MOSFET rDS PCB

This one came out surprisingly well, apart from the total faceplant with that resistor. With any luck, it’ll measure MOSFET on-state drain resistance over temperature for an upcoming Circuit Cellar column; it’s a honkin’ big Arduino shield, of course.

Drilled holes on the Sherline using the relocated tool height switch:

rDS Tester - drilled PCB

rDS Tester - drilled PCB

Front copper, after etching & silver plating:

rDS Tester - etched front

rDS Tester - etched front

Back copper, ditto:

rDS Tester - etched rear

rDS Tester - etched rear

I think I can epoxy the resistor kinda-sorta in the right spot without having to drill through the PCB into the traces. Maybe nobody will notice?

The traces came out fairly well, although I had to do both the top and bottom toner transfer step twice to get good adhesion. Sometimes it works, sometimes it doesn’t, and I can’t pin down any meaningful differences in the process.

And it really does have four distinct ground planes. The upper right carries 8 A PWM Peltier current, the lower right has 3 A drain current, the rectangle in the middle is the analog op-amp circuitry tied to the Analog common, and surrounding that is the usual Arduino bouncy digital ground stuff. The fact that Analog common merges with digital ground on the Arduino PCB is just the way it is…

, ,

  1. #1 by Peter on 2012-02-23 - 08:04

    Hi Ed. If you’re using your Sherline to drill the holes, is there any reason why you don’t just use it to mill the tracks as well?

    • #2 by Ed on 2012-02-23 - 09:00

      use it to mill the tracks

      Mostly because I don’t want to fill the Basement Laboratory and my lungs with a haze of glass fiber dust. Other than blowing the exhaust outdoors (which just moves the problem elsewhere), there seems to be no good way to collect the fine dust resulting from that operation.

      In addition, I’ve never been confident that I could level and align a large PCB to the required tolerances. The first side is no problem, the bottom side seems to pose a challenge. That’s particularly true for the teeny end mills required to cut around those teeny SMD pads; snap off a few of those puppies and milling starts looking pretty expensive.

      So I think drilling the holes, then using two ferric chloride etches, makes for an overall win. Opinions certainly differ on that, but I do take advantage of the local household hazmat collection days to get rid of spent etchant.

      • #3 by smellsofbikes on 2012-02-28 - 12:22

        I was reading the blog of the guy responsible for pycam, when he was on about milling pcb’s, and he mentioned doing so in a water bath only a few mm deeper than the copper, to catch the debris. I might try this.

        • #4 by Ed on 2012-02-28 - 15:36

          a water bath

          Turning the dust into mud makes plenty of sense!

          Sealing the bottom would be tricky; the Sherline doesn’t have much room for clamping around the edges, so you’d (well, I’d) be forced to drill through the bottom of the tray. Maybe clay or some such would seal well enough around the bolts?

          It’s not as if the XY motion is fast enough to slosh the water out…

  2. #5 by Keith Neufeld on 2012-02-23 - 08:10

    I love the look of the board vignetted by unprocessed copper, and I love the look of your multiple ground planes.

    Educate me — apart from providing some thermal relief for soldering (in addition to thermal relief on individual pads), are there other technical advantages to a gridded ground plane, or is it an aesthetic choice?

    • #6 by Ed on 2012-02-23 - 09:16

      an aesthetic choice?

      Well, I do write the Above the Ground Plane column in Circuit Cellar, so it’s definitely got some of that aesthetic thing going on… which, to be honest, is about as much aesthetic as I can handle. [grin]

      The real reasons:

      • Thermal isolation for the bazillion Z-wires stitching the top & bottom ground planes. I can’t make plated-through holes, so I get lots of practice soldering teeny little wire snippets.
      • Alignment targets across the board. I drill the holes first, then line up the toner transfer paper on a light table; it’s really helpful to have plenty of dots.
      • Etch depth indication. When all those little squares come clean, I know the traces will be isolated across the board.

      The Z-wires use square vias that exactly fill one grid square, which means they’ll survive the inevitable misalignments…

  3. #7 by Peter on 2012-02-23 - 09:37

    Why did you silverplate it?

    • #8 by Ed on 2012-02-23 - 09:44

      I think it’s more solderable than bare copper, at least after a few weeks, although I don’t have much hard data to back that up.

      Plus, there’s that aesthetic thing…

      • #9 by Peter on 2012-02-23 - 12:56

        Ah, yes, it looks nice

        Was just wondering, because silverplating doesn’t even makes sense at 23 and 13 cm, at least *not* the kind of silverplating people can do at home – the silver layer is much thinner than the skin depth. Same goes with those shiny gold SMA connectors – the gold looks nice, but far too thin to have any effect on the lower microwave bands.

        So, was wondering why you silverplated it for near-DC signals; was curious if there was any profound reason I didn’t think of :-)

        It indeed makes for easy soldering (though I spray it with some solder flux; same effect, and even easier soldering :-) ) but unless you use solder with a few % of silver in it, the PbSn solder will quickly leech away the silver plating near the joint, leaving you with a ring of bare copper print around the joint – but doesn’t matter much, apart from aesthetics, as the joint has been already made by then…..

        • #10 by Ed on 2012-02-23 - 15:18

          the silver layer is much thinner than the skin depth

          Except in high-end audio gear, where anything is possible…

          unless you use solder with a few % of silver in it

          I’d been using 4% silver, but recently gave up and went back to ordinary 60-40 lead solder. I should try the silver solder again with the Hakko iron, as the failing Weller might have been contributing to the soldering problems.

          as the joint has been already made by then…..

          And the pictures have been taken!

      • #11 by smellsofbikes on 2012-02-23 - 13:02

        It is *way* more solderable, although on our copper boards at work (and I do probably 5 a week) we use a 3M scrubby pad on the copper (lightly, so as to not rip up traces) and that does a great job for refurbishing copper for soldering.

        I’m finding the same sorts of problems you’re talking about wrt getting the board flat and level enough for milling. I’m using a shopvac with a cyclone separator and 2micron filter to catch fiberglass debris, and that’s working well, but getting a flat board (and spinning the bit fast enough to get a good cut surface) is time-consuming.

        • #12 by Ed on 2012-02-23 - 15:23

          spinning the bit fast enough

          Ah, yes, I’d forgotten about that. I drill the board with the Sherline’s 10k rpm head, which is too slow by at least a factor of two. Fine-pitch milling would be even worse. I just feed the drill too slowly into the holes, suck up the swarf with the shopvac, and it all works out fine for my simple needs.

          I have a dead shopvac that should become a cyclone separator, but that’s far down on the to-do list.

          The EMC2/LinuxCNC mailing list routinely features long discussions about which affordable 60k rpm spindles have the fewest warts…