PDA

View Full Version : Kiwi junk-box mini router...



kiwichris
02-24-2007, 07:10 PM
OK....

Only been on the site a wee while, but it would seem like a good idea to share my journey. :)

Basically been dithering for aobut 5 years on making a router, never got around to it as I kept on designing things too complex to actually build at home. I got keen again about three weeks ago, did some googling, found the rockcliff design, which lead to this site.

I've studied lots of photos, and a couple of the free plans, and decided if I studied too many more I'd never get started, and in five years time I'd still be collecting parts and dreaming of cutting bits of balsa wood.

I'm a bit short to buy bits, but really wanted to get something going, so this is what I'm doing with stuff from the garage. (I collected lots of stuff over the last few years, so I've got a lot of bits in my garage!) It'll be too small for most of what I'd like to do, but I'm sure I'll learn lots along the way.

The first pic is six of the ground shaft/bush pairs I've collected from old printers etc. Selected at random because I had pairs of them the same length/diameter with bushes. Yes the bushes should be longer for this type of use, I'm over it.

Measuring those led to the second pic, which I drew up in qcad. Appologies for the metric measurements for those who don't live in france. Total travel going to be about 230x180x60 or there abouts. Going to swing my trusty old dremel tool on it, see if I can blow the bottom bearing out.

Pic 3, the garage/workshop. after making one long cut I discovered that my skill saw blade has a wobble, so all of my cutting after that is/will be done with the drop saw, it's dead square, and makes a good tidy cut. The pile of recycled hardwood on the bench is a coffee table I should have been making for the wife. If anyone is talking to her, that's what the noise from the garage was last night, OK? Anyway, the basic frame parts are taking shape in front. All 18mm MDF unless otherwise noted on my un-dimensioned plan. I'll add dimensions as I need em. I have a 1:5 copy of the bench for reference, and I rounded all the measurements to 5mm increments to make it easy...

Pic4, rod support bracket thingit, 1 of 4 for the base. Did them this way so I can retro-fit larger rod when I get some. The X rod (This is going to be X) is only 8mm, but with the short 230mm span it should be OK for light cuts. We'll see. block idea courtesy of this site (joes & jgro & others).

Pic6, the only critical bit of the X axis, getting them vaguely parallel. I clamped, measured, re-clamped, measured and then cyno'd the edges of the 9mm MDF. Then I glued the bushes with a little cyno, super-screwed the 9mm down, and super screwed a wedge against each bush. The wedges can come out to put bigger bushes in at a later date. Got this right first time, it's as close as I can measure with my trusty steel rule to parallel.

Pic 7, tada! 3 hours later and I've got an X. (an unknown value). Blocks screwed and glued with PVA, ends screwed & glued the same. The whole thing slides quite well, and is reasonably rigid. I've got some 'U' Channel aluminium I'll add to the base later to keep it all square and solid, and to provide a better mount for the gantry.

last pic. Skip to the other shed (less dust, more electronics), and it's alive!

I put a blacklash nut (not anti-backlash!) made of a bit of cutting board offcut screwed between two lumps of MDF on the bottom of the table, and mounted a stepper on the end.

The stepper mount needs work, as I only had two M3 machine screws that fit the job, and my rod coupler is currently masking tape, leading to interesting amounts of whip and funny noises from the M6/1.0 threaded rod.

I can get a reliable 720mm/m (28ipm) out of it, with a reasonable level of push left for cutting (Finger preasure push/pull test) stepper stalls jogging at around 900mm/m(35ipm). Steppers are old printer ones rated at 3A/1.8V driven by a knocked up L/R half stepping unipolar driver at about 2.5A from a 12V supply. (Yes, the resistors are running hot!)

Software is EMC installed of the live Ubuntu build, on a junker celeron PC. Also tried it with turbocnc, which seems to 'glitch' while stepping a little on this PC, although it runs fine on an old laptop I have. Weird.

Proper leadscrew coupling, a better nut and better drive should get me 1m/m (40ipm) or better. Looks promising so far. :)

Total new bits purchased so far: small container of M6/30 machine screws and nuts for the rod supports ("gutter bolts"), and an M6 tap, as I couldn't find one in my box-o-taps, although I'm sure I have one somewhere, now I've got two!

There'll be a bit of gap before I get anything more done, as I have a busy week ahead, next installment will be next sunday probably.

Thanks to this site for the bit of inspiration that got me off my chuff after so long.

Arthur Clampitt
02-25-2007, 07:04 AM
Interesting , keep up the flow of pictures , Along the lines I was considering before I decided to cnc my existing mill/drill ( Still in the "gunna" stage :-( )

Coogrrr
02-25-2007, 09:42 AM
I actually have 22 old printers now that my son gets $5us for each one he disassembles and sticks all the parts into a bag (one for each printer) and 2 copiers to do also.

I am encouraged by the bushing solution more now for the small ones.

Keep up the pics and let me see if I can match your build (although I may stick to HDPE instead)

Coog

dan dimock
03-03-2007, 12:27 PM
I would like to know if you made your own driver or board and what about the power supply.I am geting some old HP printers model # 3 that has 100oz motors that I plan on using, keep up the good work, and keep us up to date.
Dan

toneV8
03-03-2007, 01:48 PM
kiwichris: do you have CAD drawing od machine.

kiwichris
03-03-2007, 03:00 PM
Hi-ho,

Dan:

The controller I'll be using is a simple Unipolar mosfet job, with resistor current limit... Yes, I've built it myself as I'm trying to do this out of the junk box as much as possible. Although my electronics junk box is fairly substantial as I used to run a part time business doing electrical/electronic design work... I'll post some pics of the controller later on when I update my progress on the machine itself. I don't have a circuit diagram, although I can scribble one and scan it if you're interested. I have used three PIC16C84's as the stepper translators, and can provide the source if you're keen on that as well, although you could use one larger (more pins needed) processor for all three axes if you're not lazy like me. :-).

For the power supply I'm using an old PC power supply, 12V rail for logic (Using a 7805 regulator to get 5V) and 5V rail for the steppers... When you're using a resistor limited driver anything more than around 2x the rated voltage of the steppers is really just waste heat, and my steppers are rated at 1.8V. I'm not too worried about performance with this one obviously... If it turns out that it dosn't work well enough I'll design a unipolar chopper driver for it, but based on my simple testing so far I think it'll be fine for light cuts in wood/plastic which is all i want from this machine.

Tonev8: Yup, I'll post a DXF later on, it's a bit vague as I've been winging it to some extent, but the basic layout and dimensions are mostly correct. I've been editing the cad file as I go to reflect the machine. :-) They are 2d only though, and don't show fasteners etc. As I don't currently use a 3d package, as I found I was wasting too much time drawing things!

I got some more work done on the machine yesterday, I'll post progress pics and descriptions tonight.

Cheers, Chris H.

kiwichris
03-04-2007, 05:05 PM
Hi-ho,

Some progress from the weekend….

Pic 1:

I decided that the masking tape lead screw couplings had to go, so got out the wee lathe, and started to turn up a M6 to 1/4 stepped coupling.. This is as far as I got before the belt In the lathe shredded itself. Hmmmm.

I bodged up a belt from a couple of large O rings plastic welded together, and managed to finish one coupler, but will have to get a belt to finish the other two.

With the coupler installed in the X from last week there is now almost no whip at all, and the odd ‘whop whop whop’ sound as the screw whacks against the wood is gone. I was kinda liking that sound..

Pic 2:

Some fabricated bits for the Z axis, and stepper mount/offset bits. Made much MDF dust and noise to get to this point.

Pic 3:

The Y carriage. (The flat bit of MDF will be known as the Y carriage from now on, OK?) Notice the three ‘saddle’ bush holders and the one offset one. I found out pretty quickly after I tightened the M5 machine screws that getting things lined up perfectly in MDF with the ‘through hole’ mounts for the bushes was going to be near on impossible, and it made the entire thing very tight to move.

I simply trimmed off one side of the bush mounts with a scroll saw and re-mounted them with the bush hard against the Y carriage (The flat bit, remember?). I understand now how folks have arrived at solutions using routed slots in the ‘carriage’ to align the bushes, but I don’t have a table router or table saw, so I might CNC up a better carriage with this machine when it’s finished!

The machine screws are too long because I only wanted to buy one length of them, being a tight wad.

Pic 4:

Skip forward a wee bit and we have a finished Z. Here you can just see the nice shiny aluminium coupler. I stuck two M3 grub screws through it, and if they don’t hold I’ll add another couple and some red loctite.

You can also see my fabulous ‘Backlash nut’ solution. (not anti-backlash). I figure it will get things going, and if the machine shows promise of actually doing real cutting of something harder than sponge cake I’ll splash out on some ptfe or acetyl to make some better nuts from. Probably acetyl as it’s quite a bit cheaper here in NZ from the calling around I did last week.

The Z axis is a bit stiffer to move than I would liked, as the bushes are a little tight, and I think there might be a slight bow in the Y carriage. It moves fine in testing though.

Holding on to the carriage it moves up and down without stalling at 400mm/min (15ipm) with a 1.5kg weight hanging on the bottom plate. (The motor and end plates will travel up and down with the spindle).

Spent a little time playing with EMC settings as well, to see how the acceleration figures change performance etc.

Pic 5:

Another delay and we have the ‘Y’ rods and bushes mounted on the other side of the carriage, and the front plate of the Z axis mounted. Not to forget another fabulous cutting-board lead nut assembly.

Here the Z is resting on X, showing Y. (This is draft for the beginning of a Dr Suess book…)

The lead screw protrudes through the left end so that I can add a bearing to that end if stepper end-float becomes an issue. Not sure if it will for the speeds and materials I want to work with, but it gives me the option without changing too much.

Pic 6:

For Dan, a picture of the home brew controller. I’m using three separate pic 16C84’s (16F84 replaced them, these are old) as stepper translators, driving into some LM324 opamps I’m using as level shifters for the standard mosfet gate drive.

Logic level mosfets wouldn’t require the op-amp, but I didn’t have 12 logic level ones I wanted to use in this project. The fets are BUK 456-60A’s. A bonus advantage of using the LM324’s is that they have a slowish rise time, and protect against Dv-Dt failiure of the mosfets, so no series resistor is needed for the fet gate.

I’ll be using a separate board for the six current limit resistors (1Ohm 10W). I’m using the 5V rail from an old PC power supply for the motors, and running the logic from the 12V rail of the same supply via the LM7805 on the board shown. I’m only using 5V at this stage to reduce heat loss, and anything much more than 2x the rated voltage on a LR controller is just wasted as heat anyway. The small steppers I’m using are rated at 1.8V 3A.

I’ll also add a buffer board for limit/home switches and estop. The whole thing (Three bits of strip board and the PC power supply) will get screwed into a case if I can find one in the garage, or more than likely a bit of MDF as I have lots of that.

Software in the pic is pretty simple half step code. I’m might add an ‘enable’ input that I’ll drive off EMC to reduce static heat build up.

I almost fell into the trap of re-inventing the wheel and creating a microstepping chopper circuit for this, but in reality kits like the hobbycnc one are cheaper than I can buy the components, so I’ll save my pennies and buy something like that for my next attempt. Either that or I’ll get brain fade and design something anyway, just for the sake of it!

Also attached is the DXF file. The drawing is a work in progress, so don’t expect high art… I’ve been updating the drawing as I go, so it’s currently got some vague in the area of the Z/Y carriage, and some dimensions may be patently incorrect and/or different from what I’ve built.

Total new bits purchased so far: M6/30 machine screws and nuts, M6 tap, M5/50 machine screws & nuts/washers.

Lessons learned this week:

It is better to drill/tap MDF than use wood screws for such small bits when going into the ‘end grain’. The threads hold pretty well, and you wont split both end plates for your Z axis and have to re-make them. (Ahem)

Pretty bush supports made of a (very slightly) compressible material like MDF are not much good, doing up the screws distorted them enough to misalign things.

‘O’ rings can be used as a lathe drive belt in a pinch with application of soldering iron to weld them together!

Next steps: Alloy bracing bits underneath the bed, and gantry. Finish up the controller so it's usable off the bench. More lead screw couplings.

Cheers Chris.

toneV8
03-05-2007, 12:22 AM
kiwichris: thanks for CAD drawing.

Does it z axis move only 50mm, whay dont you make at least 80mm.

kiwichris
03-05-2007, 01:15 AM
kiwichris: thanks for CAD drawing.

Does it z axis move only 50mm, whay dont you make at least 80mm.

Hi-ho,

Your english is fine... :)

The drawing has 80mm of travel, as per the image, the blue is the fixed carriage and the pink slidesup and down with stepper motor/spindle attached. On the real machine I'll get about 70mm due to the stepper mount screws sticking down from the top plate of the motor support / Z axis assembly.

Cheers, Me.

dan dimock
03-05-2007, 07:54 AM
I am going to attemp to build a board like the one you are using, Will get my printers taken apart this week and see what size motors I can find, and may need more help to get mine going, I really like what you are building and the fact that you are keeping it simple without all the high dollar items that most are talking about.
Thanks for showing others that it can be done.
Dan

kiwichris
03-05-2007, 01:44 PM
U'r welcome. :wave: . I enjoy giving back to sites like this, as I get so much out of them..

For what it's worth, the steppers I'm using here are line feed motors from a Sekoshia BP5420 printer, the 9mm rails I'm using for Y were from a couple of Brother 1409's and I think the 8mm ones came from OKi microline 84's. (what a great printer.) I got two of the dead BP5420's from my old work in the 90's (used to service them, their controller boards were a nightmare discrete design which failed when you looked at it) and another from a recycling place that strips down computer gear for metals etc, so I got two for free, and one for $10.

I also have some larger Nema 23 double stack motors that were the carriage motors from the Sekoshia printers. The Sekoshia was a large dot matrix, an excelent source of bits :). Also has some good 20mm shafting, and the power supply transformer is a mother of a thing I'll probably use in my next machine with the nema 23 double stack motors, and microstepping chopper controllers (Probably my own design, but we'll see how silly I feel on that count, re-inventing wheel and all that).

For what it's worth, if you have to buy all the components individually to make your controller board it may be cheaper to get one of the kits around, especially if you value your time at all. I have a fairly good stock of components, so building the controllers for me is a matter of picking components I have and making a circuit to fit what's in stock.

Good luck with your junk box build. :)

kiwichris
03-07-2007, 03:25 AM
Spent some time in the shed yesterday, got a bit more done. :-).

Made up one more lead screw coupler, and did some work on the controller..

Pic 1 is the controller powered up with all three motors for the first time, and the current limit resistors I intend to use temporarily mounted in the terminal strips. You can also just make out the lead screw coupler on the X motor...

I'm getting a measured 2.7A per winding, but the static voltage is a little higher than rated, so the motor that wasn't moving much (Y in this case) was getting really hot.... Almost too hot to touch after 10 minutes.

The 10Ohm load resistor is to get the powersupply to start, these old PC switchmodes wont run without a load. Should use a higher resistance, but I only have 1Ohm and 10Ohm 10W resistors to had... Will change that before I complete the job.

This is the first time I've used these steppers near their rated 3A. I've messed about with them a bit at lower current levels, and never had any heating problems. I might have a bit of a re-think on the voltage/current decision, and go to a lower current, but higher supply voltage, which should give me less heating due to less static voltage on the steppers.

Pic 2, A screenshot of emc2, the reason the 'Y' motor got so hot. This code is a simple jpeg to gcode conversion I wrote in a shell script, the X is constantly moving, so was only just warm, Z was moving a bit, so was warmer, and Y only moves at the end of every horizontal line, roughly every 15s, so got toasty warm.

With some basic number crunching, with a 1Ohm resistor I'm getting 2.7A, therefore 0.8Ohm in the stepper winding, which is a higher resistance than the implied spec of 3A at 1.8V printed on the stepper, which would give 0.6 Ohm. Either way, the 2.3V across the stepper when it's not moving is getting it nicely warm. Disipating 11.6W of heat in there somewhere, which dosn't sound too bad. Hmm, might post something over in the stepper forum on this one..

Watch this space..... :-)

kiwichris
03-12-2007, 05:21 PM
Not a lot done in the last week really, spent my free time on some house renovation/building projects...

I did however get some 3mm carbide spiral upcut two flute end mills (What a mouth full!) from a local supplier, one ball nose and one square, which will keep me happy while I test the machine out once it's in cutting form... I'll probably want some 1.5mm square ones for cutting model bits, but that can wait till I've actually managed to cut something, anything. Cardboard would do! :)

As per the pic I also dragged out an old AT formfactor PC case, and removed everything that didn't look like a CNC controller.

I've added the seperate resistor board to the setup, and changed to using the 12V rail. With 4.4Ohms per winding I'm now getting just under 3A. I also had to switch power supplies, the old AT one (which came with the case, for free) was only good for 8A on the 12V rail. The 'new' power supply is a 400W ATX job, which is good for 16A on the 12V. Which dosn't require the load resistor on the 5V rail to run, as a minor bonus.

I wired the stepper outputs back through some DB9 connectors on the back of the case, and I'll run a ribbon cable out for the parallel port input as well to make it all tidy.

I might also make up some sort of ducting for the two 80mm fans to keep the resistor board cool(ish) as it disipates about 80W if all six phases are energised and static. (3 motors x 2 phases). Havn't thought too much about that, I got as far as yanking the fans out of another failed project and connecting them, I thought that was pretty good.

I ran the 'tort' sample gcode from EMC through the three steppers on this setup for about 30 minutes. Nothing caught fire with just one of the the two fans pointing in the general direction of the resistor board, so with a little ducting I think it'll be a good thing.

Total new bits purchased so far: M6/30 machine screws and nuts, M6 tap, M5/50 machine screws & nuts/washers, two 3mm end mills, 12 2.2Ohm 10W resistors.

Please note: If you're looking for an example of how to build a 'good' cnc controller, look elsewhere. :rolleyes:

Next: Ducting for the controller, and back to making the machine itself!

Cheers, Chris H.

dan dimock
03-12-2007, 08:08 PM
All I want my controller to do is to run the unit.
I have been reading all of this for over two years, and still don't know anything.
I was hopeing that with all the know-how you are showing, you would come up with plans and details so we could build one that would work for guys like me that gets a small disabilty income.
Keep up the good work and thanks for the pictures and all the details you are posting, I may learn all I need to get one up and runing.
Dan

kiwichris
03-12-2007, 09:36 PM
Hi Dan,

I'm quite happy to knock out some plans etc for the controller if you're keen, and it could actually be simpler that what I've built for this machine if I spent more than 15 minutes making it.. :).

Come to think of it, it would give me an excuse for building a better controller rather than 'just for the sake of it'.

When you say you know nothing, what's your general electronics knowledge like, on a 1 to 10 scale, 1 being 'What is electronics' and 10 being 'I just built my own MRI machine'.

While a simple stepper driver is just that, simple, there is a low level of required knowledge about components etc that will increase your odds of success tenfold, and will definately keep you alive when it comes time to wire up the power supply to the mains. (flame2)

Cheers, Chris H.

dan dimock
03-12-2007, 11:30 PM
Chris;
When I went to college, in the 50th I work at a repair shop, doing TV and Radio repair, I know parts and the value of them, but cannot set down and draw a cirtuit that would work. I am now 72 and in a power chair, had a cabnit shop for over 40 years with as high as 32 people working for me.
I got a computer about two years ago and have been learning a lot since then. I can find the parts and know how to put them together, just need to know where they go.
Thanks
Dan

kiwichris
03-13-2007, 12:57 AM
OK. Sounds like a plan then :). I want to get the junk box router done, and at least partially working, then I'll knock up some drawings and prototype for a simple stepper controller using common/cheap bits, rather than junk from my collection!

Cheers, Chris H.

kiwichris
03-14-2007, 08:36 AM
Did a wee bit more this evening....

As per the picture 1, mounted the base onto a couple of bits of alloy channel to hold everything rigid, and to allow the gantry and bracing to have something solid to screw through.

You'll note I drilled/tapped the holes for holding the bracing bits on this time around, as per my previous post. It's better than using wood screws when going into the edge of the MDF.

The alloy is just some scrap bits that were being used to hold up some office partitions, I scored two long bits of it about 5 years ago. It made a huge noise cutting it in the drop saw.. :rolleyes: (Yes, I was wearing ear-muffs)

I've also mounted the X stepper, or is it Y? Hmmmm, I was calling this axis X, but I think I'll call it Y, and make the gantry axis X...

The second pic is just the finished layout of the controller. I mounted one fan as per the pic in the 'back' of the PC case, and mounted the other in the front, blowing out through a new hole cut for the purpose. Without ducting this seems to keep the heat from the resistors under control.

Havn't bought anything new this time around...

Just need the gantry, X motor, spindle mount, and hold-downs on the bed and I'm done.. Getting close. ;-).

kiwichris
03-15-2007, 09:48 AM
Righty, I can almost smell it.

Oh, wait.

Vege soup for tea, pardon me.

Anyway, as per the pic.. Made up a quick'n'dirty mount for the dremel to live on the carriage, and mounted it with appropriate hose clamp.

Screwed a bit of 9mm mdf down to the work area for the sacraficial, and put a row of tapped holes along side it, just outside the work area for hold downs of some sort.

Made up the sides of the gantry and mounted the X rods.. The X rod adjustment blocks/mounts are on a 45 degree angle to allow access to all the bolts, down side being it's a little tricky to adjust the rails for smooth movement. Never mind though, after 10 minutes of fiddling it seems to be fine..

Had to put the blocks for one end of the X rails inside the gantry upgright, as there isn't space for the stepper... This is one thing I would have discovered if I had have done full cad drawings, but then I'd stillbe doing cad drawings.... Not to worry, I only loose 18mm of X travel, not a biggie.

I just need to make another lead screw coupler up, and mount the 'X' stepper and I'll be in business, or at least that's the theory right now....

Oh, and extend the stepper cables, install limit swtiches, shift the PC to the garage, etc etc.... :-).

Cheers, Chis H.

dan dimock
03-15-2007, 10:08 AM
MAN"

That is looking great - I would like to have one 12x12 to cut gears for a all wooden clock - I have been working on one (a clock) for two years and the gears are killing me.

Please post a picture of the back, I would like to see how to fixed the Z axis.
Dan

kiwichris
03-15-2007, 06:03 PM
One rear photo....

Cutting clock gears by hand? My level of hand-tool skill and patience would dictate I'd give up before I started!! :o

The rockcliff design is 12x12x4, and is the same type of machine as this, although a lot more polished... The down side for me with the rockcliff is that it requires a table saw or router, neither of which I have access to easily.

FYI, I intend to re-build the Z/X carriage once I've got the machine cutting, as it's the least 'tidy' bit.. I'm not that happy with it, but in the scheme of things with the tools/skill I have with wood I think it's pretty neat to get this far. :)

As I built the carriage I could see it was turning into a bit of a mess, but without a table saw or router I couldn't see a better way of doing it.

I'm semi-confident that it'll last long enough to cut itself a new carriage, which will fix the mess. I'll cnc some channels in a bit of MDF for the bushes to snug down into, and some small saddle pieces to hold them. From the photos of the rockcliffs on here I think that's what he's done, it'd certainly be tidier than what I did!

Cheers, Me.

dan dimock
03-15-2007, 09:03 PM
I have all the wood working equipment that you will find in any shop, even a metal lathe. but I am lost when it comes to the electrial and computer end.

I have not looked at the rockliffs, I may try to look at his site if I can find it.

What I like about yours is that you have taken what you had and turned it into a unit that will build you one that would cost BIG BUCKS if you went to buy one.

I have people that builds wooden clocks that will give me the clocks already put on disk where, if I get a small CNC machine, I do not have to have a cad program.

I have two daughters that I want to build a grandfather clock for. That is the main reason I want one.

I like what you have done, my hat off to you.

Dan

kiwichris
03-16-2007, 09:20 AM
Well, What a cool time in the garage I've had tonight.

As per pic 1, I got the X axis lead screw and motor mounted, and moved the computer + driver case from the small electronics workshop, to the spare bench in the garage.

I had a little bit of difficulty getting the lead screw aligned, due to some less than precise cutting of the cutting board off-cut, and the rather silly absense of a thread... ie: I forgot to tap the hole, although that didn't stop me trying to put the threaded rod through it anyway! :rolleyes:

After some basic testing, joging etc, pic2 happened..... I put a small bit of MDF under the screws, and yahooo! it cuts. :D :D :D That was just with the jog keys, set to 100mm/minute.

I then broke out deskengrave, and created the code to cut pic3... 3mm deep at 150mm/min. No problem. It's a bit 'hairy' right after beeing cut, but a light swipe over with 200 grit paper and it looks sooooo cool.

I need to lengthen the stepper leads so the controller case dosn't have to sit up on top of another computer, and create some sort of cable holder for the dremel power lead, but outside that I'd say it's a going thing.

I have about 0.4mm total backlash in X, and slightly less in Y, although putting 0.4 into the emc backlash settings over compenstates for it quite a bit, so I suspect the slop is somewhere else, possibly the steppers, or flex in the gantry due to the lack of bracing.

Have to put the bracing on it now. Ooo, and I can cut a better mount for the dremel....

I'll do some experimentation with speeds etc, although this is never going to be lightening fast as the driver/motor combo only gives me a rapid speed of 450mm/min.

I'm sure many folks on this site already know how cool it is to see your new baby come to life. If it weren't for the fact that it's 2am here, I'd be positively buzzing right now... I'm close but that may be the sleep deprivation talking...

I'm working for part of tomorrow, but I'll see if I can actually make something with it tomorrow, not sure what... Anything will do! :)

I'll continue to post to this thread as I do the 'finishing touches' to this machine, and improve it.. I intend to add limit switches, spindle control and some other doo-dads to test ideas for a larger machine, so this is by no means the end of the junk-box router build.

Thanks again to all the people on this site for the little bit of inspiration that kicked me into gear after so long.

Cheers, Me.

kiwichris
03-17-2007, 05:00 AM
Hi-ho again.

Well, I'm hooked, so to speak.

I got a bit of spare time between other things this arvo, and took the wheelbarrow DXF that jimini posted to this thead (http://www.cnczone.com/forums/showthread.php?t=19346&page=4) a wee while ago and this is what I did:

pic1: Combined Jimini's three DXF's into one and scaled it to work within my small table and bits of 5mm MDF I have scrap from an old project. I then drew in the cutting paths for a 3mm tool, as that's all I've got at the moment. (Qcad, on Linux)

pic2: Saved the file to my LAN, and converted it to gcode with NCplot on my windows 98 laptop (NCPlot dosn't work properly on my wine setup, must get that going...). I added some tabs to some of the parts manuall in the g-code, and added some preamble bits, and comments so I'd know what the g-code was for in 2 years time when I want another wheelbarrow! I set it up for 2 cuts, one at 3mm deep, the second at 2.5mm giving -5.5 depth.

pic3: Save the gcode back to the lan and wander out to the garage. open it up in EMC2, set the home position....

pic4: And cut. I'm cutting this at 200mm/min (7ipm) as the dremel mount flexes all over the show. I'm going to work on a new mount, and cut it with the router sometime in the next couple of days, it's the biggest 'issue' with my junk box build. It works fine at low speed, but you can see the whole dremel deflecting even at 100mm/minute. At 200 it's all over the show, moves by as much as 1mm...

pic5: Ta-Da. Tabed bits, ready for a quick swipe with 200grit paper...

pic6: The finished article, with old tosh laptop keyboard for scale.

If I had bothered to cut the 'rounds' out of the corners, or thought to make the slots slightly longer this would have been a perfect press-together fit. In the entire time I've been making models (15 years ish, model aircraft mainly) I've never been able to get a press-fit in wooden bits. You can't see it now, but I've got a grin like the Joker that wont go away.

Cheers folks, next installment will be some upgrades to the machine I imagine, to fix some slop.

rippersoft
03-17-2007, 10:01 AM
Very cool! I've been wondering if anyone had EMC running a small dremel mill successfully.

I use NCPlot as well and I would like to run it under Wine. Something to play with later.

RipperSoftware

kiwichris
03-18-2007, 07:52 AM
Very cool! I've been wondering if anyone had EMC running a small dremel mill successfully.

I use NCPlot as well and I would like to run it under Wine. Something to play with later.

RipperSoftware

Hi-ho,

EMC is great once you stop playing with it! The sample stepper config file works just fine, I only had to edit the input_scale and max rate settings to get this working propperly.. When I first played with EMC the default file didn't work so well, so i wasted hours just messing about. :-).

NCPlot under Wine is a bit odd, you get the splash screen, then it closes without an error... I haven't tried the full release though, only RC4, might try it again. Most apps that work OK on 98 run fine in wine, and NCPlot works 100% of my 98 laptop. Odd.

Cheers, Me.

kiwichris
03-18-2007, 08:06 AM
OK, the first thing to go was the motor mount.. It was flexing all over the show, and causing the cutting tool to 'chatter' in MDF a bit more than I liked.

So I measured the dremel, designed a mount in Qcad, converted it to gocde, and cut the large MDF bit in under an hour. And it's a tight slide fit onto the dremel before you do up the pinch bolt. I really, really like this game.

pic1: The new mount compared to the old. You can see the backlash 'notch' on the inside of the dremel mount hole. It's about 0.5mm or so, ish. The bit of alloy will be pinched in the bottom of the dremel, with the treaded on bit. The most painful bit of this process was making the 22mm hole in the alloy. The largest drill I had was 18mm, and my tapered ream only went out to 20mm. Then the only round file I could find was a silly fine one. Much filing ensued. Guess which mount is more rigid?

pic2: Squaring up the new mount. Using the time-honoured 'TLAR' method (That Looks About Right) I lined up the seam inthe plastic case with the square. It's now closer than it was before, and near enough given the tollerances of the rest of this machine! :)

pic3: I didn't cut the second barrow to test the new mount, I cut it because my 3y old daughter told me I had to make her one if I wasn't going to let her play with mine.

Interesting lesson learnt with this. Making the more rigid motor mount caused the chatter to get worse... I backed off the RPM on the dremel and it improved to better than with the old mount... From that I assume that the rigid mount was transfering cutting vibration and loads to the X axis rails more efficiantly and they were 'bouncing'. My theory, I'm sticking to it.

There appears to be a very slight amount of slack in the X axis bushes, which could also be causing the chatter. I'm not too worried, adjusting RPM so that the tool is constantly loaded solves 99% of it. The tool will also run cooler with a higher chip loading, so it's all good. From my not very exhaustive testing I appear to have torque to spare, so making the tool work a little harder shouldn't be an issue.

Picked up some 3mm MDF from the hardware store today to play around with cutting some random objects. Might make one of the dinosaurs from the 'show us your parts' thread..

Cheers, Me.

boblon
03-18-2007, 07:38 PM
Ya know, the more I look at the wheel barrow the more I like it.

I might just have to try one of those as a test piece when I get my build done and IF it works.

Good job.
BobL.

dan dimock
03-18-2007, 08:10 PM
I made three of these wheel barrows, for my wife and daughters a couple of years ago sawing then out on a scroll-saw, and I think I worked on them for three or four weekends to get them done right.
Man I wish I lived close to you, I would get you to build me one and show me how to use it.

Keep up the good work.
Dan

kiwichris
03-18-2007, 11:49 PM
Man I wish I lived close to you, I would get you to build me one and show me how to use it.

A wheel barrow?

You put things in the space on top of it, lift it by the two handles at the opposite end to the wheel, and push it around... Don't need to live close to show you that... :rainfro:

And my teachers said I wasn't helpful!

dan dimock
03-19-2007, 05:34 AM
You made my day (AGAIN) here in the part of Texas where I live we call them the poor man 18 wheeler -I not allowed to use one as I got a DWI driving one before.

Now, back to the machine, I would like one of those also.

Have a great day.

Dan

RCH
03-23-2007, 01:02 PM
Great job Chris.
You have done something that I have been wanting to try. A true, low cost, junk box, small CNC machine. I look forward to you publishing more information about this rig as you continue to modify and improve its function.
You have also got me to finally make my first post after lurking around this forum for a long while.
Once again, great job, and thank you.

boblon
03-24-2007, 09:03 PM
Chris, thanks for the words of encouragement on my build.

I hadn't forgot about the wheelbarrow, just didn't realize it was on your build log that I saw it.

I might just get to try one after all :)

BobL.

kiwichris
03-29-2007, 08:32 AM
A bit of an update on the junk-box router... :)

As per pic 1, I've been cutting lots of different bits, finding what I can and can't do with the machine as it stands before I get keen and 'upgrade' some of it.

Up until this evening I'd cut MDF, hardwood, balsa, and even some CD snowflakes using the gcode from cncdudez site. I still need to play with speeds/feeds to get a better feeling for things, but the flex in the X guides/rods is causing some interesting 'bouncing' effects, so they might be for the chop first, so to speak.

I played around with cutting some platics as well, and some sign laminate I had lying around.

All good learning, but it really didn't answer the big question I had.. Can I cut metal?

Now, we all know that cutting alloy with an MDF mill is not going to be that great, but what about one swinging a dremel tool, with heaps of backlash and springy guides on two of it's axis...

Enter, picture 2. I stuck a bit of T3 in my small drill press vice, clamped to the bed, and wrote some quick'n'dirty circular pocket code to try it out.

Result, picture 3. The pocket is 9mm across, 1.5mm deep, cut at 200mm/min (7ipm) with 0.5mm step over and 0.25 step down (Is that the right term?). Climb milled (counter clockwise path). Used the 3mm up cut 2 flute carbide end mill, going flat out (25000 rpm?).

The finish isn't too bad considering what cut it. There is quite a bit of high frequency tool chatter apparent in the wall of the cut, which is possibly the dremel bearing, or runout in the shaft/collet. The irregularity in the face/bottom of the cut I think is the 'bounce' in my X axis.

The important thing is that I can realistically cut some alloy bearing blocks etc with this cnc machine for making a better/bigger one, once I've made it a little more rigid, and fixed the backlash. It'll take ages with the small cuts, but this is a hobby machine, so I've got all the time in the world. I have access to a friends 'real' mill, but it isn't CNC, so can't do anything outside simple shapes with my limited skill in that area.

I've also done some playing with the backlash settings in EMC2, and found that although it helps quite a bit, I really should fix the backlash, the Y axis (bed travel) is free enough that if you're climb cutting along the X axis it'll pull the bed into the cut, and the backlash compenstation in software is invalidated. I could have guessed this at the start, but it's good to see these things for yourself...

I've decided that EMC2 is a good thing, and can recommend it 100% for a small machine.

I can also recommend getting decent carbide end-mills for a small machine. I've see a few comments about using router bits, ground drills, and other tooling in small mills. I've done at least 3 hours of cutting in MDF and plastic, and now the alloy with the 1/8 flat tool, and it's still got it's edge, and shows no sign of heat damage etc. They are not cheap at $30 NZ each, but worth it I think. It was the most expensive bit I had to buy for the machine as it happens...

Now, onto some other things, I'm going to tidy up some cabling, maybe put some limit switches on, and E-stop, put some lead screw support bearings on, and look into anti-backlash nuts and a beefier pair of X axis rods. I've got some heavier rods in my collection, I'll just have to buy some bushes for them. Might call around tomorrow and see what they're going to cost.

I'll have to draw up my driver circuit for Dan as I promised I would too.. I started the process, but got side-tracked seeing if I could create a chopper controller purely in software in a PIC processor. The short answer to that question was no, but I did enjoy a few interesting hours sniffing solder fumes to get there, and I even vented some magic smoke, so not a total waste of time.

I'll upload a video of the alloy cutting experience to google video and post a link as well, for those who like moving pictures, rather than photos.

Cheers, Me.

kiwichris
03-29-2007, 09:22 AM
The video (http://video.google.com/videoplay?docid=5498363940327273120)

You might want to turn down your speakers before playing it, the combination of vacuum and dremel is a bit much. :)

Although, you can hear the tool chatter if you're paying attention....

RCH
04-01-2007, 09:03 PM
Chris,
Any idea about when you will be posting your controller plans?

kiwichris
04-01-2007, 10:04 PM
Chris,
Any idea about when you will be posting your controller plans?

Hi-ho,

Later on this week if I get some time to spend on it at home. That depends heavily on our son (11 months old) who is teething at the moment, so I'm having some sleep deprivation issues!

Cheers, Me.

kiwichris
04-07-2007, 07:46 AM
Righty, the son almost has 8 teeth now, and I've managed to spend a bit of time creating another controller... This time from bits that are still available! :)

As per the two pictures...

The Controller prototype board, the single processor, 12 IRLZ44 mosfets, 13 resistors, 2 capacitors a red LED for power light and a resonator. I could probably loose the resistors as well, but they are the one safety feature it has, so thought they should stay!!!

The second pic shows the controller driving 3 Nema 17 steppers, running off an old laptop. There are only 2 power resistors in that view as two of the steppers are rated at 5V, and they're running straight from the 5V rail on the 180W ATX power supply. The third stepper is rated at 2.6V so needs some current limiting to ensure it dosn't go up in flames after 5 minutes.

I took a video of it running the steppers, but Google video is not accepting videos at the moment, so you'll have to imagine them turning, and make silly buzzy noises to yourself.

I just need to do some more work on the software for the PIC processor before I post it, mainly so I'm happy sharing it. It's a little messy at the moment, and needs comments so folks who understand PIC assembly language can fiddle with it to their hearts content.

I'll knock up a circuit diagram tomorrow some time, and write some notes on working out the correct values for the power resistors.

Question for the two folks who were interested in this, and anyone else nuts enough to be reading, did you want a PCB layout done, or are you brave strip-board assembly people? I could lay out a PCB in Eagle, which you could then get routed by someone from this board who is local to you... Either that or I could fab the PCB's and send them, but that's getting a bit too organised!

Intrestingly, I see that Hobbycnc used to do a simple L/R stepper controller kitset, but it's no longer available. I would have thought there was a need for a low-end simple controller, obviously not. This type of controller will get you off the ground fairly well, and is far cheaper than other designs to make DIY fashion...

Cheers, Chris H.

dan dimock
04-07-2007, 02:16 PM
Chist;

I fell like this is what a lot of the DIY people like me is looking for.

A easy, cheap way to get up and runing, then if need be - work on a larger unit,

As for me, I like the pin - board that you are using, keep up the good work, you have help a lot of people like me to get started on a unit.

Thanks
Dan

Degrom
04-07-2007, 04:53 PM
Question for the two folks who were interested in this, and anyone else nuts enough to be reading, did you want a PCB layout done, or are you brave strip-board assembly people? I could lay out a PCB in Eagle, which you could then get routed by someone from this board who is local to you... Either that or I could fab the PCB's and send them, but that's getting a bit too organised!


LOL... I think the strip board will be fine. ;)

kiwichris
04-07-2007, 07:55 PM
Good-o,

Strip board it is. :-).

FYI, my 'design target' for this is up to 10khz pulse rate from three motors simultaneiously. That'd give you 1500rpm half stepping (3000rpm full stepping) which equates to 75ipm with a 1/4x20 thread or 1.5m/m with M6/1 thread. (135ipm or 1.5m/m full stepping). That should be more than enough for simple machines, and gives loads of 'head room' for the processor to ensure things don't fall appart. :-).

There are some issues with a single processor monitoring three stepper inputs, so this type of design has an inherent flaw. Not an issue if you don't want to go screaming fast though! If you do want to go screaming fast, all that is required is a single, smaller processor per stepper motor, as per my original junk box design.



Cheers, Me.

dan dimock
04-07-2007, 08:12 PM
Chris;

I fell like this will work for me.

Hope you have a great week end and a happy Easter, and may God bless you, as he has me.

Dan

kiwichris
04-10-2007, 12:08 AM
Over the weekend...

Made up some 'anti-whip' bearings for the end of the lead screws out of LDPE plastic sheet, as per the first pic. The Y lead screw in particular has been going into wild oscilations when jogging... It's just a flat bit of plastic with a 6mm hole in the middle, nothing flash, but does the job.

Also added some more bits to the simple stepper controller. I mentioned before that there was an inherent flaw in using a single processor for three stepper drives. The main issue is that the processor could 'miss' the step input from an axis, while processing the step for another. which would give missed steps without having to over-speed or apply too much resistance to the motor, which is a bit silly. Each 'loop' through the software is only taking 5us, but it's still a problem if your pulse outputs are only 5us across..

I could have fixed it by saying 'you must use 50us pulse width output' but I googled mach3 and found it can only go up to 5us (That might be wrong, I just did a cursory search and found an article stating that) so I figured I had to fix it in hardware, as mach3 is pretty popular.

Anyway, 'cause I've currently got nema 17 steppers running on only 5V on the controller, they don't go fast enough to test up to my 10Khz goal. I wrote a test version of the software that requires 8 step pulses to get 1 on the motor. Now I can drive it at a higher frequency for testing, without having to use a bigger test setup. In real software terms, I'm using a lookup table which is normally 8 steps long, I made it 64 steps long, with the same data repeated 8 times.. Sort of like micro stepping without the steps! :)

Anyway, I wrote a quick program to generate some random gcode for soak testing the code, and left it running with the three steppers running at max 12Khz for just under an hour, and didn't drop a step.

The test code can be seen in the second pic below, and I even took a video, because I've just started using google video, and thought it might be cute.

http://video.google.com/videoplay?docid=-7400150368503616353

Next step is to clean up my bench and bring the junk box cnc machine in with it's PC and controller so I can stick this unit in to test with the larger steppers and power resistors in circuit. I also want to do some testing with emc, and maybe the demo of mach3, as just testing with turbocnc isn't a very good sample set, so to speak.

While the machine is on the bench I'll be putting limit switches on it, and putting an e-stop of some sort together, along with putting a couple of power relays in for spindle and dust collection. Call me lazy, but having to turn on the dremel by hand is sooo time consuming! :).

I'm busy the next couple of nights, but I'll see if I can get the circuit diagram and bill of materials together for people to pick holes in before the end of the week.

Cheers, Me.

kiwichris
04-11-2007, 07:38 AM
One draft schematic.... A meeting I was meant to be at was cancelled. :)


If you go to: http://ohmark.trash.co.nz/3step/ There's one PDF file, and the current eagle schematic file for those with eagle (I'm using the free version).

I've not uploaded it directly to the forums here, as I've seen comments about not being able to delete/edit posts after they get a bit old, and I don't want a draft diagram stuck on here, never to change... If it changes I'd prefer people were getting the newer version automagically.

In the schematic, the spindle relay drive and limit switch circuits are optional if you don't want them, and the jumpers I've shown can just be the pins on the IC connected to gnd or vdd. The F/H jumper (JP1) will set full or half stepping, open/high for half stepping, closed/low for full. You can ignore the 'spare' one, I put that in the schematic for the PCB version, just in case I thought of a need for another jumper!

The input schmit triggers are also optional, but you have to keep the resistor + diode + cap network which I'm using as a simple latch 'pulse stretcher' to get around the limitations of having 3 steppers on one processor. With the schmit triggers in place you need to set your 'step' signals to be 'active low' in your CNC software.

You can just drive the parallel port pins straight into the 1K resistor, although I can't say it will be reliable with all parallel ports or computers, and a 74HC14 should be under a dollar which is cheap reliability insurance.

I've labled this a draft as I've not actually found out how limit switches interact with EMC2 or other cnc software yet, so I might be off on that count, and becuase I've not added the E-stop latch and enable circuit yet which I want for my build.

I'll post the source code when I've done some more testing later in the week, if anyone is really mad keen to get this under way.

I'll also post some notes on power resistors, the power supply hookup, and general building of the thing.

The motors / power supply / power resistors are not shown in this either, but it goes without saying that if you drive your 3A 2V steppers from a 12V power supply without any resistors you're going to create smoke, and if you use a 24V power supply and poke it into the 5V supply for this circuit you'll also create a nice smoky haze in your workshop!!!!

Let me know if this is total gibberish. :)

kiwichris
04-15-2007, 09:27 AM
I've added the limit switch functionality to the controller board for myself, as well as the spindle relay drive circuit and e-stop. I've updated the PDF of the circuit on my website at: http://www.ohmark.co.nz/3step

Also, as per the pic, I've got the driver mounted into the junk-box controller case, and tested.

And a video is worth a thousand (or more) words:

http://video.google.com/videoplay?docid=1186347701240282258

I discovered some interesting things with emc2 (And I assume Mach, as it's based on EMC, although I've not tested with Mach yet).

You can't turn on the spindle output unless you've set a spindle speed, even though your spindle may not be speed controlled. You can waste a bit of time figuring out that...

And when you hit a limit switch, the spindle output dosn't turn off... Which I'll have to look into, as I would have thought the spindle would be turned off for any error condition.

Anyway, I've got it all running, the steppers in the video are doing 700rpm, equivalent to 700mm/min for my junk box router, or 28ipm with 1/4x20 rod for the imperial folks.. That is max speed for those steppers using the resistors I have in the box for the single-stack motors I used on the junk-box router. I suspect I'd get 1000rpm or there abouts with matched resistors rather than the ones for the smaller steppers.

Now that I've got it set up to test at 'high' power levels, I'll try it with TurboCNC and Mach3 to make sure it's all good, and then tidy up the source code, and post it to here. Taken a bit longer to get here as I've picked myself up a head cold, and have had some days sitting inside in the warm playing with the kids.

That, and I've been using whiskey to keep a very sore throat under control on a couple of occasions. Soldering and spirits dosn't sound like a good mix! :cheers:

There is one issue cropped up, in that the ATX power supply will not start if the control PC is on first. This is because the 5V rail is being held up slightly by the parallel port signals feeding through the 'clamp' diodes on the 74HC14's. The ATX power supply is detecting this a a fault condition. I'll add some series input resistors to fix this issue, you may or may not see this depending on your power supply used.

Cheers, Me.

kiwichris
04-15-2007, 09:29 AM
Those steppers are using half-stepping, and at that rate were running at 4.666Khz step pulse rate. Well within the target or 10Khz, and even lower than the test frequency I used for the software of 12khz...

Cheers, Me.

kiwichris
04-19-2007, 11:03 PM
Hi-ho,

An update if you will...

I've finished the software, and made one more subtle change to the hardware. The LED on the e-stop circuit now runs off the PIC processor so I can have it flashing. Makes it a little more cute. :-).

The assembly source code file is now on the web, along with the compiled .hex file if you just want to run the code in without compling using one of the parallel port programmers off the web.

I've not updated the circuit diagram with the LED, and I still need to write a quick bit of Docco about selecting power resistors.

I'll dig out my writing hat and do something over the weekend.

Have a look at let me know if it's clear as mud. :-).

BTW, I used MPLAB 6.3 to write/assemble the code, but it should work fine in the current version (7/5). MPLAB is free from the Arizona Microchip site: www.mirochip.com.

Cheers, Chris H.

kiwichris
04-24-2007, 07:37 AM
As per the pics... The new version of the controller is now full installed and cabled into the case for the router. On the left of the pic you can see the new controller and the resistor board from the originial controller. The original controller board is sitting in the middle of the case..

The small board in the bottom right is the relay board for the spindle control. The second pic shows where the relay is wired to.

The last pic shows the e-stop box which I've stuck to the side of the router chassis. The stop is on top for easy access, and the reset is on the back of the case so it can't be bumped 'on'. The terminal strip is for the limit switch wiring. You can see on the front edge of the router one of the Y limit swtiches. I've wired two of them in series. The 10mm 'top sheet' on the bed touches them about 2mm before the end of travel.

Also a short video, for the video-fiends out there: -video- (http://video.google.com/videoplay?docid=5594809606355326574)

I've now tested the driver with Mach3 as well.. Interestingly enough I get better stepper performance on EMC2 than Mach3. I'm now getting 900mm/min (35.4"/min) on X and Y, and 700 on Z with EMC2, but without lowering the acceleration in Mach3 quite a bit I can only get 600mm/min on X and Y. I suspect this is due to EMC2 having the 'S curve' acceleration, and Mach3 dosn't. Although it's coming according to their website, which would make sense as I understand Mach is based on the EMC code base.

I also put the scope on the step outputs and noticed more irrgularity in the pulse train than I see with TurboCNC or EMC2.. but TurboCNC dons't have continuous profiling, so EMC2 is the winner for me... Of course Your mileage may vary. Use what works for you, and what you find easy to work with...

It was interesting trying Mach3, as a lot of people rave about it on here. It's easy to configure, and seems very powerful, but the bang bang acceleration curve and being based on Windows are detractors for me.. Also the user interface is too busy for my liking, although 1000% better than Mach2, which I also tested out of curiosity...

Anyway, the junk-box router is moving back to the garage so I can cut some bits with it, and start to look at building a better Z carriage, as I mentioned earlier in the thread.

I also broke down and ordered some 'real' backlash nuts from www.dumpsterCNC.com to replace my fast wearing bits of cutting board.

I'll cut a new Z carriage using the router to make it's own parts, installing the real backlash nuts, and upgrading the rails. I have some 15mm ground rod, and I'll pick up some bushes to match... I didn't use the 15mm originally as I had bushes lying around for the 8 and 9mm rod.

That'd remove the slop from the Z and X, and I'll think about what I'll do with Y, I might by some rod for it.. We'll see.

I've just about finished the docco for the controller, I'll stuff something on the web by the end of the week for that if anyone is still following that part of the process...

samco
04-24-2007, 09:09 AM
Very nice work.. :)

Emc2 has a torcher nc program called tort.ngc also. It was used to test the new trajectory planner. I don''t know the size of it off hand but I know it was run on a mini style mill (maxnc)

Good to see it all working so well for you.

sam

kiwichris
04-24-2007, 04:58 PM
Ya, I used a modified version of tort for doing some of my testing with the new controller..

tort has some quite slow feeds in it, so I removed all of the feed blocks and set F2000 at the start so I could get the max. pulse rate I was looking for in testing.

It was interesting seeing how the different controller software coped with it as well Not too suprisingly EMC was the smoothest, even with G61 enabled.. Mach3 might as well have been in 'exact stop' mode with G61, I had to lower the feed rate & acceleration to avoid stals at transitions in the path, and TurboCNC doesn't do CP, so it faired worst in run time, but never stalled as it always runs in 'exact stop' mode due to having no trajectory planner.

I'm not really too sure how many people bother tesitng their machines/controllers in this way, so I didn't go right into the gory details in the thread... Because I wanted to make sure the controller was going to work before I provided software I went to quite a bit of trouble to test it. :)

kiwichris
07-26-2007, 09:48 PM
Hi-ho,

Applogies for not posting my progress on this, things have been a little hectic in our house, and hobbies are the first thing to suffer...

Anyway, progress has been slow on this, but I've been tinkering on and off.

I got some dumpstercnc backlash nuts, and I intend to upgrade the rods on the X and Y as the router 'bounces'...

Havn't said that, I've made quite a few dolls beds and chairs for my daughter, and some other friends, so in the scheme of things the router is a done deal.

The stepper drive design hasn't moved much further either, but I'll tidy up the webpages I started for that and get them online in the next week or so.

Cheers, Chris H.

kiwichris
03-24-2008, 05:51 AM
Wow, over a year since I said I wasn't getting anywhere.....

Anyway, I've still not done much on the junk-box CNC or controller, my real job has changed substantially, and my free time has been almost nil.

On the up side I'm back keen on CNC again, and fired up the junk-box router today to check it was still working so I can make some small robot bits. It's still working as well as it did back in about March last year when it was last turned on!

I have shifted my website, so I think some of the links back in this thread are now broke, but I'll have a look into that in a week or two when I re-shift the website to a new server... The stepper controller design might be of use to someone (or not).

Cheers, Me.

dan dimock
03-24-2008, 10:00 AM
I have been working on my unit for a year and a half and have not got it going. I have the chip and other parts but have not got the board put together.
I got a HobbyCNC board and blew it up the first time I went to hook it up so I have been on hold, as money has been tight.
Dan

mactec54
03-24-2008, 11:58 AM
Hi Kiwichris

Where are you located in NZ is electronics your thing/job

kiwichris
03-24-2008, 03:04 PM
Where are you located in NZ is electronics your thing/job

I'm just south of Christchurch, near Leeston... My job is IT/Computing at the moment.

I used to run a business called 'Ohmark electronics' doing micro development work, and making some niche products for the R/C market.

Hi Dan...

Long time no type, sorry to hear about the hobbycnc board, that'd be a bit of a set-back!

And I think we can all relate to the shortage of cash, that's one of the reasons I've not progressed much with mine, it's on the bottom of the pile priority wise.

kiwichris
04-05-2008, 06:09 PM
FYI, the 3step design is now at http://ohmark.co.nz/3step.html

PIPEORGANPHIL
01-23-2010, 10:06 AM
Hi Chris and all on this thread.
Well it seems like I’m a bit slow on the take up.
Anyway here I go with my junk box unit.

I’ve just set out my version of Chris’s diagram in Circuit Wizard but it seems I have made it look a bit complicated compared to Chris’s as it’s got 4 boards, unforchantly I didn’t see any of this post until today. I came across Chris’s work when searching Google and found his schematics then wend to work on it then found this thread.

Anyway photo 1,2,3 shows the LPT control side of my layout all the same board jus that one is the component side 2 is the component and track side and 3 is just the tracks.

Photo 4 is the stepper motor driver board component and track side there’s 3 of these one for Y,Z,X.
Photo 5 is just the tracks.

I’ve almost finished the mechanics of my CNC and now got all the parts together for the electronics; PIC is programmed just got to etch the Boards now.

Maybe if there still any interest in the thread I my post photo’s of mechanics.

http://i92.photobucket.com/albums/l27/reedorganphil/DSCF0027.jpg

http://i92.photobucket.com/albums/l27/reedorganphil/DSCF0023.jpg

http://i92.photobucket.com/albums/l27/reedorganphil/DSCF0025.jpg

http://i92.photobucket.com/albums/l27/reedorganphil/DSCF0024.jpg

http://i92.photobucket.com/albums/l27/reedorganphil/DSCF0026.jpg

PIPEORGANPHIL
02-09-2010, 04:34 PM
Just to let you all know I’ve just about given up on this project now, has I’ve had no reply from kiwichris regarding information needed before etching. I’ve sent about 10 emails to Chris through various websites and forums with no response.

kiwichris
05-29-2011, 07:14 PM
Hi Phil,

Sorry about the lack of reply, hobbies got shunted well down the list for me over the last couple of years...

I'm getting back into it a bit lately, and I'm looking to build a larger/better CNC as well..

On your PCB's...

Good work laying it out, but if you had have built it I'd have concerns about putting the mosfet's off-board like that. Generally want the circuit/wire lengths for the powersupply caps to mostfet, and drive from the PIC to be as short as possible to avoid any issues with inductance/capacitance or resistive loss in the circuit.

Cheers, Chris H.