Reply with QuoteW O W!!!!! Looks awesome already!!!!! And doesn't even look like a router yet!
Chich
Alrighty...some pics
The bed is done, 50mm grid spacing drilled and ready to tap, my mate had a little trouble with the rigid tapping (broke one tap) so he started the thread and left it at about 5mm deep and I will get busy with the battery drill tonight and finish it off. (8mm thread)
The whole thing is straight off the VMC so it all needs a little cleaning up. The second lot of parts make up the base for the machine, 5 of in total, they sit on the bottom, and the bed, X axis linear rail base's slot in, the bed then bolts to the base pieces and should all be nice and rigid.
The aluminium is all 16mm thick, as mentioned the bed is 800mm by 400mm. With a bit of luck the bed can be converted to a vacuum table eventually if required, otherwise I will just make a small jig with vacuum for milling PCB's, about the only need I have for a vacuum table (for now anyway)
Once I clean things up I will sit the pieces together and post more pics which will make things a little clearer. It is a cantilever design so it might look a little strange for a while (maybe forever, lol).
I put the spindle in one pic for interest sake.
Cheers.
Russell.
W O W!!!!! Looks awesome already!!!!! And doesn't even look like a router yet!
Chich
Looking great Russell.
16mm plate will make it strong as. It is good to have your parts made on a CNC machine.
Is it finished yet? Sorry couldn't help myself.
cheers,
Rod
Perth, Western Australia
Thanks Chich, hopefully soon it will start looking like oneOriginally Posted by chich2
Cheers Rod, I think (hope) it should be rigid enough and yeah it is nice to watch a big machine cutting out parts for a router !Looking great Russell.
16mm plate will make it strong as. It is good to have your parts made on a CNC machine.
Is it finished yet? Sorry couldn't help myself.
Lol you guys are sounding more and more like my kids...are we there yet ???
The last machine took a couple of years and then some to make so don't hold your breath, though I have learned a thing or two in the process, and I have all the parts now ready to assemble more or less apart from the rest of the plate to be machined so you never know
Oh yeah there is going to be provision for a lathe headstock/4th axis that will literally bolt on to the end of the bed, plans at this stage are for a large servo to drive the spindle, something along the lines of a C40 generator/starter converted to a servo, with a bit of luck and research power should be between 1 and 2 KWatt.
The design would probably be able to use some kind of gang tooling but a quickchange toolpost is probably the most likely option for now. Should make for some interesting video's if it doesn't catch fire.
Enough pie in the sky...better get back to it...
Cheers.
Russell.
Nice in aluminum.
Don't forget to add a temperature sensor, so that Mach3 (scaling) can automatically compensate for the hot weather you get up there.
LM35 is perfect. 10mv/deg C. That will be some accurate machine, as well as pretty.
Hi Neil, nah it isn't hot up here...middle of winter and it is a balmy 25 deg, may go for a swim later
Bedsides, you guys would probably need temp compensation to stop ice buildup
Hope all goes well for you guys at the BBQ, I expect to see lots of pictures !
Cheers.
Russell.
More pics, I cleaned up the pieces and tried them for fit, so far so good, the bed will sit a little lower as I need to chamfer the bed to fit in the rounded part of the base pieces but you get the idea.
Next is to mark and drill the base pieces ready for bolting together, also chamfer the bed then it is onto making the linear rail bases and clamps, you can see the slots on the end of the 5 base pieces, this is where the linear rail bases will slot in, as with everything else they will be made from 16mm plate.
Cheers.
Russell.
Hi Russell,
I couldn't work out how your base supports fitted but it is clear as mud now. Will the area between the table and rails be where the tooling will sit?
cheers,
Rod
Perth, Western Australia
Hi Rod, that is kinda a bit of no man's land, the plan is that by the time the Z axis bolts onto the end of the Y axis, the spindle will sit exactly on the edge of the table, this way the cutter can cover the entire table area.
There were a few unknown's when the design was done, mainly the Z axis "thickness" I guess you could call it, which included the spindle, so it was left on the safe side with enough room to fit everything, if that gap is too large then all that is needed is to make the Y assembly larger to reach the edge of the table, but too small would mean the cutter would hang over the edge and some cutting area would be lost.
As for tooling, if the toolchanger goes ahead there should be about 50mm or so at one end that the spindle can reach past the table (on the x axis travel) so that would be where the rack would go, probably enough for 8 tools maximum but I would think that is more than enough.
Did lots of tapping today, not really any photo's to take as it all looks the same, next week should see some rails getting mounted, all going well.
I have decided to build DSPIC servo drives, and also build some Elm Chan servo drives with identical pinouts/connectors as I wish to compare the two, and then eventually once those have been put through their paces I will have a crack at a MESA card with closed loop to the PC.
Cheers.
Russell.
Hi Russell,
Makes sense what you are saying about the overhang for the edge of the table. Perhaps if you can make it adjustable in case you change spindles or decide to mount a small mill head (X2 ?) on there. I expect that you would have to allow a fair amount of room behind/besdie the machine so that your Y axis has somewhere to go. I do worry a bit about flex on a canterlever design but knowing you it is covered by a real beefy Y axis and post.
It does sound interesting and it is always good so see a different design evolve - I will watch with interest.
You love your servos.
cheers,
Rod
Perth, Western Australia
Hi Rod, yeah there will need to be clearance behind the machine, I would guess about 400mm which is the Y axis travel, not really an issue for me, the advantages of having clear access to the table should be worth it. I guess as my first machine had a fixed gantry I am more annoyed than most (with moving gantry's) when it comes to leaning over the machine to clamp jobs onto the table.
My mate that is helping me with this has built a couple of semi commercial versions of cantilever machines, one had only slightly larger spacing for the X axis rails and had a Y travel of 1200mm and it worked a treat. It does seem a little odd at first but I think it will be a very practical format, here's hoping it works
Yup I am pretty sold on servo's, I found them much simpler than steppers, both in setup and circuit board design but each to their own, half the fun of this CNC stuff is working out what works for the individual person, I for one enjoy seeing how different people tackle the same issues.
Cheers.
Russell.
While the mechanical has gone a bit quiet I thought I would have a crack at the drives.
I have posted a pic of the power stage of my current working router, if there are any electronic guru's out there that have a minute, I am basing my next power stage on this layout more or less.
The question I have is that there is always mention of connecting the power ground and logic ground at one point only, as you can see from my layout I have used a ground plane that sort of covers the entire board and the motor ground connects to this at one end and the logic at the other end, is this bad ?
I think it can't be too bad as my drives are working and haven't blown up yet after much use, but I imagine as I am only using a 30V supply to 24V servo's I am getting away with it, even if the design is poor.
As I am thinking of increasing the voltage on two servo's on this machine to around 50Volts, while not in the high power league it will certainly amplify any faults I have created.
So to make a short story long
I will have more of a ground plane coverage as per every guru's advice I have seen and on the next version I will be using different gate driver chips but the idea is the same. I will also have the processor on the same board and include current sense/blanking, the board in the photo didn't have this as the micro took care of current limit, not using sensing.
Any tips appreciated.
Russell.
You have fluked it.
Think of all tracks as resistors, and the volts that drops across them.
From below the $ in E$7 split the earth track so that a small separate track runs down and then left to pin 3 of E$25 and the capacitors E$40 etc.
The upper earth track on has digital current to the opto, so it works OK because it is an opto, but you should split the track from below $ of E$10 similar to above. This means all your earth references run from the midpoint now.
Basically the rule is consider HIGH CURRENT tracks as resistors, and make sure you connect to the component where you want to reference a voltage from.
If you never form loops on a PCB, (and you have) no currents can be induced into the loops (you don't have), nor are there loops to radiate.
The loop in your situation is probably not an issue, but do note that the current from the negative terminal of E$28 goes the long way to E$2.
Pretty, symmetrical layouts are not the aim. Making the current go where you want it to is what it is all about, even if you must add some links. They are cheap, especially in China.
If it already works, then the worst that will happen is you might send some interference out of your +5/+15 ground wire.
Pity you missed the BBQ. because you might have learned what a mountain goat does.
Super X3. 3600rpm. Sheridan 6"x24" Lathe + more. THREE ways to fix things: The RIGHT way, the OTHER way, and maybe YOUR way, which is possibly a FASTER WRONG WAY!
Hi Neil, thanks for the reply, I see what you mean about the loop around the FET's, I guess I tried to make all the gate driver circuitry as neat and symmetrical as possible as this was recommended on the datasheet, mainly as in keeping the track lengths the same while keeping as close to the FET's as possible, then I added as much groundplane as I could, I didn't think about forming a loop.
I see what you mean about the current path for the E$2, a link straight from where E$39 and E$50 are situated would be a better solution ?
I always try to use as little links as possible, maybe I should re-think this...
In the working drives I actually bypassed the opto, I didn't see the sense of protecting a $1.70 micro with a $3.00 opto, not to mention possibly introducing timing problems. Also the processor board had an opto to protect the upstream BOB/PC from extra smoke.
OK since I am starting afresh on my next board layout, I should keep all grounds linking at one (centralised ?) point ? Also for the groundplane I will try cover as much unused space as possible but not form any loops. High current paths as direct as possible even if it means putting in links.
Mountain Goat ?
Cheers.
Russell.
The opto.
Micros are very noise sensitive, so isolating with an opto, which if you look at Rocby Electronics, is probably a good move.
http://www.rockby.com.au/searchres.c...stock_no=35454
Buy 100 for $36 or 10 for $5. Is that expensive? + << $10 pp. Good bulk purchases here, and just a mailbag softpack away. Good to deal with.
If you don't use an opto, you have to slug the signal with a capacitor to keep out the noise and that will give you more timing problems than the opto.
Remember, the opto can eliminate earth noise issues. Common mode noise, just goes away.
Is this a budget project? High volume project? $$$ ??? If you can fly to a BBQ you can afford (or could) an opto.
Super X3. 3600rpm. Sheridan 6"x24" Lathe + more. THREE ways to fix things: The RIGHT way, the OTHER way, and maybe YOUR way, which is possibly a FASTER WRONG WAY!
Put it this way...did you see me at the last BBQ ? that's because I spent all my money on opto'sIs this a budget project? High volume project? $$$ ??? If you can fly to a BBQ you can afford (or could) an opto.
Seriously though, it is just a project, I am claiming my Elm Chan drives as a success (even if I fluked it
I will probably build a handful of these for my own use and that is about it so I guess a few opto's won't break the bank. If it turns out not to be a dud and not entirely badly designed I will also post the board layout for others that may be interested.
I have opto's (HCPL 2630) on the step/dir input to the drive and PC817 on the enable and fault so the drive IO is opto'd, I had just bypassed the opto on the PWM and NOT PWM signals from the AVR to the H Bridge.
Cheers.
Russell.
Last edited by epineh; 07-14-2009 at 08:27 AM.
I am thinking of building the table frame for this beast out of timber, mainly plywood, as it is just so easy to make and modify as needed. If I want to gloss it up I can always whack a bit of laminate or similar on at the end of the build...
Since the machine will be more or less a desktop machine, even if it is pushing 100KG, the table it sits on is really just that, with an electronics enclosure in one end and cable pathways throughout. I want to take advantage of the cantilever design so front and two side access to the machining area is critical for ease of job loading/setup/removal, at this stage the footprint is looking like about 1200mm by 1200mm, bit larger than I had hoped but there is room in the shed
I was going to make the controls for this thing look like something from the space shuttle, but have calmed down a little and instead will be going with minimalist, the area at the front of the machine under the bed has about a 30mm gap so the plan is to mount LED's with the vital info, power supply status and servo drive status. I have counted six independent supply voltages for this machine so six green led's will give instant indication of their status.
The servo drives will have a couple of indications, enable, fault. I was thinking of having a drive reset button on the front but to be honest if the drive is tripping then something is wrong and needs repair, so the reset can sit in the electronics enclosure, away from the dust.
Other than that a pendant for setup is about all I would like, also the traditional screen/keyboard/mouse is hard to get away from.
I have ordered a compact flash to IDE converter (5 of them actually) and will try use a 16GB CF card as a hard disk, that should prove interesting...
Cheers.
Russell.
Okay scratch most of the last post, I have just aquired an electronics enclosure that will make this thing something else...a little back on the space shuttle concept
What I have got is a control panel from an old machining centre that a friend is retrofitting, it is a free standing kind of thing and is hinged awkwardly in the middle and front, with a little creative thinking I can see a way to use it to make a very simple and professional looking frame/machine base/electronics enclosure/tool rack... that should cover everything I need
I will pick it up tomorrow with a bit of luck so I can get some pics happening then hopefully. Very excited, especially as it still has all the controls for the old VMC, including a green screen and a pop out section for the tape reels (yes paper tapes !) complete with two paper tape rolls, I am not sure what they will make but if EMC will read them...lol.
It will need a coat of paint and some cleaning but should at least look the part once that is all done.
Hopefully the linear rail bars are next to be machined so that will mean LOTS of drilling and tapping for me to do, but the upside is that the rest of the axis's should follow fairly fast from there...woohoo !!! (at least the linear rail parts)
Cheers.
Russell.
Good to hear Russell - a cabinet worthy of holding all that fine work.
cheers,
Rod
Perth, Western Australia
Cheers Rod, I thought you may have liked that idea
I also received my Compact Flash to IDE converters and CF card in the mail during the week, I thought I ordered a 16GB card but it was actually an 8GB card, still heaps for an EMC install, I might try it out on my other machine and see how it goes, supposedly the PC still thinks it is a HDD so is none the wiser.
Cheers.
Russell.
