[GreenLead]

I'm assuming you did try to get into the Yahoo group, right? Like I stated, they only allow people have have already bought products to register/sign up. Dave even deleted a question I had posted on his main website.

I've thought about Gecko drives, too. The Will Chapman I mentioned uses them as well. But they're pretty pricey, at least so I remember the last time I checked.

(I need to stress that, living in New Zealand and being a student with a small budget means that I can't afford the best of the best even for the the size of the setup I'm working towards, as much as I want to)

So is support for Hobby CNC really that bad? What a shame, considering it's the cheapest I could get for something that could work easily with EMC2.

[ZipSnipe]

Ya might wanna just save, save , SAVE and SAVE. Increase your budget by at least double (really ya need to quadruple) and ya might get something that is worth it. Also check government auctions, I got my cnc machine for $300(Sherline/Spectralight) brand (basically)new from a school.

Good Luck!!!

P.S. remember when you search Ebay you really should try different search words, example"cnc router,cnc mill, vertical mill, and etc. Plus your boss gave you great advice, become a machinist first , at least a couple of years in the trade just to get your feet wet and comprehension of how things are made.

[TOTALLYRC]

List of Things that you will need to build an inexpensive Mill or Lathe.

3 Gecko 251’s (I just bought 6 at the special intro price) Mariss, Marcus, and the team are something special.
$69 x 3 = $237US http://www.geckodrive.com/product.aspx?c=3&i=14471
3 of these wired bipolar. 3 x $15 = $45US
http://www.alltronics.com/cgi-bin/it...-stepper-motor.
1 Breakout board. $84US Marty is great guy and a real straight shooter.
http://www.candcnc.com/UBOB.htm.
Total plus shipping $367US
This will give you all the power and protection that you will need to run any machine from real basic learning to a surprisingly large machine. When you get a lathe, and you will, it would be child’s play to get more steppers and switch the wires between machines when you want to run the lathe or mill or even a router. Mach three will allow you to set up a different profile for each machine so it would be easy to do. I am not sure about EMC2.

This will run circles around the Xylotex package for only a few dollars more. I have the Xylotex package on my Taig. It works but suffers from mid range resonance which limits the rapids. With the small diameter tooling that you are using it might even limit you machining speeds. I will be upgrading to the gecko stuff that I just bought. I also wouldn’t run with out a breakout board. It will protect the PC and makes it much easier to add relays for motor on and off, coolant, vacuum cleaner, lights, or anything else that you might want to run from the pc while machining.

Software
I would suggest Mach3. The demo version works without a time constraint. Just be a nice guy and buy the software when you can.

Power supply min 15v max 50v min 6.3 amps.
You can gut many a 12 automotive battery charger and get the transformer. Easy to find at yard sales, swap meets, your parent’s basement etc. It is what I did for a small lathe I am building. Mine was 12x2 output so I wired it for 24v which after rectification came out at about 30+volts.
If you can find a 24v charger so much the better.
If you know basic electronics you might even be able to use the diodes from the battery charger. I went with a discrete bridge rectifier in mine because it was so cheap. Just remove the voltage regulation from the circuit.
I even have several old (read freebie) printer power supplies that look like they would be usable. At 32v and a couple of amps each you could use 1 for each axis or gang them up if you identical ones. Someone would be better able to tell you if ganging them up would work.



http://www.radioshack.com/product/in...entPage=search
It is right at the limit amp wise but you could always use 2 in parallel to double the ampacity. I am using 1 as it is for a 2 axis lathe which should be fine.

A couple of Ebay capacitors to filter the power would work.

Try to scrounge all of the stuff that you can. I always let people I know that I am interested in used computers, machine tools and other stuff..
After awhile you will have more stuff that you know what to do with. I had so many old computers and other stuff that when I took it to the scrap yard I got $150US in just old computer cases and CD drives and the like. Scrap prices are sky high in the USA right now.

Machine
As for a Mill, an inexpensive XY Table and a flea market drill press will get you going for short money. It will give you 2 axis machining to start because of the taper holding the chuck is not designed for side loading like you get when milling. Plus there is a lot of cast iron there. If you are doing light machine work such as with a Dremel tool or die grinder, you could adapt it to mount on the side of the spindle and eliminate the taper all together

Or even one of these http://cgi.ebay.com/Cross-Compound-S...d=p3286.c0.m14
And one of these( http://cgi.ebay.com/HeavyDuty-Compou...d=p3286.c0.m14) would work out real well. Use the large one for the table and I think with a little ingenuity you could use the small on for the Z axis by mounting it vertically. A couple of pieces of scrap from a friend or the local scrap yard and you would have something with a fair of mass for short money.
All that being said, for a few more dollars you could just by an X1 or X2.

http://www.harborfreight.com/cpi/cta...emnumber=47158
I am not sure if they are available in you neck of the woods.

Other important stuff.

These are in no particular order

1. Friends to help out.
2. A local mentor is probably the greatest asset you will ever have.
3. Basic hand tools. This is where the friends or mentor come in handy if you don’t have the tools.
4. Read as much as you can. I have learned so much just by reading. A lot of this comes from reading posts that really don’t seem to apply to what I am doing but it can surprise you what you will learn.
5. Realize early on that with that small of a budget it will almost certainly go over budget. You could easily spend 3 time your budget in just tooling.
6. Have fun
7. I find that screaming at the walls really helps out during those difficult times when nothing is going right. LOL.
8. Find places that you get scrap for free. I am lucky because where I work we have the best dumpster in the world. I routinely pull aluminum, brass, 304 stainless, cast iron objects, and steel scrap out of it. Not to mention all kinds of drives and electronics. I was lucky enough to grab several motors as well. Always ask for permition.
9. Local classifieds are a great resource for small tools and the like.
10. As far as the shoestring budget goes, if you buy lager shoes they have larger shoestrings so your budget will get larger. LOL

11. I always save the “walwart” power supplies from any device that you thruway. I use them all the time on my CNC stuff.
12. Don’t be afraid to scrounge on trash day while you are driving around. In the USA it is amazing what you can either pickup as trash or people will put to the side of the road with a free sign on it.
13. Never stop learning. I am 44 years old and I am learning more today than I ever thought possible.


Just my 1.9 cents. We must account for inflation.

[Jason3]

Hey Greenlead,

That's some really great advice from Totallyrc.

Your new budget completely changes the picture - I don't think your original plan to scratch build a mill was ridiculous (maybe it was a stretch at your original budget tho ).

For those not in New Zealand, affordable used or new machines suited for a retrofit are really REALLY scarce. We're a long way from most countries that have this stuff to spare and it's all heavy so shipping cost is horrific. A cheap mill to retrofit could use up half the budget or more easily. An X3 would probably cost $2500 here, an RF45 clone could be $3000 easy.

Scratch built is a different story, you can readily get all the parts you'll need for that - I've done it before (maybe towards the upper end of your price range though, depending how much you do yourself and how long you have to wait for good deals...) Another reason to build it yourself would be the work you intend to do with it - moulds need high accuracy and repeatability, and no backlash. Much easier to achieve if you aren't fighting dovetail ways and sloppy acme screws. Personally, I find scratch building easier, you don't need to work within the limitations of an existing machine.

What sort of work can you do - do you have access to a manual mill? You could even buy a manual mill, make the parts you need and sell it again. How about soldering? As an engineering student, surely this is an ideal project...

Would 330 x 220 x 220 mm be enough travel for you? I can hook you up with some nice THK linear bearings & a set of 20 x 5 mm pitch ground ballscrews - not quite free, but for a good price and hey, they're in the country already I might have a spare power supply that would get you started too.

I found Mach3 to be really easy to set up and use. I haven't tried EMC but would be surprised if it were as easy as Mach3, and this is a big project whichever way you cut it - no need to make it harder than it needs to be.

Best regards,

Jason

[GreenLead]

Thanks everyone!

Just so everyone knows, I've had very little experience with mills, just one day's workshop practice through university. My workpiece came out perfectly, although maybe because it was a easy one to start with (a duralinium spirit level). I'm not aware of any holiday courses available for milling/lathing in general.

As Jason3 has stated, it sure is difficult to get stuff to here in NZ from overseas. I looked at Machinery House's (www.machineryhouse.co.nz) catalog the other day (page 17), and saw that the X3 goes for NZD$2,095, the Super X3 for NZD$2605. Quite steep, considering that I'm not sure if I could retrofit them.

I'm not sure about Mach3, as I heard Windows-based control software tends to run into real-time issues. And I don't think I afford it. Since I'm familiar with Linux, I've already started playing around with EMC2 - mostly simple parport pinouts and stuff.

I'm quite wary of the G250s' and 251s' - looks like I'll need a BOB, and considering the cost of three of those plus a CandCNC BOB, I might as well get the G540 because I nearly smoked my parport with my scratchbuilt Allegro 5804 drivers (the setup works, but too slow for EMC2 to handle etc), and I'd rather keep things foolproof for someone like me.

So is the G540 a good buy?

Would 330 x 220 x 220 mm be enough travel for you? I can hook you up with some nice THK linear bearings & a set of 20 x 5 mm pitch ground ballscrews - not quite free, but for a good price and hey, they're in the country already I might have a spare power supply that would get you started too.

Now that I've noticed you're also from NZ, I'm beginning to see the light at the end of the tunnel ;D

The travel seems about right, as I'm only doing tiny plastic toy molds and small mechanical parts.

I'll have a think about a few things here and there, but I really appreciate everyone's continued help and support. Keep giving me those suggestions and advice!

[TOTALLYRC]

I totally forgot about the G540.
I have not used a G540 but all I hear are good things about them on the zone.
It has 4 g250s inside which are just OEM style g251's.

3 Gecko 251’s (I just bought 6 at the special intro price) Mariss, Marcus, and the team are something special.
$69 x 3 = $237US http://www.geckodrive.com/product.aspx?c=3&i=14471
3 of these wired bipolar. 3 x $15 = $45US
http://www.alltronics.com/cgi-bin/it...-stepper-motor.
1 Breakout board. $84US Marty is great guy and a real straight shooter.
http://www.candcnc.com/UBOB.htm.
Total plus shipping $367US

For $367USD plus shipping you can get one hell of a system.

If you are unsure of you wiring skills, the g540 will make the task easier and save a few dollars to boot. You can't go wrong either way.
Plus you get the benefit of having a fourth driver for a 4th axis when the need arrives.

[GreenLead]

I totally forgot about the G540.
I have not used a G540 but all I hear are good things about them on the zone.
It has 4 g250s inside which are just OEM style g251's.
For $367USD plus shipping you can get one hell of a system.

If you are unsure of you wiring skills, the g540 will make the task easier and save a few dollars to boot. You can't go wrong either way.
Plus you get the benefit of having a fourth driver for a 4th axis when the need arrives.

I see. Okay, so that's something I'll read up more about ASAP.

I may need motors of a higher holding torque (say 260~300oz-in), unless the 175oz-in Lin Engineering 5709M-02 unipolar steppers you suggested is already enough to drive a metalworking hobby mill.

[Stepper Monkey]

I may need motors of a higher holding torque (say 260~300oz-in), unless the 175oz-in Lin Engineering 5709M-02 unipolar steppers you suggested is already enough to drive a metalworking hobby mill.

Depending on the size of the mill, these unipolar motors may actually work quite well for you wired up in bipolar series. Those are six wire steppers, and I think Lin uses white and yellow for the center tap wires. If so, just tape them off and only use the two other wire pairs for series bipolar.

The game isn't about initial holding torque, its about how far up the speed curve it can keep it. 175 oz/in is actually more than enough if the torque curve is relatively flat and extends out to a good rpm, which this one does. This will give you great performance and torque up to about 150rpm even at only 24v before you get any real torque roll off at all, and even much faster than that at higher voltage (and I would strongly suggest higher if your controllers will handle it).

It looks like it will pull 2.1A at 5.8mh/ph in series - this motor would be about perfect if it could run in parallel (4.2A, 1.5mh/ph) but it only has six wires so you have to work with what you've got. In any case the inductance is still halfway decent and it can certainly be offset a lot by going to 36v or 48v.
Not precisely what I would pick if starting from scratch, but not actually a bad choice for a small mill at all. Since you already have it it can't hurt to try, I think you'll actually be pleasantly surprised by what it can do.

[GreenLead]

So that means that, as long as the torque doesn't fall off at the usual range of speeds the steppers are running at, they'd be okay?

Oh, and I actually haven't got those motors, or anything at all - I was merely answering totallyrc's suggestion.

By the way, does anyone sell steppers with DB9 connectors that will work with the Gecko 540? Or will I have to wire up my own, and deal with shielding and stuff?

[Stepper Monkey]

Steppers don't come pre-cabled, it's no big deal though.

The 540 ships with the DB9 connectors needed to interface with it, you do have to wire them up to your steppers though. No big deal, 18ga four conductor shielded cable is stupidly easy to get, even cheap foil shielded will work. People tend to go nuts with overly huge wire gauges, but if you actually do the math 18 ga. works perfectly fine for the current you'll be jamming through it. You can go 16 ga if you have it available, but there won't be much of a resistance difference so don't bother hunting it out specially for this. Either one works fine. Leave the car battery sized cables for cars...
Just remember to ground the shielding (or the drain wire if the cable has one) to the connector end, obviously don't run them right next to your data signal cables, and keep the cable to a reasonable length and you'll be fine for EMI. Not much more to it than that really.

[TOTALLYRC]

That site has more motors than the ones that I suggested.

Steppermonkey is exactly right. Plus I didn't realize that if you run the six wire motor biploar it is in series but what do I know? LOL.

What you need is a motor that will turn the screws of the machine so you can make parts. I suggested those motors for 2 reasons. Price Which was the major concern because of your budget. And reason 2 was availability, as I was just doing some quick reasearch to show that there are very reasonable priced motors out there that don't come from Ebay.

My Taig has 425oz motors which sounds impressive but based on there inductance they need about 82 volts for max performance. 8 wire in paralel. Frankly there performance stinks with 30v on a Xylotex board.
After I bought My Taig, which I bought from someone who really didn't know what he was doing and I got a great deal, I really learned a lot. After doing reasearch and learning the formula for max stepper voltage
(32 X Sqrtmh= Max Voltage) I soon figured out why my machine was so slow.

The formula I belive comes from Mariss of Geckodrive. It states that the max voltage that you can safely put inot a stepper motor is 32 times the square root of the motors inductance expressed in micro henries. My rule of thumb is if you can't get that number before you buy the motor, don't.

As a general rule as the stated voltage of a motor rises the inductance goes way up. A 20 v stepper is usually way out of the range that you can use and have it go anywhere on the voltages that we use. Thats why if you look most motors that would work are rated in the low single digits as in 2 or 3 v.

In the G540 manual it gives you a nice range of motor inductances to use.

My machine is so slow for 2 reasons.
1. 30? volts into a motor that ahs a max voltage of about 82.
The Xylotex board doesn't have midrange resonace compensation. I have read that making a "rattler" will really help(search for it on the Zone) but I like better electronics over bandaids.

Since you don't have a lot of tools hanging around, it is going to be tought to do what you want, why make it any harder.

I will be doing more reasearch in the next couple of weeks to find an inexpensive motor that will work on my taig. you just don't need anything as large on a taig with 20tpi screws. I just may buy the ones that I suggested for you and try them out with the Xylotex board first, and then with the ag251's I just recieved.
Of course this is after I finish theretro-fit of my BTC-1 which is about ready to make parts.
I have found that for steppers at least you are better using a smaller motor at close to max or at max voltage and suppling them with there rated amps than running a larger motor at 1/2 volts and less than name plate voltage. I think Mariss said it best all you are really doing is turning nema 42 into a nema 23 when you run it way below specs or something like that.

I will probably be selling the Taig as a complete unit before Chrismas as I will need to buy a bunch of tooling for my BTC-1. If I do I will post it on the Zone but It will probably go for more than your budget.

Keep reading and keep asking questions. As my father used to say "The only stupid question is the one you haven't asked"

Mike

[GreenLead]

Stepper Monkey: Sounds great!

Totallyrc: I played around with that formula too last night - Mariss is really good at explaining things clearly.

But I guess the most important thing I've learnt about was the mid-range resonance - which may explain why my old UCN5804 driver struggled to push steppers beyond 400Hz (in hindsight, all my old steppers were the round cased rare earth magnet type, which Mariss noted weren't as good as the newer square-cased PM steppers)

So yeah, the G540 sounds like a good buy overall, with optoisolated parport and inputs etc.

Just wondering, though. Has anyone worked with E-stop in EMC2? I started a thread for this a while back (http://cnczone.com/forums/showthread.php?t=51212), figured something out that sounded reasonable at the time, but need clarification:

Basically, I came to the conclusion that parport pin1 was EMC2's Estop out, which via an optoisolated BOB/the G540 would be wired to a relay that kept the spindle and other aux devices (wired in series) on. When EMC2's Estop in (parport pin 10, usually LOW) goes to HIGH from a Estop button being hit, E-stop out goes low and shuts down the relay, and with it the aux devices.

Now I know the G540 has an Estop for its four axes, and it can also accept switch inputs that get sent back to parport pin 10, so I should be able to, with some wiring, use one Estop switch to kill the G540 and tell EMC2 to stop, right?

[cadmonkey]

EMC2's EStop can be configured numerous different ways, high or low for any signal and not all signals need to exist even. You can certainly configure it to do what you just suggested.

That said I'd comment that whether the default setup you seem to refer to (there are defaults that appear in Stepconf but you can change them to whatever you need, I believe that some of the manufacturers and hobbyists have just followed a basic standard at some point in the past so that arrangement has just been applied as the defacto standard - I made my own Opto-Isolated BOB so I put things in what seemed like a logical order for me and easiest to etch on a single sided through hole PC board wihout needing a lot of jumpers to route signals while still keeping a ground isolation strip between traces) in the config is low or high, you take into consideration shorts and broken wires - what you just described has the estop switch sending a HIGH signal to the machine so a broken wire, dead power source, short or other malfunction could result in no Estop signal being present. Safe practice dictates that you set it up to be a LOW signal seen at the controller is the Estop condition. What needs to go in and what EMC puts out are configurable either way. Looking for a low signal would protect against a dead power source, broken wire, popped fuse, short, etc. IIRC my Gecko drives (203v) required a low signal to be active which I didn't like, my thought is that they shouldn't go live until they have their supply power and a positive voltage signal from me (the control computer) to go live and start resolving position pulses.

Many people (from the EMC mailing list atleast - myself not included though - at least not on a benchtop mill - a knee mill like a Bridgeport and I'd probably agree with them) prefer to shut EVERYTHING down on an Estop except the computer, and use the Estop switch to run a charge pump (looks for a constant X frequency signal present as a watchdog) and if that shuts down (which if the computer goes haywire or the estop were pressed would happen) then the power supplies and drivers and spindles and everything are shut down.

I use the charge pump to disable the power source to the drivers and spindle but not on Estop, only when EMC is shutdown or lost (ie - if estop doesn't stop the problem for some reason then I have to hit the PC reset or power button) and Estop just disables EMC from sending any signals. If the computer gets lost (which it did once while loading a really complex .ngc file since my PC is an old curbside find - it's never gotten lost running a program, that one it got lost on was a complex engraving that had multiple loops to take something like 100 passes in a loop for each path using the O word commands EMC has available - it was the Detroit Redwings logo so if you're familiar with the Winged Wheel you know there is a lot of fine detail to that resulting in detailed program and the loops were too numerous for the machine and it timed out trying to figure it out, I tried it on my test machine - my office desktop setup dualboot in EMC and it loaded quickly without issue - P3 650mhz w/ 380mb vs a P4 2.8 w/ 3GB) the charge pump signal stops and all the drives shut down. This also allows me to have the computer on and work on other things without the mill being powered up, so long as EMC is off, and I can't power up the mill without the PC which prevents extraneous signals from doing anything unexpected. So basically I have one power switch (the computer) and then EMC is the virtual power switch for the mill.

Just my cautious suggestions - being safe now is easier than getting patched up later. YMMV.

Greg

(edited [repeatedly] for some clarity - I tried to touch on so many different points and this got LONG - I just recently took my meds so my brain isn't fully with it - sorry - hope you understand where I am trying to steer you )

[GreenLead]

cadmonkey: Hmm...so what would be the suggested wiring method to what you have described? A wiriing diagram may explain it to me better.