Homebrew PCBs

> I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....
>
> ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....
>
> So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...
>
> The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....
>
> I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...
>
> Suggestions?
>
> Best, Charlie
>
>

>
> You would just shift to another set of problems.
> Instead of paper and stuff you will think about milling bits, burr,
> milling depth, etc....
> 
> ST
> 
> On Sun, Sep 5, 2010 at 6:40 AM, gandolfreefer
> <synchronousmosfet@...> wrote:
> > I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....
> >
> > ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....
> >
> > So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...
> >
> > The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....
> >
> > I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...
> >
> > Suggestions?
> >
> > Best, Charlie
> >
> >
>

>
> Thanks, Stefan,
> 
> Yes, there are two worries that I mainly have about using a 3D router:
> 
> burring and minimum trace width/clearance.
> 
> I'm pretty sure surface mount would be "out"...but maybe not.
> 
> On the other hand, once it's setup and debugged, it's repeatable, fast, and the automated hole drilling is not to be sneezed at; I find the hole drilling part by far the biggest hassle and would love to have it done automatically....I suppose, if nothing else, I could just have the 3D router just do the drilling, that would be a huge time and hassle saver....
> 
> However, if the hole diameters were standardized to one size (in some circuits this is doable, even if it isn't optimal), one bit could do it all: routing and drilling...all in one multi-stepped set of passes. That would be pretty cool. It would have to be a carbide or even diamond bit, at a very high speed, maybe even with water cooling, but I'm sure it's achievable.
> 
> And, now that I think about it more, actually larger holes would, of course, be achievable: just program it in to to mill the hole out larger. It might be useful to make the table 4D, with rotary capability as the 4th dimension.
> 
> I was hoping someone on this list had already mastered the craft of routing boards instead of etching them. It might have to be me...
> 
> Best, Charlie
> 
> 
> --- In Homebrew_PCBs@yahoogroups.com, Stefan Trethan <stefan_trethan@> wrote:
> >
> > You would just shift to another set of problems.
> > Instead of paper and stuff you will think about milling bits, burr,
> > milling depth, etc....
> > 
> > ST
> > 
> > On Sun, Sep 5, 2010 at 6:40 AM, gandolfreefer
> > <synchronousmosfet@> wrote:
> > > I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....
> > >
> > > ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....
> > >
> > > So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...
> > >
> > > The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....
> > >
> > > I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...
> > >
> > > Suggestions?
> > >
> > > Best, Charlie
> > >
> > >
> >
>

> Thanks, Stefan,
>
> Yes, there are two worries that I mainly have about using a 3D router:
>
> burring and minimum trace width/clearance.
>
> I'm pretty sure surface mount would be "out"...but maybe not.
>
> On the other hand, once it's setup and debugged, it's repeatable, fast, and the automated hole drilling is not to be sneezed at; I find the hole drilling part by far the biggest hassle and would love to have it done automatically....I suppose, if nothing else, I could just have the 3D router just do the drilling, that would be a huge time and hassle saver....
>
> However, if the hole diameters were standardized to one size (in some circuits this is doable, even if it isn't optimal), one bit could do it all: routing and drilling...all in one multi-stepped set of passes. That would be pretty cool. It would have to be a carbide or even diamond bit, at a very high speed, maybe even with water cooling, but I'm sure it's achievable.
>
> And, now that I think about it more, actually larger holes would, of course, be achievable: just program it in to to mill the hole out larger. It might be useful to make the table 4D, with rotary capability as the 4th dimension.
>
> I was hoping someone on this list had already mastered the craft of routing boards instead of etching them. It might have to be me...
>
> Best, Charlie
>

> One stumbling block is the software to generate tool paths.
> It can be 
> quite expensive. The free offerings are generally rather
> quirky. I quite 
> like Visolate <http://sourceforge.net/projects/visolate/> though it
> 
> takes some trial and error to get it working correctly.
> Instead of 
> slavishly following the track outlines it calculates the
> minimum cut 
> length to achieve the desired connectivity. The resultant
> tracks look 
> odd but it works well. Cutting time and cutter wear is
> considerably less 
> than most other methods. There are some pics here 
> <http://www.mit.edu/~vona/Visolate/Visolate-info.html>.
> Visolate won't 
> do the drilling.

On 07/09/2010 03:39, Herbert E. Plett wrote:
> VISOLATE?
> went to the pages, saw the pictures and find it more than interesting.
> now, downloaded the file but there is no documentation whatsoever and I have no clue as how to start it.
> any help, PLEASE..
>
>
>
>    
>> One stumbling block is the software to generate tool paths.
>> It can be
>> quite expensive. The free offerings are generally rather
>> quirky. I quite
>> like Visolate<http://sourceforge.net/projects/visolate/>  though it
>>
>> takes some trial and error to get it working correctly.
>> Instead of
>> slavishly following the track outlines it calculates the
>> minimum cut
>> length to achieve the desired connectivity. The resultant
>> tracks look
>> odd but it works well. Cutting time and cutter wear is
>> considerably less
>> than most other methods. There are some pics here
>> <http://www.mit.edu/~vona/Visolate/Visolate-info.html>.
>> Visolate won't
>> do the drilling.
>>      
>
>
>
>
>
> ------------------------------------
>
> Be sure to visit the group home and check for new Links, Files, and Photos:
> http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>
>

>
> Hi Charlie,
> > Yes, there are two worries that I mainly have about using a 3D router:
> >
> > burring and minimum trace width/clearance.
> >    
> 
> With the right equipment you can get quite fine traces. I have seen 
> boards made this way with two tracks fitting between 0.1" pitch pads. To 
> achieve that sort of resolution you need a very accurate spindle as even 
> 0.005" out of center can make a big difference. If your spindle is not 
> perfectly concentric the cut width will increase and you often get 
> problems with burring. If you get burring, stop cutting and rotate the 
> cutter in it's collet by 45 degrees and try again. Repeat until you find 
> the 'sweet spot'.  A worn cutter will of course also cause burring.
> 
> > I'm pretty sure surface mount would be "out"...but maybe not.
> >    
> 
> Fine pitch such as QFP packages is out but other surface mount is fine.
> 
> > On the other hand, once it's setup and debugged, it's repeatable, fast
> 
> Not as fast as you would think. Routing is slow but you can just leave 
> the machine to get on with it. It is also generally quite noisy due to 
> the high spindle speeds (15,000RPM to 100,000RPM, depending on how deep 
> your pockets are).
> 
> > , and the automated hole drilling is not to be sneezed at; I find the hole drilling part by far the biggest hassle and would love to have it done automatically....I suppose
> >    
> 
> Yes, drilling is far faster and more accurate than hand drilling. Before 
> I went over to routing boards I used to photo etch then CNC drill. 
> Aligning the board was always a PITA.
> 
> > However, if the hole diameters were standardized to one size (in some circuits this is doable, even if it isn't optimal), one bit could do it all: routing and drilling...all in one multi-stepped set of passes. That would be pretty cool. It would have to be a carbide or even diamond bit, at a very high speed, maybe even with water cooling, but I'm sure it's achievable.
> >
> > And, now that I think about it more, actually larger holes would, of course, be achievable: just program it in to to mill the hole out larger.
> 
> Don't go there. Thin routers break VERY easily. In my experience 
> anything below 1.5mm (1/16") is asking for trouble. Cut the tracks with 
> a V cutter like these 
> <http://cgi.ebay.co.uk/5-x-45-Carbide-PCB-Engraving-Bits-CNC-Router-Tool-T14-/250522253434> 
> then drill with normal carbide drills. You can get plastic collars that 
> fit on drill shanks to set the depth. Once you have the machine and 
> collars set up you can change cutters quite quickly because they always 
> fit in the collet at a fixed depth.
> 
> You don't need any form of coolant. It just makes a mess. Use decent 
> vacuum extraction to keep the abrasive dust out of your machine and out 
> of your lungs. As far as hold-down is concerned, vacuum works quite well 
> if the vacuum is strong enough. A vacuum cleaner is definitely NOT 
> strong enough. When drilling and routing parts out you need to cut all 
> the way through plus a bit so you need a backing sheet of some sort. For 
> vacuum this sheet needs to be porous. 6mm (1/4") MDF works and is pretty 
> cheap. I just use double sided adhesive tape to hold the boards down. It 
> is quick and easy. You need to experiment with the tape. White 'carpet' 
> tape sticks well but is a pain to remove and tends to clog cutters. 
> Transparent tapes generally work quite well but can be difficult to find.
> 
> For double sided boards, drill a couple of locating holes in your blank 
> board and use them to locate the board on pegs on your table. These 
> precisely locate the board when you flip it over.
> 
> One stumbling block is the software to generate tool paths. It can be 
> quite expensive. The free offerings are generally rather quirky. I quite 
> like Visolate <http://sourceforge.net/projects/visolate/> though it 
> takes some trial and error to get it working correctly. Instead of 
> slavishly following the track outlines it calculates the minimum cut 
> length to achieve the desired connectivity. The resultant tracks look 
> odd but it works well. Cutting time and cutter wear is considerably less 
> than most other methods. There are some pics here 
> <http://www.mit.edu/~vona/Visolate/Visolate-info.html>. Visolate won't 
> do the drilling.
> 
> I use SheetCam <www.sheetcam.com> for drilling. Actually what I do is 
> generate the tracks in Visolate them import the tracks and drill files 
> into SheetCam. I can then lay them out as I want them before cutting.
> 
> Les
>

>
> Hello, 
> Elektor Magazine (now available in the USA) has a kit form PCB milling-drilling unit at about 750Euros. Optional extras include front panel milling ect at an extra cost. According to the spec sheet, 0.2mm tracks with 0.2mm spacing is possible as well as multiple size hole size drilling. The schematics, mech plans, firm & software are available at a small fee from the mag. They will also supply the ready made PCS,s and pre-programmed Micro-controllers, again at a small fee.
> check out the site at http://www.elektor.com for previews, reviews and US costs. chester
> 
> 
> 
> 
>       
> 
> [Non-text portions of this message have been removed]
>

>
> Hi Charlie,
> > I can save a serious chunk of dough making the 3D ... maybe even 4D rig myself, with surplus steppers and my own circuit boards from scratch...
> >    
> Yes, it isn't difficult. Preferably you want to use ball screws for the 
> drive. Acme screws will work but any backlash will cause problems. HTD 
> toothed belts can also work quite well. Don't try allthread rod. It 
> isn't worth the hassle.
> 
> Don't use rails that are supported only at the ends. There are a lot of 
> machines out there that have round rails that are only supported at the 
> ends. In the middle of their travel they flex and bend all over the 
> place. It is amazing how much even a thick rail will flex. Linear rails 
> like these 
> <http://cgi.ebay.co.uk/Bosch-Rexroth-linear-bearing-70-mm-rail-R044281301-/170525543558> 
> are the best but they are pretty expensive. You can achieve pretty good 
> precision using ground flatstock (tooling plate) and skateboard 
> bearings. The biggest problem with that setup is keeping the dirt out.
> 
> I would suggest that you buy the stepper drives rather than trying to 
> make your own. Microstepping drives are much better than full or half 
> step. If you want maximum dependability and reliability, look at servos. 
> They are a bit more expensive but they are worth every penny. IM Service 
> do some nice motors and drives for reasonable money 
> <http://www.imsrv.com/>. If you do go for steppers, get the modern 
> square type, not the old fashioned round type. The square ones work much 
> better.
> 
> > and spend the money where it really matters: the spindle itself. Something ridiculously fast with close to zilch run-out.
> >    
> 
> There is an interesting discussion of cheap Chinese spindles on the Mach 
> Yahoo group at the moment. 
> <http://groups.yahoo.com/group/mach1mach2cnc/>. For your application you 
> don't need more than about 150W but a bigger one, say up to 800W would 
> be fine.
> 
> > Thanks for all the links and tips; this is soooo the way to go, I'm never looking back at etching again...just gonna use the protoboards that match up to breadboards until I get the PCB milling rig done, and then convert straight to milling. (routing).
> >    
> 
> Routing isn't a magic bullet. It solves some problems but it has it's 
> own steep learning curve and problems. However a CNC router is a very 
> handy tool to have anyway. If you are into model making etc you will 
> soon wonder how you ever did without one.
> 
> > My God, I can't even imagine .005 run-out! They actually make spindles that bad?
> >    
> 
> Oops, not enough zeros. I meant 0.0005" :-[
> 
> > Just to theorize a little....it did just recently occur to me that this could also be done by using, instead of a rotating spindle with a V bit and drills, a very fine, very high speed/pressure jet of abrasive...either air based or water based. I know such things are made....I remember reading of one that had a diamond head, because the jet of water that came out was under such high pressure and at such an insane speed it cut through inch steel plate like it was butter...soft butter.
> >
> > Now, THAT would be a fast method. In fact, the table would have to move FAST to keep the jet from just boring right through....heh heh heh...but it could be done, no doubt about it. I'll have to google if there are any micro-jet rigs/heads for such a thing. And incredibly fine abrasive.
> >    
> 
> Waterjet is quite impressive. I have seen it cut through 2" steel plate 
> using garnet abrasive. Cadbury's Crunchies (A sweet made with foamed 
> sugar covered with chocolate) are cut out using an oil jet.
> There are a few problems with jet cutting. The first is the pressure 
> needed. Waterjet uses BIG pumps (15hp - 200hp), running insane pressures 
> (40,000 - 90,000PSI). The other difficulty is in maintaining the correct 
> cut depth. Copper is harder to cut than fibreglass so the feed rate 
> would be ultra critical. You will have major problems on corners as the 
> machine slows down to take the corner. Also as the nozzle wears you 
> would need to tweak the feed rate to compensate.
> 
> Les
>

>
> Hi Charlie,
> > I can save a serious chunk of dough making the 3D ... maybe even 4D rig myself, with surplus steppers and my own circuit boards from scratch...
> >    
> Yes, it isn't difficult. Preferably you want to use ball screws for the 
> drive. Acme screws will work but any backlash will cause problems. HTD 
> toothed belts can also work quite well. Don't try allthread rod. It 
> isn't worth the hassle.
> 
> Don't use rails that are supported only at the ends. There are a lot of 
> machines out there that have round rails that are only supported at the 
> ends. In the middle of their travel they flex and bend all over the 
> place. It is amazing how much even a thick rail will flex. Linear rails 
> like these 
> <http://cgi.ebay.co.uk/Bosch-Rexroth-linear-bearing-70-mm-rail-R044281301-/170525543558> 
> are the best but they are pretty expensive. You can achieve pretty good 
> precision using ground flatstock (tooling plate) and skateboard 
> bearings. The biggest problem with that setup is keeping the dirt out.
> 
> I would suggest that you buy the stepper drives rather than trying to 
> make your own. Microstepping drives are much better than full or half 
> step. If you want maximum dependability and reliability, look at servos. 
> They are a bit more expensive but they are worth every penny. IM Service 
> do some nice motors and drives for reasonable money 
> <http://www.imsrv.com/>. If you do go for steppers, get the modern 
> square type, not the old fashioned round type. The square ones work much 
> better.
> 
> > and spend the money where it really matters: the spindle itself. Something ridiculously fast with close to zilch run-out.
> >    
> 
> There is an interesting discussion of cheap Chinese spindles on the Mach 
> Yahoo group at the moment. 
> <http://groups.yahoo.com/group/mach1mach2cnc/>. For your application you 
> don't need more than about 150W but a bigger one, say up to 800W would 
> be fine.
> 
> > Thanks for all the links and tips; this is soooo the way to go, I'm never looking back at etching again...just gonna use the protoboards that match up to breadboards until I get the PCB milling rig done, and then convert straight to milling. (routing).
> >    
> 
> Routing isn't a magic bullet. It solves some problems but it has it's 
> own steep learning curve and problems. However a CNC router is a very 
> handy tool to have anyway. If you are into model making etc you will 
> soon wonder how you ever did without one.
> 
> > My God, I can't even imagine .005 run-out! They actually make spindles that bad?
> >    
> 
> Oops, not enough zeros. I meant 0.0005" :-[
> 
> > Just to theorize a little....it did just recently occur to me that this could also be done by using, instead of a rotating spindle with a V bit and drills, a very fine, very high speed/pressure jet of abrasive...either air based or water based. I know such things are made....I remember reading of one that had a diamond head, because the jet of water that came out was under such high pressure and at such an insane speed it cut through inch steel plate like it was butter...soft butter.
> >
> > Now, THAT would be a fast method. In fact, the table would have to move FAST to keep the jet from just boring right through....heh heh heh...but it could be done, no doubt about it. I'll have to google if there are any micro-jet rigs/heads for such a thing. And incredibly fine abrasive.
> >    
> 
> Waterjet is quite impressive. I have seen it cut through 2" steel plate 
> using garnet abrasive. Cadbury's Crunchies (A sweet made with foamed 
> sugar covered with chocolate) are cut out using an oil jet.
> There are a few problems with jet cutting. The first is the pressure 
> needed. Waterjet uses BIG pumps (15hp - 200hp), running insane pressures 
> (40,000 - 90,000PSI). The other difficulty is in maintaining the correct 
> cut depth. Copper is harder to cut than fibreglass so the feed rate 
> would be ultra critical. You will have major problems on corners as the 
> machine slows down to take the corner. Also as the nozzle wears you 
> would need to tweak the feed rate to compensate.
> 
> Les
>

>
> I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....
> 
> ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....
> 
> So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...
> 
> The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....
> 
> I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...
> 
> Suggestions?
> 
> Best, Charlie
>

From: mat_henshall <mat@...>
Subject: [Homebrew_PCBs] Re: thinking of routing....
To: Homebrew_PCBs@yahoogroups.com
Date: Wednesday, September 8, 2010, 4:54 PM







 



  


    
      
      
      

Charlie,



I recently purchased a 3axis router kit from www.zentoolkits.com. A good price and straightforward to set up.



I can recommend it for both routing small parts out of wood and plastic, as well as for pcb tracks and drilling holes.



I used a small finish router for wood and plastic, and spindle motor zenntoolworks sells fro pcbs.



Mat

--- In Homebrew_PCBs@yahoogroups.com, "gandolfreefer" <synchronousmosfet@...> wrote:

>

> I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....

> 

> ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....

> 

> So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...

> 

> The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....

> 

> I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...

> 

> Suggestions?

> 

> Best, Charlie

>





    
     

    
    


 



  






      

[Non-text portions of this message have been removed]

>
> 
> Charlie,
> 
> I recently purchased a 3axis router kit from www.zentoolkits.com. A good price and straightforward to set up.
> 
> I can recommend it for both routing small parts out of wood and plastic, as well as for pcb tracks and drilling holes.
> 
> I used a small finish router for wood and plastic, and spindle motor zenntoolworks sells fro pcbs.
> 
> Mat
> --- In Homebrew_PCBs@yahoogroups.com, "gandolfreefer" <synchronousmosfet@> wrote:
> >
> > I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....
> > 
> > ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....
> > 
> > So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...
> > 
> > The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....
> > 
> > I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...
> > 
> > Suggestions?
> > 
> > Best, Charlie
> >
>

>
> Hi Charlie,
> 
> > take some thought as to how to do it cheap but good. If I can get my hands on a Bridgeport...or there are machine shops here in Manilla, I'm hoping to find out they're ridiculously cheap...a decent aluminum extrusion can be carefully milled to achieve a decent "straightness", something like that could be a great cost saver.
> >    
> 
> By the time you get it machined, extrusion is likely to be more 
> expensive than ground flat stock and less accurate. It is usually 
> available in standard 12" and 18" lengths. When I was looking into 
> designing a small router GFS was the cheapest solution I could find that 
> was reasonably accurate. In the end I converted a CNC Bridgeport. More 
> expensive and it takes up a lot more room but you can also do a lot more 
> with it.
> 
> > I don't know what a ball drive is...but I was thinking a good toothed belt should work fine...I'll look up ball drives at McMaster-Carr's.
> >    
> 
> Try this 
> <http://cgi.ebay.co.uk/NSK-C1-ground-ballscrew-12-4mm-L469mm-BALL-SCREW-/140448845277>. 
> Ball screws are very low friction and fairly accurate. Lower precision 
> rolled screws aren't too expensive. If you want precision preloaded 
> screws, your wallet is going to take a hammering!
> 
> > Les, I can achieve .001" (linear table motion) steps fairly easily with steppers.
> >    
> 
> 0.001" is fine for PCB work, as long as that is 0.001" per step, not 
> 0.001" per microstep. Under load you can only gurarantee a stepper will 
> be within +/- 1 full step of the commanded position, no matter how many 
> microsteps you use. Microsteps create smooth motion and reduce the 
> chance of resonance induced lost steps. They don't increase accuracy.
> 
> > Depending upon the toothed drive belt, I would assume the precision would be at least as good as +/- 0.0001"
> >    
> 
> No chance I am afraid. For that sort of precision you are looking at 
> precision preloaded ball screws ($$$) and careful frame design. If you 
> are lucky repeatability will be within about 0.002". Absolute accuracy 
> could be as bad as 0.01" or worse over the whole travel. That sounds bad 
> but it should be adequate for PCB work.
> 
> > PS: wowzer, that linear rail link from ebay you gave is a knockout!!! And only 10 pounds...that's a deal, in my book. :)
> >    
> 
> Yeah but it is very short. Decent length rails cost a lot of money.
> 
> Les
>

On 09/09/10 09:47, gandolfreefer wrote:
> Les, thanks, I see your point about flat ground stock.
>
> I do remember being amazed at how inexpensive flat ground stock really is when I used to order it for various milling tasks, especially compared to trying to make some oneself!!
>
> What would you recommend for bearings to run on the ground stock? Will regular plain old sealed ball bearings from McMaster-Carr do?
>
> yeah, unfortunately, the only ball-screw drives at McMaster-Carr are horribly expensive, seriously expensive, that's what got me retreating to toothed belts. I don't suppose....steel chain will work? It does wear over time and "stretch", but it seems to me this is a pretty light duty application for chain, and perhaps if the chain was a bit oversized for the app, it would wear very well?
>
> Best, Charlie
>

>
> Hi Mat,
> 
> WOWZER, their 7 x 7 kit is so cheap it's gotta be the most screamin' deal on the planet!!!!
> 
> I have only one cavil: it is very frustrating for them not to list any $*#@!!! specs!! Like, what the heck is the step size for each axis? How fine a resolution is it capable of?
> 
> And.........while I'm at it, is the driver a microstepper, or what? Do they get whatever resolution they get by gearing down the steppers, or by using microstepping drive? And, if it's microstepping, how many microsteps per steps of the stepper motor, and how many steps does the stepper motor have to start with???
> 
> Sheez.
> 
> It looks great and the cost is almost unbelievable, but even at $500+ I'm not plunking my hard earned cash down for it without some stinkin' specs!
> 
> But thanks for the tip, this could well be the best answer. I just don't know 'till I see the specs.
> 
> Best, Charlie
> 
> --- In Homebrew_PCBs@yahoogroups.com, "mat_henshall" <mat@> wrote:
> >
> > 
> > Charlie,
> > 
> > I recently purchased a 3axis router kit from www.zentoolkits.com. A good price and straightforward to set up.
> > 
> > I can recommend it for both routing small parts out of wood and plastic, as well as for pcb tracks and drilling holes.
> > 
> > I used a small finish router for wood and plastic, and spindle motor zenntoolworks sells fro pcbs.
> > 
> > Mat
> > --- In Homebrew_PCBs@yahoogroups.com, "gandolfreefer" <synchronousmosfet@> wrote:
> > >
> > > I really appreciate all the answers; very helpful, and I'm still rereading them all and ruminating on it all....
> > > 
> > > ....part of figuring out what method I'm going to commit to is I'm now in the philipines, doing medical mission work and R & D, here for quite a while, and I have to consider local supplies, and it turns out getting all my PCB fab stuff sent to me is not as simple as I thought it would be.....
> > > 
> > > So, I'm now considering building a 3D router table/stepper-controlled, and just route the PCB's....which would also include being able to drill the holes automatically...
> > > 
> > > The nice thing about this, is that even though it's expensive to setup, once it's done, no worries about type of paper, type of printer, etchant, etc., etc....
> > > 
> > > I just wouldn't be able to do very fine stuff, I presume...but I wasn't planning on making boards for surface mount IC's anyway...
> > > 
> > > Suggestions?
> > > 
> > > Best, Charlie
> > >
> >
>

> 
> WOWZER, their 7 x 7 kit is so cheap it's gotta be the most
> screamin' deal on the planet!!!!
> 
> I have only one cavil: it is very frustrating for them not
> to list any $*#@!!! specs!! Like, what the heck is the step
> size for each axis? How fine a resolution is it capable of?
>

>
> > Hi Charlie,
> > 
> > I would use skateboard bearings. They are available in ABEC7 accuracy
> > for very little money. Skate bearings are standard 608 size metric
> > bearings, 8mm bore 22mm OD 7mm thick.
> 
> 
> ABEC7 (in a skate shop) is a marketing gimmick.  For skating you want 'crap'
> bearings as free running is more important than run-out.  That rating is for
> high-speed spindles, hence the price.  ABEC1 or 'precision' bearing are good
> enough.  
> 
> The rating has nothing to do with quality either.  Consider sealed (rubber)
> instead of shielded (metal), they keep the dust out better.
> 
> See www.vxb.com, they'll sell you 100 sealed bearings for $35.  (Skates need
> 16 bearings, boards 8, hence the odd pack sizes.)
> 
> Tony
>

>
> look at
> http://www.probotix.com/FireBall_v90_cnc_router_kit/
> 
> 
> --- On Thu, 9/9/10, gandolfreefer <synchronousmosfet@...> wrote:
> > 
> > WOWZER, their 7 x 7 kit is so cheap it's gotta be the most
> > screamin' deal on the planet!!!!
> > 
> > I have only one cavil: it is very frustrating for them not
> > to list any $*#@!!! specs!! Like, what the heck is the step
> > size for each axis? How fine a resolution is it capable of?
> >
>

On 11/09/10 02:37, Herbert E. Plett wrote:
> look at
> http://www.probotix.com/FireBall_v90_cnc_router_kit/
>
>
> --- On Thu, 9/9/10, gandolfreefer<synchronousmosfet@...>  wrote:
>    
>> WOWZER, their 7 x 7 kit is so cheap it's gotta be the most
>> screamin' deal on the planet!!!!
>>
>> I have only one cavil: it is very frustrating for them not
>> to list any $*#@!!! specs!! Like, what the heck is the step
>> size for each axis? How fine a resolution is it capable of?
>>
>>      
>
>
>
>
>
> ------------------------------------
>
> Be sure to visit the group home and check for new Links, Files, and Photos:
> http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>
>

> Tony, outstanding link.
> 
> Thanks. The same bearings from McMaster Carr are twenty times the
> money. :-0
> 
> Turns out the specs on the ABEC bearings has only to do with OD and ID,
> etc., nothing on runout I can find. Not that it matters much, really;
> I'll use bearings that are rigid and bearings that are spring loaded to
> oppose, and flat stock, and side bearings to maintain alignment, and it
> should work fine; the only thing I have to be careful to avoid is super
> fast bearings with "loose" balls, but that shouldn't be much of an
> issue.
> 
> Best, Charlie
> 
> --- In Homebrew_PCBs@yahoogroups.com, "Tony Smith" <ajsmith@...> wrote:
> >
> > > Hi Charlie,
> > >
> > > I would use skateboard bearings. They are available in ABEC7
> accuracy
> > > for very little money. Skate bearings are standard 608 size metric
> > > bearings, 8mm bore 22mm OD 7mm thick.
> >
> >
> > ABEC7 (in a skate shop) is a marketing gimmick.  For skating you want
> 'crap'
> > bearings as free running is more important than run-out.  That rating
> is for
> > high-speed spindles, hence the price.  ABEC1 or 'precision' bearing
> are good
> > enough.
> >
> > The rating has nothing to do with quality either.  Consider sealed
> (rubber)
> > instead of shielded (metal), they keep the dust out better.
> >
> > See www.vxb.com, they'll sell you 100 sealed bearings for $35.
> (Skates need
> > 16 bearings, boards 8, hence the odd pack sizes.)

>
> By the time you add all of the options to get a running machine it is 
> nearer $1000. Also it is uses those horrible cheap unsupported rails 
> with Oilite bearings. In the middle of it's travel that thing is going 
> to flex and bend all over the place.
> 
> Les
> 
> On 11/09/10 02:37, Herbert E. Plett wrote:
> > look at
> > http://www.probotix.com/FireBall_v90_cnc_router_kit/
> >
> >
> > --- On Thu, 9/9/10, gandolfreefer<synchronousmosfet@...>  wrote:
> >    
> >> WOWZER, their 7 x 7 kit is so cheap it's gotta be the most
> >> screamin' deal on the planet!!!!
> >>
> >> I have only one cavil: it is very frustrating for them not
> >> to list any $*#@!!! specs!! Like, what the heck is the step
> >> size for each axis? How fine a resolution is it capable of?
> >>
> >>      
> >
> >
> >
> >
> >
> > ------------------------------------
> >
> > Be sure to visit the group home and check for new Links, Files, and Photos:
> > http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
> >
> >
> >
> >
>

>
> I've just checked out the site and all the spec are listed under the "specifications" tab ??
> If you scroll down below the picture there are various tab options, the X-Y spec are in inches as well as the resolution in various stepping modes.
> 
> 
> 
>       
> 
> [Non-text portions of this message have been removed]
>

> guaranteed, but only if you put a 3HP router on the Z axis and try to cut through steel.
> if you use a reasonable router and don't work metal, the flex is unnoticeable...
>
>

On 12/09/10 09:25, gandolfreefer wrote:
> 3. scale....while I fully understand that a bigger machine allows me to play around with various modeling, etc., and have...a...goood...time, anything larger than what I precisely need is a waste of money and a compromise in performance. I know precisely the largest PCB's I have any intention of ever making, at my level of design and production, so I will scale the machine to precisely that purpose, and I will get outstanding results for small money.
>

> You'll be surprised how little force is needed to give a noticeable
> amount of deflection. If you ever get hold of one of these machines,
> move it to the middle of it's travel and place a pointer near the
> carriage. Even pressing gently with your finger will make it move
> visibly.
> 
> Les
> 
> > guaranteed, but only if you put a 3HP router on the Z axis and try to
> cut through steel.
> > if you use a reasonable router and don't work metal, the flex is
> unnoticeable...

> Don't go too small. Most of the cost
> goes in the bearings, motors, 
> electronics, spindle etc. Increasing the length of the
> rails and 
> screws/belt doesn't add much to the overall cost.

On 15/09/2010 02:33, Herbert E. Plett wrote:
> Not at all true.
> If you lengthen a rail you will also have to step up the size ->  more expensive!
>