Loading ...
Sorry, an error occurred while loading the content.

Building up the IOB6120; component question

Expand Messages
  • dfnr2
    Hello all, While my wife is having her girls night out for her friend s birthday, I decided to finally build up my IOB6120 which I received along with parts
    Message 1 of 5 , Oct 2, 2011
    • 0 Attachment
      Hello all,

      While my wife is having her girls' night out for her friend's birthday, I decided to finally build up my IOB6120 which I received along with parts from Rob Ribbeck quite some time ago.

      My general technique is to dump all the parts on the bench where I can see them all at once, then grab them according to their geometry rather than part number. This usually results in the most efficient build. Well, of course I got to the end and I couldn't find a couple of parts. I'm certain I'll find them next week, or some time when I'm organizing my bench. In the meanwhile, I looked in the BOM in the manual, and also in the iob9c schematic, and can't find the reference for RN2, which is used for the VGA interface. The resistor values aren't marked on the schematic, either. Can somebody fill me in on this?

      Otherwise, judging from the size and the circuit, it looks like something like the CTS 744C083271JP should fit (8 pin, 4 x 270ohm resistor, 0.126" x 0.2"). Since it's VGA being driven from the FPGA, I'm guessing 270ohm resistors will work fine. If I were the impatient type (I am), and if I weren't missing the 2.5V LDO as well, I'd just solder in 3 1206 resistors on their sides, but since I'm throwing the parts onto a mouser order anyway, I may as well order the right part.

      It might be nice to update the BOMs with this info as well.

      By the way, I'd like to shout out a big "THANK YOU" to Vincent Slyngstad and Bob Ribbeck, who put a lot of effort into making these boards and painstakingly assemble the parts kits, so all I have to do is the fun part. THANKS!

      Best regards,

      Dave

      If I thought my soldering skills were up to it, I'd stick a
    • Vincent Slyngstad
      ... Cool! ... RN2 is a set of 4 330 ohm resistors. These were R11, R12, and R13 in the original IOB6120. (The fourth resistor is not used.) ...
      Message 2 of 5 , Oct 2, 2011
      • 0 Attachment
        > While my wife is having her girls' night out for her friend's birthday, I
        > decided to finally build up my IOB6120 which I received along with parts from
        > Rob Ribbeck quite some time ago.

        Cool!

        > My general technique is to dump all the parts on the bench where I can see
        > them all at once, then grab them according to their geometry rather than part
        > number. This usually results in the most efficient build. Well, of course I
        > got to the end and I couldn't find a couple of parts. I'm certain I'll find
        > them next week, or some time when I'm organizing my bench. In the meanwhile,
        > I looked in the BOM in the manual, and also in the iob9c schematic, and can't
        > find the reference for RN2, which is used for the VGA interface. The resistor
        > values aren't marked on the schematic, either. Can somebody fill me in on
        > this?

        RN2 is a set of 4 330 ohm resistors. These were R11, R12, and R13 in the
        original IOB6120. (The fourth resistor is not used.)

        > Otherwise, judging from the size and the circuit, it looks like something like
        > the CTS 744C083271JP should fit (8 pin, 4 x 270ohm resistor, 0.126" x 0.2").
        > Since it's VGA being driven from the FPGA, I'm guessing 270ohm resistors will
        > work fine. If I were the impatient type (I am), and if I weren't missing the
        > 2.5V LDO as well, I'd just solder in 3 1206 resistors on their sides, but
        > since I'm throwing the parts onto a mouser order anyway, I may as well order
        > the right part.

        744C083331JPCT-ND is the parts number I've used with success.

        > It might be nice to update the BOMs with this info as well.

        Not sure where you got your BOM, but if it is the PDF at
        http://www.jkearney.com/sbc6120/iob6120.pdf, that does an excellent
        job of documenting the original boards, but doesn't say anything about
        the revisions since then.

        > By the way, I'd like to shout out a big "THANK YOU" to Vincent Slyngstad and
        > Bob Ribbeck, who put a lot of effort into making these boards and
        > painstakingly assemble the parts kits, so all I have to do is the fun part.
        > THANKS!

        You are welcome! Let us know how it turns out. I found the flash
        chips challenging to solder correctly, and to a lesser extent the FPGA
        chips. The rest of the stuff mostly went right together. Debug was
        an interesting learning experience :-).

        Vince
      • dfnr2
        ... Super. So it looks like my guess was not so bad. ... Yes, that s the BOM I was looking at. The document is excellent; I just didn t know that R11-13
        Message 3 of 5 , Oct 2, 2011
        • 0 Attachment
          > 744C083331JPCT-ND is the parts number I've used with success.
          >
          Super. So it looks like my guess was not so bad.


          >
          > Not sure where you got your BOM, but if it is the PDF at
          > http://www.jkearney.com/sbc6120/iob6120.pdf, that does an excellent
          > job of documenting the original boards, but doesn't say anything about
          > the revisions since then.

          Yes, that's the BOM I was looking at. The document is excellent; I just didn't know that R11-13 corresponded to the network.


          >
          > > By the way, I'd like to shout out a big "THANK YOU" to Vincent Slyngstad and
          > > Bob Ribbeck, who put a lot of effort into making these boards and
          > > painstakingly assemble the parts kits, so all I have to do is the fun part.
          > > THANKS!
          >
          > You are welcome! Let us know how it turns out. I found the flash
          > chips challenging to solder correctly, and to a lesser extent the FPGA
          > chips. The rest of the stuff mostly went right together. Debug was
          > an interesting learning experience :-).
          >

          I will definitely report back. I can't try anything without the 2.5V LDO, so it will be later this week.

          Regarding the soldering; to be honest, at the time I ordered the board, I was still intimidated by SMT work; I had always had my prototypes built up by the same local assembly house that does my production jobs. I had fiddled with hot air, and with fine soldering tips and wire, and no-clean flux. Then, over a few visits to the assmbly facility, I took note of what they were doing: No hot air (except to remove chips), no IR, no ovens (except their big reflow machine.) These guys were using regular irons with regular chisel tips, basically dipping the parts in water-soluble flux and still populating the board with beautiful joints faster than I could do through-hole. I copied what I saw.

          Now, I am using my same old Metcal iron, with a pretty big low-temp hoof tip that can hold a bunch of solder. I just squirt on a bunch of flux along the leads using a squeeze bottle with syringe-needle tip, load up the hoof with a bit of solder, hold the chip in place by hand, tack down leads at either end, then drag the solder along the line. The shorts will drag right along to the end, then I use a solder wick to clean up the last two pins. Makes a perfect joint. The full-spaced parts are so easy to solder, I can do them faster and better quality than through-hole.

          For the discretes, I used the same small chisel tip as for through-hole. I placed all the resistors of a given value on the board next to their pads; fluxed up the pads, applied a tiny bit of solder to one side of each footprint with the iron tip; then go through and align each resistor, heat up the solder, and place the resistor. Finally, turn the board around and solder the other side of all discretes by keeping the tip loaded with a bit of solder and touching it to each unoldered joint.

          Afterwards, I wash the whole board in warm running tap water, and scrub really well with a toothbrush until the foaming stops. The boards look really professional.

          Actually, although I hesitated forever before jumping on the bandwagon, I find SMT much faster and cleaner than through hole: No missed joints; no bending leads, taping the part in place; no clipping leads, no tedious pin-by-pin soldering, and fewer touch-ups. The only new tools required are a $1.50 pair of sharp-tip tweezers from the dollar store, a head-band magnifier, and (I consider optional) a $15 hoof tip for my soldering iron.

          I guess that was pretty long-winded, but despite having designed several SMT-based boards, and having done some SMT rework, this was the first decent-sized SMT project with fine-pitch parts I actually built up by myself. I was surprised at how easy it was (with a lot of flux). It helped that the board was well laid out with lots of space around most of the parts.

          Best regards,

          Dave
        • Bob Armstrong
          ... the end . It s called drag soldering or drag and wipe . It s a bit scary, but it s super fast if you have the right technique. Practice first on a few
          Message 4 of 5 , Oct 2, 2011
          • 0 Attachment
            >then drag the solder along the line. The shorts will drag right along to
            the end .

            It's called "drag soldering" or "drag and wipe". It's a bit scary, but
            it's super fast if you have the right technique. Practice first on a few
            parts and boards that you don't care about :-)


            http://www.howardelectronics.com/jbc/dragsoldering.html
            http://www.diyforums.org/GrubDAC/GrubDACsolder.php

            Bob
          • Vincent Slyngstad
            ... You were definitely on the right track. ... I lack a really good way to create PDF documents. The best I can do is to compose something in Word or
            Message 5 of 5 , Oct 2, 2011
            • 0 Attachment
              >> 744C083331JPCT-ND is the parts number I've used with success.
              >>
              > Super. So it looks like my guess was not so bad.

              You were definitely on the right track.

              >> Not sure where you got your BOM, but if it is the PDF at
              >> http://www.jkearney.com/sbc6120/iob6120.pdf, that does an excellent
              >> job of documenting the original boards, but doesn't say anything about
              >> the revisions since then.
              >
              > Yes, that's the BOM I was looking at. The document is excellent; I just
              > didn't
              > know that R11-13 corresponded to the network.

              I lack a really good way to create PDF documents. The best I can do is to
              compose something in Word or whatever, then print to PDF. And I haven't
              any way to edit an existing PDF.

              > Regarding the soldering; to be honest, at the time I ordered the board, I was
              > still intimidated by SMT work; I had always had my prototypes built up by the
              > same local assembly house that does my production jobs. I had fiddled with
              > hot air, and with fine soldering tips and wire, and no-clean flux. Then, over
              > a few visits to the assembly facility, I took note of what they were doing: No
              > hot air (except to remove chips), no IR, no ovens (except their big reflow
              > machine.) These guys were using regular irons with regular chisel tips,
              > basically dipping the parts in water-soluble flux and still populating the
              > board with beautiful joints faster than I could do through-hole. I copied
              > what I saw.

              I'm using the hot air and solder paste, and no-clean flux with fine solder
              for touch-up.

              > Now, I am using my same old Metcal iron, with a pretty big low-temp hoof tip
              > that can hold a bunch of solder. I just squirt on a bunch of flux along the
              > leads using a squeeze bottle with syringe-needle tip, load up the hoof with a
              > bit of solder, hold the chip in place by hand, tack down leads at either end,
              > then drag the solder along the line. The shorts will drag right along to the
              > end, then I use a solder wick to clean up the last two pins. Makes a perfect
              > joint. The full-spaced parts are so easy to solder, I can do them faster and
              > better quality than through-hole.

              I did some of that, but had some problems with it. One was nearly invisible
              whiskers of solder, where the pads hadn't quite electrically separated. But
              more significant, for me, was that I'd occasionally drag a pin as well as the
              solder. It is a total bear to try to straighten a bent fine pitch pin that has
              been soldered to the wrong pad.

              > For the discretes, I used the same small chisel tip as for through-hole. I
              > placed all the resistors of a given value on the board next to their pads;
              > fluxed up the pads, applied a tiny bit of solder to one side of each footprint
              > with the iron tip; then go through and align each resistor, heat up the
              > solder, and place the resistor. Finally, turn the board around and solder the
              > other side of all discretes by keeping the tip loaded with a bit of solder and
              > touching it to each unoldered joint.

              I had a lot of problem with "tombstoning", where the molten solder drags
              the tiny SMT part vertical, due to surface tension effects. Hot air mitigates
              that by melting both sides at once, or nearly so. (Also, the solder paste
              helps tack down the un-melted side.)

              > Afterwards, I wash the whole board in warm running tap water, and scrub really
              > well with a toothbrush until the foaming stops. The boards look really
              > professional.

              I also worried about the amount of flux getting under the larger SMT
              components, and whether I could successfully scrub under there.

              > Actually, although I hesitated forever before jumping on the bandwagon, I find
              > SMT much faster and cleaner than through hole: No missed joints; no bending
              > leads, taping the part in place; no clipping leads, no tedious pin-by-pin
              > soldering, and fewer touch-ups. The only new tools required are a $1.50 pair
              > of sharp-tip tweezers from the dollar store, a head-band magnifier, and (I
              > consider optional) a $15 hoof tip for my soldering iron.

              In general I agree, with the exception of the 0.5mm stuff, which is just too
              small for my fine motor tremor, and aging eyes.

              > I guess that was pretty long-winded, but despite having designed several
              > SMT-based boards, and having done some SMT rework, this was the first
              > decent-sized SMT project with fine-pitch parts I actually built up by myself.
              > I was surprised at how easy it was (with a lot of flux). It helped that the
              > board was well laid out with lots of space around most of the parts.

              I also used quite a bit of flux, especially in rework :-).

              Most of the layout is inherited from Jim's original version, so I can't take
              any credit for that.

              Vince
            Your message has been successfully submitted and would be delivered to recipients shortly.