Thinking of constructing the Zoltar CNC. PCB fabrication.

[beastmaster]

I'm in need of a desktop CNC for my electronics projects. Specifically, I need something to mill PCBs. After weighing my options, it seems like my best bet is to build from pre-existing designs. I came across the "Classic Zoltar CNC", and it looks to be exactly what I need. But, before I dive into buying stepper motors, control boards, hardware and such, I have a few questions for the people that have gone through the construction process.

Approximately how much is the laser/water jet cutting going to cost for these parts? Obviously, the price is going to be dependent on the shop, but I'm curious as to what other's have paid for this work.

Are there any pitfalls I should be aware of when making the Zoltar? What was your biggest stumbling block?

How well does the Zoltar perform when making PCBs? Is there anyone out there who can tell me of their experiences with making PCBs with this machine?

I'd also be interested in hearing about user experiences with other CNC setups.

[signIT]

Hi,

It is a machine that can make parts from almost any material but it needs some modifications to be OK.
Go to cnc home web site and inform yourself there you can find the best tips and hints for this machine, buy at least 2A current steppers not to heavy as this is a small and a bit floppy machine so it must have a very stable base frame to stand on.
Dont change the MDF router base as it very tricky to change, the MDF is cheap and firm just lay another MDF above as waste board.

This machine can be trimmed to be as good as any hobby router for small money.

It is very nice to use and can make anything as a big expensive router can do as long you use soft metal as brass and aluminium.

Warning! be very careful when buying parts as it is very easy to buy parts that you can't use.

Have a nice weekend!

Signit

[ozan]

In Turkey, i paid around USD 200 for lazer company, of course they supplied me 3 mm ST-52 too, i know stainless stell is better but the cost of it exactly doubles the price even a bit more...

[signIT]

The Zoltar Classic router has stainless steel only for the linear sliding parts for x,y axiz.

I recommend that you use stainless only where you really need it, of course it is better to use stainless steel overall but it is overkill.

If you afford linear bearings for all axiz go for it and stay with ACME screws as a cheap good solution on all axiz you can change it later if you want.

If you use THK type of linear bearing you have to modify the parts to make place for those bearings

Signit

[sansbury]

I built a PCB mill for the same reason. I'll share a couple of thoughts with you.

1. A milled PCB is a lot less bother once you're all set up than doing it with an etching tank and drill press. No competition at all.

2. A home-made PCB is not as nice to work with as a professionally-fabbed one. There's no soldermask, and the holes aren't plated through, so sometimes you need to do some tricky double-sided soldering. This is a PITA with DIP sockets, terminal blocks, etc. You also have to make your own vias. None of these things are the least bit unmanageable, but depending on the type of work you do, they can add up to a lot of work.

3. The cheapest you have any chance of getting away with is ~$500. With services like Batch PCB, the cost for pro-fab boards is coming down quite a bit--for $500 you could probably buy at least 10-15 one-off custom boards, and if you are doing any kind of volume, that comes way down.

Now, since none of those were sufficient to stop me, I'll tell you how I built mine

I bought the "Brute" plans from John K leinbauer, and built it 99% stock. There is some real genius in this machine in that it requires almost no precision parts fabrication to get something that works. This was my first build, and it worked well pretty much the first time I turned it on. I can reliably get .008" accuracy on it, which is good enough to run signals between pins on a 0.1" pitch grid.

I'm working on a larger machine now for heavier work running a laminate router, and trying to decide whether to rebuild it more precisely, as some people have gotten as good as .003", which brings a lot of SMT devices within usable reach, if you have good hands and eyes. I got build-itis towards the end and hacked some of it together, but it turned out very well despite me. If I do rebuild, I will use the larger machine to cut the parts. However, there's really nothing in the plans I'd change, and I highly recommend them for other beginners. The advantage for a first-timer is that John tells you "do it this way" anywhere it matters. $36 is cheap insurance against wasted time and money.

I now run it with a HobbyCNC board with 24V feeding 2A 125oz motors, and today I ran it at 48ipm, which is plenty considering it has a 6" X-Y area. This is large enough for most DIY electronics projects. I used to run it with a simple R/L controller and a 12V PC PSU, which gave me 14ipm. Fast moves are fun to watch, but with a Dremel, you probably won't be cutting faster than 10-12ipm most of the time, so don't get obsessed with rapids. I use the free version of Eagle which has a 2x4" or so board size limit and the PCB-GCODE plugin to generate milling and drilling programs, and I run the machine with EMC2.

The thing I had the most trouble with was getting my depth to be really, really consistent. In order to cut narrow traces well, you need to control the depth very well. Surfacing your table and mounting the copper board carefully (I use carpet tape) are essential. I found that by watching the cutter as it first started cutting, I could pause, tweak the depth, and continue, and get very good results.

I am not sorry that I built the Brute because I learned a lot and had fun, but I've found that having my own board-manufacturing capability is not as useful as I thought, while having a CNC router is, so I'm upgrading.