Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
01/24/07 16:04
Read: times


 
Msg Score: +2
 +2 Informative
#131405 - PCB Prototyping
Responding to: ???'s previous message
Hi Mike,

I think I understand the heart of your question now. You want to know how small you should be able to go without destroying your work. The answer depends very much on the process you are using to make the PWB.

PWB stands for Printed Wiring Board. The bare board, with just traces and pads and holes, is called a PWB. Once you populate the board with components it becomes a Printed Circuit Board (PCB). Just a side note that you might be interested to know.

If you are using computer generated transparencies to expose photosensitive etch resist with UV light as the process for producing the trace pattern on the board, followed by acid etching of the exposed copper, then you can (with practice) get down to at least 0.010" traces and 0.010" spaces. I know because I have done this, and was able to do so consistantly. If you're using an etch resist pen to draw the trace pattern onto the board by hand, I don't know how small you can get, but it won't be near that small. If you're using dry transfer patterns to rub the trace pattern onto the board, then you can get as small as the dry transfer trace patterns.

The issue that seems to concern you more is the physical drilling of the board. If you are using a power drill to bore the holes by hand, you aren't going to be able to get very small. You will need to buy a small Dremmel style tool with a miniature 3-jaw chuck and a drill press stand. You will then be able to use very fine "wire size" drill bits small enough to drill your holes. And yes, you will easily be able to drill the header row you're describing. Do a search for "wire size bits" and I'm certain you'll be able to find plenty. You might even be able to buy them at the local hardware store. If so, you may have to ask them for bit sizes 61 through 80 for them to know what you're talking about. They will also probably be able to order you small quantities of only the sizes you need, instead of you having to buy a set of wire size bits.

Here's the rub. If you use normal wire-size bits, the board's fiberglass will dull them very quickly. Once they start to dull they overheat and that heat will melt the adhesive between the copper and fiberglass resulting in pads being lifted off of the board. So, you will need to make certain that you change out the bit as soon as you feel it getting dull (harder to drill through the board).

The alternative is to buy special carbide PWB drill bits. These are bits ground down from a standard size 1/8" shank, and they stay sharp for a very long time. The problem is that the carbide is extremely brittle. They will last for hundreds and hundreds, maybe thousands, of holes without dulling. That is, of course, if you don't break them first. While they can withstand huge forces along their vertical axes, they can withstand approximately zero force laterally. If you wiggle the board, even the tiniest little bit, while drilling, the bit will snap without hesitation. Most people break bits by getting in too much hurry to move the board to the next hole, just a fraction of a second before the bit clears the hole on the upstroke. The point is this. If you practice (yes, you will break a bunch of them first) you can get the hang of using these bits. You have to hold the board in position fast, without any wiggle. And you have to consciously not move the board to the next hole until the drill is completely retracted. With practice you can get proficient, but even then I've never known anyone who could get the full lifetime out of one of these bits while drilling by hand. In fact, I've only known of a couple of people who could make them last longer than the cheaper wire-size bits and then nowhere near enough to justify the added expense. Given the expense of these carbide bits, even with practice, it's still much cheaper to use the standard wire-size bits. Just keep plenty of them on hand and don't be shy about changing them when they start to dull.

The last tip I will give you about drilling is the pilot hole. Any good software package will print the pads with a small opening in the center. This allows the acid to etch a small divot, a pilot hole, in the copper of the pad. Without this pilot hole the drill bit will wander off-center when you try to drill. If your art work doesn't include these pilot holes in the pad centers, find out what you need to do to include them and do so. They are not optional. If you have to, put small divots in the pad centers with a small punch or brad or something before you try to drill.

Do these things and you should be able to achieve the pitch you're looking for with no trouble.

Now before I quit writing and get back to my own work, I want to give you a bit more help, specifically on how you can make 2-sided boards. If you are using photosensitive etch resist and transparencies, you can include fiducials on the transparencies that mark two opposite board corners. This will provide the alignment you need. The only other thing you need is to provide continuity from one side to the other (the hole plating). An easy way to do this is to thread a piece of bare 28AWG wire (wire wrap wire) through the hole and solder it on both sides. If the hole must also accomodate a component lead you will need to ensure there is sufficient clearance. With DIP packages this isn't a problem since they have flat leads. You just need to make sure and solder the wire to the side of the lead.

Good luck,
Joe


List of 70 messages in thread
TopicAuthorDate
BUS            01/01/70 00:00      
   does not make sense, pls explain            01/01/70 00:00      
      100 mils width is too small            01/01/70 00:00      
         Use 8052.com            01/01/70 00:00      
         in one direction, in the other it is .100            01/01/70 00:00      
         Drilling holes with 100 mil pitch should not be...            01/01/70 00:00      
            None at all!            01/01/70 00:00      
            The largest drill diameter is 0.050"            01/01/70 00:00      
               It's all coming back...            01/01/70 00:00      
                  What do you mean "coming back?" It's still here!            01/01/70 00:00      
                     the rest of us regard this a trip down memory lane            01/01/70 00:00      
                        Yes, paper clips, zippers ... they're still around            01/01/70 00:00      
                           yup, slow things down            01/01/70 00:00      
               It is 0.80mm NOT 0.080"            01/01/70 00:00      
                  One's too small, the other's too large            01/01/70 00:00      
         It\'s really not too small ...            01/01/70 00:00      
         Terms; Width v Pitch            01/01/70 00:00      
   if it is all or none            01/01/70 00:00      
      I'm using a Single Sided board...            01/01/70 00:00      
         I don't understand...            01/01/70 00:00      
            yes            01/01/70 00:00      
               If you mean what I think you mean.            01/01/70 00:00      
               use an IC socket then...            01/01/70 00:00      
                  a good ides does not suffer from being repeated            01/01/70 00:00      
                     I missed that, sorry...            01/01/70 00:00      
                        what is a "component platform"?            01/01/70 00:00      
                           thanks            01/01/70 00:00      
                              here we go            01/01/70 00:00      
                                 oh now I remeber            01/01/70 00:00      
                                    they are available in 0.3" as well            01/01/70 00:00      
         nowadays, ridiculously small is REALLY small            01/01/70 00:00      
            Broken bit.            01/01/70 00:00      
   PCB Prototyping            01/01/70 00:00      
      Drilling...            01/01/70 00:00      
         great, Rob            01/01/70 00:00      
            1 sided...            01/01/70 00:00      
               you are aware, I hope            01/01/70 00:00      
                  SiLabs, etc...            01/01/70 00:00      
                     not at all            01/01/70 00:00      
                        So is the 89C51ED2            01/01/70 00:00      
                           there are PLCC sockets that spread the pins to 0.1            01/01/70 00:00      
                              Yep.            01/01/70 00:00      
                                 ok            01/01/70 00:00      
                                 you don't have a proper devboard            01/01/70 00:00      
                                    Devboard            01/01/70 00:00      
                                       do yourself a favor            01/01/70 00:00      
                                          Thanks            01/01/70 00:00      
                              There are adapter boards that make it easy            01/01/70 00:00      
                                 You can add a groundplane!            01/01/70 00:00      
                                    not just Kai            01/01/70 00:00      
                     SiLabs vs Atmel            01/01/70 00:00      
                        elaborating            01/01/70 00:00      
                           deviates            01/01/70 00:00      
                              the problem there is            01/01/70 00:00      
                                 diffs twixt SiLabs families            01/01/70 00:00      
                                    To Andy            01/01/70 00:00      
                                       LPC9xx???            01/01/70 00:00      
                                 RCAP2 vs TMR2RL            01/01/70 00:00      
                                    acronyms...            01/01/70 00:00      
                                    If that had been it            01/01/70 00:00      
                                       F120 not genuine either            01/01/70 00:00      
                                          the difference is 'addition' vs 'change'            01/01/70 00:00      
            SMT Soldering is MUCH easier than PWB drilling.            01/01/70 00:00      
               wrong address            01/01/70 00:00      
               surface mount thru-hole            01/01/70 00:00      
                  Original SMT parts were modified DIP            01/01/70 00:00      
                     True but...            01/01/70 00:00      
                        The objective wasn't to make them smaller.            01/01/70 00:00      
                           O-Who?            01/01/70 00:00      
            Just to follow up...            01/01/70 00:00      

Back to Subject List