem, could you post the link to the cutters again, I have been using full carbide 3mm and 1.5mm, didnt try 6mm yet.