PCB milling machines/ accurateCNC

[Lance_RFdude.com]

accuratecnc

I'm again looking at PCB milling machines.... I've used them in the past (t-tech) and really enjoyed having this capability.

Has anyone tried the accuratecnc.com stuff? Specs look great to me.

I'm not really interested in investing the time to build my own.

I thought somebody here might have had experience with these machines since they're pretty popular with RF/microwave folks.

Thanks.

Lance

 

[Azulykit]

pcb milling machine

Here is another good one:

w*w.lpkfusa.com

T-Tech is just fine too.

 

[Mazz]

desktop pcb mill

I've used t-tech (old one) and lpkf (new one).
LPKF is much better than t-tech. LPKF is a german company that makes also laser based machine and, as far as I know, t-tech guys where in lpkf before making their own company.
Mazz

 

[ananth]

2nd hand lpkf

I've used lpkf and didn't have too many issues with it. As long as you understand the setup required to do the processing, it works great..

 

[madengr]

used lpkf

Hey Lance,

We just bought a LPKF to replace our T-tech. It's still in the box so it may be a while until I try it. What I can say is that 80% of (these machines) ease of use is in the software.

It looks like the accuratecnc product is using DeskCNC ( **broken link removed** )This is the same software/harwdare combo that I used to re-build my router at home (I believe I sent you those pictures a while back).

The biggest beef with I have with deskcnc is that there is no rub-out function, just trace isolation. The author is very adament that rub-out is not needed, despite many requests from DeckCNC users, and T-tech and LPKF both have it. On the plus side it uses G-code unlike some proprietary language like the others.

I would suggest downloading the demo here and playing around with it:

https://www.imsrv.com/deskcnc/

Another thing that irks me is that they closed the DeskCNC forum. It's still readable but who knows for how long:

https://www.imsrv.com/cgi-bin/discus/discus.cgi

It's still a good piece of software and the only choice I have found if yoiu want to roll your own system. I believe nothing else is available that imports gerbers.

I noticed on the accuratecnc specs that spindle runout is not listed. I have a Ryobi (Home Depot) router and it is not good for PCB work. It has about +/-4 mils of run-out. It uses 1/4" tooling so I had to buy a collet to get it down to 1/8". I suppose run-out can be compensated for in software just like cutter comp (tool wear) but I have not tried it. That would also not help with small holes. I'm not sure if DeskCNC has cutter compensation. I'd ask for a milled sample from accuratecnc and measure it under a microscope to check the trace widths.

 

[Lance_RFdude.com]

vacuum table pcb

Just as a follow up...

I ended up buying a machine from accurateCNC. the machine is really solid. I've only milled one test board so far.... but there's little to complain about with the machine.

i particularly like the milling depth limiter mechanism... seems like a big improvement over the t-tech machine I used many years ago.

Now I'm trying (with help from some kind folks) to work through some problems with gerbers and how the software (DeskCNC) that runs the machine sees them.... Once I can get this part solved I'll be in better shape...

Lance

 

[E-design]

www.accuratecnc engineering.com

I use a Roland mill which limits my board size to 6X4 but it is good enough for most prototypes. The biggest problem I found with my mill is to get the pcb completely flat and setting the Z0 position before milling. At a $1000 I paid, I can't complain as it includes a 3D scanner to copy other 3D models for milling duplicates.

Lance, what did you pay for your mill new?

 

[Lance_RFdude.com]

make a pc board milling machine

Roland,

The machine was a little over $6k -- introductory price.

The depth limiting for PCB work is done with a depth limiter relative to a ring that contacts the board. The machine applies a few pounds of pressure to the ring, which holds the board flat (if it bows up at all). This is similar to how the t-tech machine did it (with a single foot) but superior -- since it is more force and a ring around the tool rather than one single foot with less pressure.

I really struggled with an investment of this size, but I thought a new machine with specs like this was a safer move than a used machine at similar/more money from a bigger company.

I expect that this company will be very successful as the name and reputation is established -- at least it will put a fair amount of pressure on t-tech and LPKF (not that I have anything against them, I just couldn't afford their machines!).
(only thing I recall being an issue with the t-tech that I used was the foot/height mechanism -- problems like you suggest)

Lance

 

[artem]

build your own pcb milling machine

I use a Roland mill which limits my board size to 6X4 but it is good enough for most prototypes. The biggest problem I found with my mill is to get the pcb completely flat and setting the Z0 position before milling. At a $1000 I paid, I can't complain as it includes a 3D scanner to copy other 3D models for milling duplicates.

Lance, what did you pay for your mill new?

if you have spare Z travel - you can make homemade vacuum table. This easely can be done with with 2 plexiglass sheets where the one (on mount on top) is drilled with multiple holes and another (mounted to bottom) provide provides path for vacuum . Attach vacuum cleaner to side hole and get pcb pressed to botton .
I see this construction on internet but don't remember the address .

 

[E-design]

best tool for removing fr4 dust from air

Last time I looked the Roland mill like mine was about $2500 new. I snatched one off ebay from an Engineering school who upgraded. The $1000 I paid included lots of bits and a PC + monitor in the package. Not a bad deal.

 

[artem]

budget pcb mill

I meant the vacuum table as attachment to your Roland CNC to position pcb flat to the cutting tool.

Of course 1K usd is the price i would pay immediately for a such machine if it will be available. I don't like to pay for PCB workshop every time i need to make test pcb .

 

[E-design]

fr4 silicosis

Artem, the vacuum table is something to look at. Thanks for the suggestion.

 

[artem]

pcb milling table

You are welcome .

And for Z position don't you use ring on milling bits ?
If you have home switch on Roland - that should solve you problem . If not and if it is possible to implement it - there are 1 micron repeatability switches you can mount to get proper vertical alignment.

 

[madengr]

homemade pcb mill

E-design,

How are you generating the toolpaths for your Roland machine?

 

[E-design]

t-tech milling

**broken link removed**

For small boards with no holes I just import a bitmap image of the layout.

 

[rpm73]

g-codes for pcb milling

Hey Lance,

We just bought a LPKF to replace our T-tech. It's still in the box so it may be a while until I try it. What I can say is that 80% of (these machines) ease of use is in the software.

It looks like the accuratecnc product is using DeskCNC ( **broken link removed** )This is the same software/harwdare combo that I used to re-build my router at home (I believe I sent you those pictures a while back).

The biggest beef with I have with deskcnc is that there is no rub-out function, just trace isolation. The author is very adament that rub-out is not needed, despite many requests from DeckCNC users, and T-tech and LPKF both have it. On the plus side it uses G-code unlike some proprietary language like the others.

I would suggest downloading the demo here and playing around with it:

https://www.imsrv.com/deskcnc/

Another thing that irks me is that they closed the DeskCNC forum. It's still readable but who knows for how long:

https://www.imsrv.com/cgi-bin/discus/discus.cgi

It's still a good piece of software and the only choice I have found if yoiu want to roll your own system. I believe nothing else is available that imports gerbers.

I noticed on the accuratecnc specs that spindle runout is not listed. I have a Ryobi (Home Depot) router and it is not good for PCB work. It has about +/-4 mils of run-out. It uses 1/4" tooling so I had to buy a collet to get it down to 1/8". I suppose run-out can be compensated for in software just like cutter comp (tool wear) but I have not tried it. That would also not help with small holes. I'm not sure if DeskCNC has cutter compensation. I'd ask for a milled sample from accuratecnc and measure it under a microscope to check the trace widths.

I have an Accurate 360 machine and I am very happy with the results.

The software & USB CNC controller are PhCNC (https://www.accuratecnc.com/software.html) - not DeskCNC !

 

[biff44]

depth limiter +cnc

Just a warning to anyone doing this stuff at home, especially with the low cost desktop cnc routers. The fiberglass dust produced is pretty toxic stuff. You do not want to breath it. You need to have a REAL hepa vaccum cleaner running to scoop up all the dust or you will have a contaminated lab in no time.

 

[Lance_RFdude.com]

large pcb milling machines

I recently upgraded to an accurate360 as well -- mainly to get video. I got the professional vacuum system as well. I think AccurateCNC is poised to make some real gains in this market over the competition.

Most of my prototyping is done on 9-10mil thick dielectrics, which I wouldn't have really thought usable on a prototyping machine until I tried it.


While I agree that the dust is bad I suggest you do your own research before going too crazy about it. A few of us researched the toxicity of FR4 dust (among other byproducts) and came away not so worried at all. On my accurate350 machine I used a standard consumer vacuum and fine paper filters.

biff44 -- can you cite any references? Even though I have a full vacuum system now I would be interested in reading about the hazards.

We had to justify this for some clowns that we worked for (err... I mean with) who were in the facilities department and intent about keeping us from getting the machine (in the end they won -- one *small* reason why I don't work there anymore).

Lance

Added after 3 minutes:

Hey Lance,

We just bought a LPKF to replace our T-tech. It's still in the box so it may be a while until I try it. What I can say is that 80% of (these machines) ease of use is in the software.

It looks like the accuratecnc product is using DeskCNC ( **broken link removed** )This is the same software/harwdare combo that I used to re-build my router at home (I believe I sent you those pictures a while back).

The biggest beef with I have with deskcnc is that there is no rub-out function, just trace isolation. The author is very adament that rub-out is not needed, despite many requests from DeckCNC users, and T-tech and LPKF both have it. On the plus side it uses G-code unlike some proprietary language like the others.

I would suggest downloading the demo here and playing around with it:

https://www.imsrv.com/deskcnc/

Another thing that irks me is that they closed the DeskCNC forum. It's still readable but who knows for how long:

https://www.imsrv.com/cgi-bin/discus/discus.cgi

It's still a good piece of software and the only choice I have found if yoiu want to roll your own system. I believe nothing else is available that imports gerbers.

I noticed on the accuratecnc specs that spindle runout is not listed. I have a Ryobi (Home Depot) router and it is not good for PCB work. It has about +/-4 mils of run-out. It uses 1/4" tooling so I had to buy a collet to get it down to 1/8". I suppose run-out can be compensated for in software just like cutter comp (tool wear) but I have not tried it. That would also not help with small holes. I'm not sure if DeskCNC has cutter compensation. I'd ask for a milled sample from accuratecnc and measure it under a microscope to check the trace widths.

I have an Accurate 360 machine and I am very happy with the results.

The software & USB CNC controller are PhCNC (https://www.accuratecnc.com/software.html) - not DeskCNC !

Indeed the new machine has a new controller and doesn't use DeskCNC. The older Accurate350 uses a DeskCNC controller, and when I first got the machine it only supported DeskCNC software. A few months later (very early 2008), PhCNC came along and supported the controller. PhCNC is much better than the other software I have used (deskCNC and isopro from t-tech -- but that was several years ago).

drop me an email sometime if you want to swap any info on the machine -- consulting AT rfdude.com.

Lance

 

[biff44]

deskcnc demo

The primary reason I say this is that aTtech machine I once used came with a Nilfisk hepa vaccum and some dire warnings.

I also know from other types of materials handling experience that this IS the sort of dust that can be a real problem. The dust is not just "dust". It is little tiny shards of glass fibers. When you breath these in, some of them stick in the lungs and just never come out again. The dust builds up in your lungs, and eventually you get emphysema. It is the very small size of these dust particles that is the problem. If they were big long strands of glass fiber, they would not be inhaled so easily.

Things like this make me worried:
https://en.wikipedia.org/wiki/Silicosis

If you are on a tight budget, just get a big woodworkers exhaust fan and have it blow everything outside when the machine is running. A wise use of 100 bucks!
Rich

www.MaguffinMicrowave.com

 

[Lance_RFdude.com]

pcb microwave milling

The primary reason I say this is that aTtech machine I once used came with a Nilfisk hepa vaccum and some dire warnings.

I also know from other types of materials handling experience that this IS the sort of dust that can be a real problem. The dust is not just "dust". It is little tiny shards of glass fibers. When you breath these in, some of them stick in the lungs and just never come out again. The dust builds up in your lungs, and eventually you get emphysema. It is the very small size of these dust particles that is the problem. If they were big long strands of glass fiber, they would not be inhaled so easily.

Things like this make me worried:
http://en.wikipedia.org/wiki/Silicosis

If you are on a tight budget, just get a big woodworkers exhaust fan and have it blow everything outside when the machine is running. A wise use of 100 bucks!
Rich

www.MaguffinMicrowave.com

eek, another RPI guy!

I agree-- nobody wants to be breathing in the dust. It seems though that most of the issue is, as your reference says, with irritation and the fact that there is dust, not that the dust itself is a toxin that reacts with the body. I would think that in the case of FR4, the epoxy is actually beneficial.

Either way, at least the AccurateCNC machines have the vacuum sucking the material through a chamber at the site of milling. This is certainly more effective than the older machines I used ~10 years go.


When we did our research we looked at MSDS sheets as well as tried to get info from board houses but came up with little that was terribly worrisome (again, nobody wants to breath it but we sought to quantify the risks and protection required).

it looks like the historical issues with fiberglass insulation were largely dismissed (again, nobody wants to inhale that either).

http://www.lungusa.org/site/pp.asp?c=dvLUK9O0E&b=35439

here are some good common sense references as well
**broken link removed**

**broken link removed**


Aside from the potential health risks I think you always want to be removing the dust for the health of your equipment and the quality of the product as well (the buildup can either damage the equipment or just cause sloppy results).

Thanks for the info

-Lance