Suggestions for in-house PCB/FPC lab?


We are a hardware startup based out of Greece, looking for in-house PCB/FPC fabing solutions/equipment. LPKF's ProtoLaser U3/S look great but they cost $140K and $70K (if I'm not mistaken) respectively is prohibitive. We want to make our own in-house lab with a budget raging from €20 to €35K ($25~$45K).

Do you have any suggestions for: pick-and-place and (especially) laser etching machines fitting for such a budget? It would be very helpful if said machines originate from the EU.

PS: It was suggested to me that we could buy the parts and make our own, which can be the case for the PNP machine but the laser etching/drilling/depaneling machine will probably need to be bought because robustness, support and performance (SMD with 0.2mm pith) are crucial.

Thank you very much.

EDIT: The reason we want such equipment in-house is to considerably shorten the engineering-to-testing circle. In Greece, it takes weeks for us to get the PCBs/FPCs we ordered from an outside vendor, which is the problem we are trying to circumvent by making that as short us possible.

Best Answer

From what you say, the better answer seems to be NOT to set up this capability in-house, but use outside vendors when needed. Let them pay the substantial capital equipment cost, and you just pay the relatively low incremental price they charge. It would take many many PC boards with the machines running many hours/day for this to make economic sense.

There are many places at the other end of the internet that will fab boards quite cheaply and quickly. With shipping, this is still so cheap that doing it yourself makes no sense. Look around and you can surely find some that are nearby or in the EU so that the shipping time is short, if that's a concern. I use a manufacturer in Serbia a lot, and I know they have had my boards made in Macedonia before, but I don't know the company.

Likewise, there will also be manufacturers in your area or only a small shipping charge away. For small volume prototypes, you just need someone with a soldering iron and a hot air station that knows how to use them properly. That equipment is relatively cheap and good to have around for one-off rework, so worst case you can get the equipment and train someone, then have them come in to do the soldering part time as needed.

Your $25k-$45k is much better spent setting up a well stocked lab that enables you to do the parts that you are specialized in that you can't farm out. Get a good stock of parts, scopes, temperature controlled soldering stations, a hot air station, etc. In other words, spend the money building a good design and testing lab, and let someone else spend the money on the parts that are not specific to you, like board fab, pick and place machines, reflow oven, etc.

If you think you need this for instant gratification, think again. First, you don't if you plan properly. Second, doing it yourself won't be all that fast either, and will suck up the time of your engineers that would be better spent doing engineering.

Without doing anything special or expensive, I get bare boards in my hands usually about 1 1/2 weeks after sending the board files to the other end of the internet. I'm in Massachusetts but usually use a place in China for that, about as far away as it gets. Physical distance matters little. After sending out the board files, I work on the BOM, then have someone else purchase parts we don't have in stock and put the kit together. There are some inherent lead times in this, with it taking about 1 1/2 weeks to get the BOM, kit, and other build files together. In other words, the boards usually show up just about when we're ready for them anyway without any extra rush cost. Now you give the kit and build files to someone local with a soldering iron and hot air station, or the part-time person that uses yours.

The 1 1/2 week gap doesn't involve much engineering time, so you get to work on other projects. Usually there will be multiple projects overlapping in different phases, so no time is wasted waiting for final boards to be available. Surely your engineers have something useful to do in the two weeks from sending out board files to getting built boards in their hands.

For higher volume manufacturing once you've got the design done, much longer times will be tolerable. As I said, I use a factory in Serbia often for that, which is actually a lot closer to you than me. Still, the physical distance doesn't really matter. Look around for places that do turnkey manufacturing.