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

Own aduc7020 pcb design

Expand Messages
  • pellepl
    Hi, anyone out there ever made a (working) pcb of their own with the aduc7020 LFCSP package? The same which is on the minikit. I ve successfully done some
    Message 1 of 11 , Nov 10, 2008
    • 0 Attachment
      Hi,

      anyone out there ever made a (working) pcb of their own with the
      aduc7020 LFCSP package? The same which is on the minikit.

      I've successfully done some boards for the aduc7128 and 7129 lqfp
      packages, which work smoothly. But for the 7020 the uC is more or less
      dead. The internal clock seem to be alive, but all other IOs are high-z
      on 2.7 volts.

      Anyone? I'll gladly share gerbers or eagle-files ;)

      Cheers / Peter
    • paloalgodon
      ... Didn t have any issues myself. We did accidentally order the -I parts with the I2C instead of uart bootloader - that gave us a headscratch for a little
      Message 2 of 11 , Nov 10, 2008
      • 0 Attachment
        --- In ADuC_ARM@yahoogroups.com, "pellepl" <pellepl@...> wrote:
        >
        > Hi,
        >
        > anyone out there ever made a (working) pcb of their own with the
        > aduc7020 LFCSP package? The same which is on the minikit.
        >
        > I've successfully done some boards for the aduc7128 and 7129 lqfp
        > packages, which work smoothly. But for the 7020 the uC is more or less
        > dead. The internal clock seem to be alive, but all other IOs are high-z
        > on 2.7 volts.
        >
        > Anyone? I'll gladly share gerbers or eagle-files ;)
        >
        > Cheers / Peter
        >

        Didn't have any issues myself. We did accidentally order the -I parts
        with the I2C instead of uart bootloader - that gave us a headscratch
        for a little while. As far as layout, it was on a 4-layer pcb with
        groundplane, and a 0402, 0.1uf cap immediately adjacent to every DVcc
        and AVcc pin...

        Steve
      • pellepl
        ... less ... high-z ... parts ... DVcc ... Hi Steve, thanks for answering! There s hope then if you got it working! :) I totally understand if you don t find
        Message 3 of 11 , Nov 11, 2008
        • 0 Attachment
          --- In ADuC_ARM@yahoogroups.com, "paloalgodon" <stevefranks@...>
          wrote:
          >
          > --- In ADuC_ARM@yahoogroups.com, "pellepl" <pellepl@> wrote:
          > >
          > > Hi,
          > >
          > > anyone out there ever made a (working) pcb of their own with the
          > > aduc7020 LFCSP package? The same which is on the minikit.
          > >
          > > I've successfully done some boards for the aduc7128 and 7129 lqfp
          > > packages, which work smoothly. But for the 7020 the uC is more or
          less
          > > dead. The internal clock seem to be alive, but all other IOs are
          high-z
          > > on 2.7 volts.
          > >
          > > Anyone? I'll gladly share gerbers or eagle-files ;)
          > >
          > > Cheers / Peter
          > >
          >
          > Didn't have any issues myself. We did accidentally order the -I
          parts
          > with the I2C instead of uart bootloader - that gave us a headscratch
          > for a little while. As far as layout, it was on a 4-layer pcb with
          > groundplane, and a 0402, 0.1uf cap immediately adjacent to every
          DVcc
          > and AVcc pin...
          >
          > Steve
          >

          Hi Steve,

          thanks for answering! There's hope then if you got it working! :)

          I totally understand if you don't find the time to pursue this, but I
          would be ever so grateful if you did of course.

          I'm running a two-layer pcb with 0603 caps. The decouplings and the
          470nF caps are on the bottom side connected with vias close to the
          pins. The Vdd and analog Vdd are 3.3V and have ~3 mV ripple. The
          minikit has 5 mV ripple so that should be ok. The LVDD outputs 2.5V
          as expected. The digital and analog ground planes are connected under
          the ADuC7020 with some small strips of copper as stated in the PCB
          recommendations of the datasheet.

          The main problem is that the bootloader won't activate when
          bootmoding and resetting. The UART connection is ok. I've soldered 3
          boards with the same result. Tried some different power and UART
          communications solutions without any difference.

          Produced some images of the schematic and the board, but I really
          can't find how to attach them. Will try to link them to my server
          later.

          Any clues?

          Again, thanks for the time!

          Cheers / Peter
        • katagia
          ... I had no bigger problems with it. Maybe a silly question but have you checked every pin is soldered correct? The uC has no legs and distances are critical.
          Message 4 of 11 , Nov 12, 2008
          • 0 Attachment
            --- In ADuC_ARM@yahoogroups.com, "pellepl" <pellepl@...> wrote:
            > anyone out there ever made a (working) pcb of their own with the
            > aduc7020 LFCSP package? The same which is on the minikit.
            >
            > I've successfully done some boards for the aduc7128 and 7129 lqfp
            > packages, which work smoothly. But for the 7020 the uC is more or less
            > dead. The internal clock seem to be alive, but all other IOs are high-z
            > on 2.7 volts.

            I had no bigger problems with it. Maybe a silly question but have you
            checked
            every pin is soldered correct? The uC has no legs and distances are
            critical.

            Do you have some protection layer over your copper lanes? Sorry, I
            don't know the correct english term for it. This could cause a gap
            between pcb and chip.

            In my first design I had problems with uart pins (bad contact) and I
            had to resolder them (I soldered the chip by hand).

            Everything else is fine? Pullup at reset, Pulldown at download and
            most important 470nF connected to LVdd?
          • pellepl
            ... less ... high-z ... you ... No questions are silly in this context - in fact, I hope the problem is silly so it will be easy to fix ;) Checked the pins -
            Message 5 of 11 , Nov 12, 2008
            • 0 Attachment
              --- In ADuC_ARM@yahoogroups.com, "katagia" <TiloLutz@...> wrote:
              >
              > --- In ADuC_ARM@yahoogroups.com, "pellepl" <pellepl@> wrote:
              > > anyone out there ever made a (working) pcb of their own with the
              > > aduc7020 LFCSP package? The same which is on the minikit.
              > >
              > > I've successfully done some boards for the aduc7128 and 7129 lqfp
              > > packages, which work smoothly. But for the 7020 the uC is more or
              less
              > > dead. The internal clock seem to be alive, but all other IOs are
              high-z
              > > on 2.7 volts.
              >
              > I had no bigger problems with it. Maybe a silly question but have
              you
              > checked
              > every pin is soldered correct? The uC has no legs and distances are
              > critical.
              >
              > Do you have some protection layer over your copper lanes? Sorry, I
              > don't know the correct english term for it. This could cause a gap
              > between pcb and chip.
              >
              > In my first design I had problems with uart pins (bad contact) and I
              > had to resolder them (I soldered the chip by hand).
              >
              > Everything else is fine? Pullup at reset, Pulldown at download and
              > most important 470nF connected to LVdd?
              >

              No questions are silly in this context - in fact, I hope the problem
              is "silly" so it will be easy to fix ;)

              Checked the pins - handsoldered it also, it was some work. Made three
              boards with three different uCs, same result. Nice to hear that other
              people handsoldered it successfully also.

              Checked the pins with a microscope and tested the critical with a
              multimeter. I might have missed with the probes, but it seemed
              connected on the correct spots and disconnected on the wrong.

              Hmm, you mean soldermask? Yep, got that, but the microscope indicates
              everythings ok.

              The pullups and pulldowns are there, I was "inspired" by the minikit
              design - ripped it right off I did ;) That's what bothers me the most.

              I'll upload some schematics and gerbers soon enough in the file area.

              Thanks!
            • pellepl
              Created a photoalbum of the bad design. More stuff coming up here. http://groups.yahoo.com/group/ADuC_ARM/photos/album/946558215/pic/list
              Message 6 of 11 , Nov 12, 2008
              • 0 Attachment
                Created a photoalbum of the bad design. More stuff coming up here.

                http://groups.yahoo.com/group/ADuC_ARM/photos/album/946558215/pic/list
              • paloalgodon
                ... I handsoldered mine, but used hot air, then retouched with an iron. Don t know the effects of leaving the solder slug underneath unconnected. I hope all
                Message 7 of 11 , Nov 12, 2008
                • 0 Attachment
                  --- In ADuC_ARM@yahoogroups.com, "pellepl" <pellepl@...> wrote:
                  >
                  > Created a photoalbum of the bad design. More stuff coming up here.
                  >
                  > http://groups.yahoo.com/group/ADuC_ARM/photos/album/946558215/pic/list
                  >

                  I handsoldered mine, but used hot air, then retouched with an iron.
                  Don't know the effects of leaving the solder slug underneath unconnected.

                  I hope all those vias in green underneath are all ground. I'd suggest
                  that as an issue if not...

                  Don't see anything else glaringly wrong, unfortunately.

                  Steve
                • martin_schoenegg
                  pellepl ... list We use to make longer pads so we are able to use a solder iron to heat the pad beside the IC. Our group made several layouts (even single
                  Message 8 of 11 , Nov 13, 2008
                  • 0 Attachment
                    pellepl

                    > http://groups.yahoo.com/group/ADuC_ARM/photos/album/946558215/pic/
                    list

                    We use to make longer pads so we are able to use a solder iron to
                    heat the pad beside the IC. Our group made several layouts (even
                    single sided) without mayor problems. All hand soldered with a simple
                    solder iron, no hot air or such specials. A few hints:

                    Please do not connect or solder the metal surface on the bottom side.
                    The mass of tin in the hole will easyly overheat your IC.

                    Don't make wires nor vias under the chip. The metal surface is
                    internally connected to GND and may give a short cirquit.

                    Colophonium in alcohol is your friend for such handmade things. Use
                    enough of such solution or an other flux to prevent shorts

                    Good luck

                    Martin
                  • pellepl
                    ... longer pads so we are able to use a solder iron to ... simple ... side. ... Hi, thanks for all tips! Using plain solder iron (and some wick) also. The vias
                    Message 9 of 11 , Nov 20, 2008
                    • 0 Attachment
                      --- In ADuC_ARM@yahoogroups.com, "martin_schoenegg" > We use to make
                      longer pads so we are able to use a solder iron to
                      > heat the pad beside the IC. Our group made several layouts (even
                      > single sided) without mayor problems. All hand soldered with a
                      simple
                      > solder iron, no hot air or such specials. A few hints:
                      >
                      > Please do not connect or solder the metal surface on the bottom
                      side.
                      > The mass of tin in the hole will easyly overheat your IC.
                      >
                      > Don't make wires nor vias under the chip. The metal surface is
                      > internally connected to GND and may give a short cirquit.
                      >
                      > Colophonium in alcohol is your friend for such handmade things. Use
                      > enough of such solution or an other flux to prevent shorts
                      >
                      > Good luck
                      >
                      > Martin
                      >

                      Hi,

                      thanks for all tips!

                      Using plain solder iron (and some wick) also. The vias and wires
                      under the chip are GND so a short should not cause any problem.
                      Rechecked and remeasured everything, the only thing I can come up
                      with is that I've heated the ADuCs too much/long. The hole was never
                      filled with tin. I'm using 280 celcius on the iron - what temp are
                      you guys using?

                      I'm thinking of redoing the PCB using a bigger board with less
                      constraints. Just to see if I can get the processor to kick in.

                      Thanks / Peter
                    • martin_schoenegg
                      ... never ... My Weller Magnastat prefers 370 °C and I never had any problems with this. It may be better to solder quick and hot than too cold and needing
                      Message 10 of 11 , Nov 21, 2008
                      • 0 Attachment
                        > with is that I've heated the ADuCs too much/long. The hole was
                        never
                        > filled with tin. I'm using 280 celcius on the iron - what temp are
                        > you guys using?

                        My Weller Magnastat prefers 370 °C and I never had any problems with
                        this. It may be better to solder quick and hot than too cold and
                        needing longer, but only as long as you use very little tin. The most
                        important thing may be a high quality solder tip.
                        BTW I prefer for prototyping the cheapest colophonium based Pb-tin
                        with 1mm diameter ;-)

                        Regards
                        Martin
                      • pellepl76
                        Answering my own question as I found the problem some long while ago - stupidly enough I d forgot to tie the JTAG pins high. Duuuh.... Cheers / P
                        Message 11 of 11 , Nov 28, 2010
                        • 0 Attachment
                          Answering my own question as I found the problem some long while ago - stupidly enough I'd forgot to tie the JTAG pins high. Duuuh....

                          Cheers / P

                          --- In ADuC_ARM@yahoogroups.com, "pellepl" <pellepl@...> wrote:
                          >
                          > Hi,
                          >
                          > anyone out there ever made a (working) pcb of their own with the
                          > aduc7020 LFCSP package? The same which is on the minikit.
                          >
                          > I've successfully done some boards for the aduc7128 and 7129 lqfp
                          > packages, which work smoothly. But for the 7020 the uC is more or less
                          > dead. The internal clock seem to be alive, but all other IOs are high-z
                          > on 2.7 volts.
                          >
                          > Anyone? I'll gladly share gerbers or eagle-files ;)
                          >
                          > Cheers / Peter
                          >
                        Your message has been successfully submitted and would be delivered to recipients shortly.