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

Pride came before the fall

Expand Messages
  • fixitsan2
    Terrible sorry if this affects you, but it would seem that some pics with IV-17 Smartsocket code have been sent out with a baud rate of 38400. If you have been
    Message 1 of 17 , Dec 24, 2011
    • 0 Attachment
      Terrible sorry if this affects you, but it would seem that some pics with IV-17 Smartsocket code have been sent out with a baud rate of 38400.

      If you have been having a problem getting your Smartsocket to work, please consider testing with 38400 baud data.

      The PIC datasheet has some confusion about the setting of the baudrate generator BRGH bit on an earlier device which I used for prototyping, before I went to the 18f25k22 and the subsequent changes I made may have been carried over by the inclusion of a library file containing the change.

      At 38400 baud the 18f25k22 woorks flawlessly, unlike previous devices, which is possibly due to the better quality internal oscillator used in this device, but I haven't looked too closely at it to confirm that. I just know that I haven't ever had to tweak the clock of an 18f25k22 like I have had to with, say, the 16F690 in order to get reliable serial data.

      I am tempted to leave the speed at 38400, but it is no big deal to make it a pin selectable function to include 38400 and 9600 baudrates.

      Guus, if this is the source of the problem with your kit, I can only apologise and will send it back with a photo of my Rudolph-red face.

      Chris
    • guus.assmann@wolmail.nl
      Hello Chris, Let s hope it is something as simple as that. And if anything, it only shows that you re human as well. As such, you re allowed to make mistakes.
      Message 2 of 17 , Dec 24, 2011
      • 0 Attachment
        Hello Chris,

        Let's hope it is something as simple as that.
        And if anything, it only shows that you're human as well. As such, you're
        allowed to make mistakes.

        BR/
        Guus

        >-- Oorspronkelijk bericht --
        >To: smartsockets@yahoogroups.com
        >From: "fixitsan2" <fixitsan@...>
        >Date: Sat, 24 Dec 2011 13:35:47 -0000
        >Subject: [smartsockets] Pride came before the fall
        >Reply-To: smartsockets@yahoogroups.com
        >
        >
        >Terrible sorry if this affects you, but it would seem that some pics with
        >IV-17 Smartsocket code have been sent out with a baud rate of 38400.
        >
        >If you have been having a problem getting your Smartsocket to work, please
        >consider testing with 38400 baud data.
        >
        >The PIC datasheet has some confusion about the setting of the baudrate generator
        >BRGH bit on an earlier device which I used for prototyping, before I went
        >to the 18f25k22 and the subsequent changes I made may have been carried
        over
        >by the inclusion of a library file containing the change.
        >
        >At 38400 baud the 18f25k22 woorks flawlessly, unlike previous devices, which
        >is possibly due to the better quality internal oscillator used in this device,
        >but I haven't looked too closely at it to confirm that. I just know that
        >I haven't ever had to tweak the clock of an 18f25k22 like I have had to
        >with, say, the 16F690 in order to get reliable serial data.
        >
        >I am tempted to leave the speed at 38400, but it is no big deal to make
        it
        >a pin selectable function to include 38400 and 9600 baudrates.
        >
        >Guus, if this is the source of the problem with your kit, I can only apologise
        >and will send it back with a photo of my Rudolph-red face.
        >
        >Chris
        >
      • fixitsan2
        ... Would anybody have a problem if 38400 becomes the default speed and 9600 became the option ? I certainly am happy with the performance of the 18f25k22 at
        Message 3 of 17 , Dec 24, 2011
        • 0 Attachment
          --- In smartsockets@yahoogroups.com, "guus.assmann@..." <guus.assmann@...> wrote:
          >
          > Hello Chris,
          >
          > Let's hope it is something as simple as that.
          > And if anything, it only shows that you're human as well. As such, you're
          > allowed to make mistakes.
          >
          > BR/
          > Guus


          Would anybody have a problem if 38400 becomes the default speed and 9600 became the option ? I certainly am happy with the performance of the 18f25k22 at the higher speed.

          I think the Arduino can run at 38400, for the builders preferring to make Arduino controllers.

          My mind has just started to mull over how best I could implement 'approximate clock' time displays as an selectable option - and I was hoping for a quiet Christmas eve playing bboard games with my family.

          Things will look better once I have a drink !
          Cheers everybody

          Chris
        • Derryck Croker
          I m still constructing the top deck , haven t got as far as that yet! Those IV-14 wires are pesky ;-) Slight error on that PCB BTW, C6 is labelled C5 (which
          Message 4 of 17 , Dec 25, 2011
          • 0 Attachment
            I'm still constructing the "top deck", haven't got as far as that yet! Those IV-14 wires are pesky ;-)

            Slight error on that PCB BTW, C6 is labelled C5 (which is also present on the controller board), and possibly its pads are slightly on the wider side of spacing.

            I'd probably prefer a lower baud rate to make interfacing with lower-spec PICs easier (I'm considering adding a FLW generator). I do have a PICkit2 programmer.

            --

            Cheers

            Derryck



            On 24 Dec 2011, at 13:35, fixitsan2 wrote:

            > Terrible sorry if this affects you, but it would seem that some pics with IV-17 Smartsocket code have been sent out with a baud rate of 38400.
            >
            > If you have been having a problem getting your Smartsocket to work, please consider testing with 38400 baud data.
            >
          • fixitsan2
            ... Have you tried an IV-17 or IV-4 instead Derryck, might be easier ;) ... Heck. Thanks for pointing out the naming error. The pad size issue will need to be
            Message 5 of 17 , Dec 25, 2011
            • 0 Attachment
              --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
              >
              > I'm still constructing the "top deck", haven't got as far as that yet! Those IV-14 wires are pesky ;-)

              Have you tried an IV-17 or IV-4 instead Derryck, might be easier ;)



              >
              > Slight error on that PCB BTW, C6 is labelled C5 (which is also present on the controller board), and possibly its pads are slightly on the wider side of spacing.

              Heck. Thanks for pointing out the naming error. The pad size issue will need to be tweaked in another version, but probably supplying larger sized capacitors will make it easier to live with until then.



              >
              > I'd probably prefer a lower baud rate to make interfacing with lower-spec PICs easier (I'm considering adding a FLW generator). I do have a PICkit2 programmer.
              >

              I know I mentioned it here first, but now that I look harder most pics can do the speed easily enough, even a bottom line capable 16f631/16f690 can do upto 115.2k at 3.6864MHz, but it is trivial to make it pin selectable in future versions


              If anyone has been negatively affected by a change of speed up to the faster 38400 baud and needs a lower speed pic in the meantime please don't hesitate to contact me immediately for a quick solution.My error is my responsibility :)


              Chris
            • Terry
              ... Here s the method I use - starting on the left side of the tube (where the index pin 22 is), I isolate the 7 leads that will go on the left strip of pads
              Message 6 of 17 , Dec 25, 2011
              • 0 Attachment
                --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
                > I'm still constructing the "top deck", haven't got as far as that yet! Those IV-14 wires are pesky ;-)

                Here's the method I use - starting on the left side of the tube (where the index pin 22 is), I isolate the 7 leads that will go on the left strip of pads and curl the other leads back slightly. I then hold the tube at an angle (tilted toward the front) and tilt it back, inserting each of the 7 leads in sequence from front to back. Once all 7 are in, I bend over the front and back leads around 1/4" from the end (to keep the tube from popping out). Next I do the front and rear strips of 3 from left to right. These can be slightly curved to get them into the holes. Again, I bend over the last 2 pins on each row to keep the tube from popping out. Lastly, I do the 8 pins on the right side, which will need a more pronounced bend to get them into the holes.

                Once all of the pins are in, I rock the tube from front to back to push it further into the board. Using needlenose pliers, I gently tug on any leads that try to bunch up between the tube and the board. Once I have the tube down approximately as low as I want it, I bend over the 4 corner leads to hold it in position.

                I repeat the procedure for the other 3 tubes. After all 4 tubes are on the board, I align the first tube as I want it and use solder to tack only the 4 corner leads to the board. I then do the same with the other 3 tubes, making sure that the tubes are in the same relative position between the front of the board and the rear, that the bottom of each tube's segments are level with each other, the tubes are facing forward exactly and not tilted toward / away from each other. Lastly, I look down the row of tubes to make sure they're all aligned. Minor adjustments can be made at this point without needing to desolder the 4 pins per tube already soldered. A major misalignment may require one or more pins to be desoldered (or at least melted enough to move the tube).

                I decided to position my tubes further away from the board than in the illustrations in the assembly manual. This may reduce stress on the leads where they pass through the glass on the base of the tube. Of course, if you are planning on fitting the completed board into an existing case, verify that it will still fit (and look OK) with the tubes mounted further away from the board.

                Here's a photo showing the lead detail from one of the boards I assembled: http://www.tmk.com/transient/6F5S8472-s.jpg

                > Slight error on that PCB BTW, C6 is labelled C5 (which is also present on the controller board), and possibly its pads are slightly on the wider side of spacing.

                I actually found that capacitor easier to solder than the resistors on both boards due to the "overhang" of the pads: http://www.tmk.com/transient/6F5S8466-s.jpg
              • shklaw75
                Hi Chris, How about keeping the default at 9600 and having the option to change to 38400 if people desire? That keeps the compatibility with slower devices, as
                Message 7 of 17 , Dec 27, 2011
                • 0 Attachment
                  Hi Chris,

                  How about keeping the default at 9600 and having the option to change to 38400 if people desire? That keeps the compatibility with slower devices, as Derryck mentioned, and more stability at the slower speed?

                  You mentioned that there was some issues at the higher speed: http://groups.yahoo.com/group/smartsockets/message/800
                  would it be easier to keep it at the slower speed? (if it ain't broke!)

                  Cheers.
                  Simon



                  --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:

                  > Would anybody have a problem if 38400 becomes the default speed and 9600 became the option ? I certainly am happy with the performance of the 18f25k22 at the higher speed.
                  >
                  > I think the Arduino can run at 38400, for the builders preferring to make Arduino controllers.
                  >
                  > My mind has just started to mull over how best I could implement 'approximate clock' time displays as an selectable option - and I was hoping for a quiet Christmas eve playing bboard games with my family.
                  >
                  > Things will look better once I have a drink !
                  > Cheers everybody
                  >
                  > Chris
                  >
                • fixitsan2
                  ... Thanks for the comment Simon. The reason I ask about it is because, as I pointed out already, the reliability of 38400 on the 18f25k22 is faultless. I ve
                  Message 8 of 17 , Dec 27, 2011
                  • 0 Attachment
                    --- In smartsockets@yahoogroups.com, "shklaw75" <shklaw75@...> wrote:
                    >
                    > Hi Chris,
                    >
                    > How about keeping the default at 9600 and having the option to change to 38400 if people desire? That keeps the compatibility with slower devices, as Derryck mentioned, and more stability at the slower speed?
                    >
                    > You mentioned that there was some issues at the higher speed: http://groups.yahoo.com/group/smartsockets/message/800
                    > would it be easier to keep it at the slower speed? (if it ain't broke!)
                    >
                    > Cheers.
                    > Simon
                    >
                    >

                    Thanks for the comment Simon.
                    The reason I ask about it is because, as I pointed out already, the reliability of 38400 on the 18f25k22 is faultless. I've put it down to the excellent oscillator which is used in this device, much better than the older internal RC device.

                    The Smartsocket doesn't communicate with anything but Smartsockets and hosts, so it is only the host which needs to be considered for compatibility. Any host which has enough memory to to make effctive use of the Smartsockets message display capabilities is probably already compatible.

                    I was the one who suggested keeping the speed low first, but now that I've thought about it more I can't see any reason not to use higher speeds if they are available, which goes towards reducing latency on arrays using several SS's daisychained togather.

                    I would be interested in hearing from someone who can't run at 38400 to say what the reason is ?

                    Chris
                  • Derryck Croker
                    ... Thanks - my method was to start at the index and work clockwise and anti-clockwise bending over each wire as it went in to keep in place. Flat nose pliers
                    Message 9 of 17 , Dec 27, 2011
                    • 0 Attachment
                      On 26 Dec 2011, at 04:12, Terry wrote:

                      >
                      >
                      > --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
                      >> I'm still constructing the "top deck", haven't got as far as that yet! Those IV-14 wires are pesky ;-)
                      >
                      > Here's the method I use - starting on the left side of the tube (where the index pin 22 is), I isolate the 7 leads

                      Thanks - my method was to start at the index and work clockwise and anti-clockwise bending over each wire as it went in to keep in place. Flat nose pliers and wide bladed tweezers to straighten the wires and pull through. Have to be careful that a rotational set doesn't take place.

                      Construction has had to take a back seat this week.

                      --

                      Cheers

                      Derryck
                    • Derryck Croker
                      ... Inflation ;-) ... No problem - anyone with the skills to solder this kit will soon sort it out. Hmm, I d probably label the bags with the component numbers
                      Message 10 of 17 , Dec 27, 2011
                      • 0 Attachment
                        On 25 Dec 2011, at 10:36, fixitsan2 wrote:

                        >
                        >
                        > --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
                        >>
                        >> I'm still constructing the "top deck", haven't got as far as that yet! Those IV-14 wires are pesky ;-)
                        >
                        > Have you tried an IV-17 or IV-4 instead Derryck, might be easier ;)

                        Inflation ;-)


                        >>
                        >> Slight error on that PCB BTW, C6 is labelled C5 (which is also present on the controller board), and possibly its pads are slightly on the wider side of spacing.
                        >
                        > Heck. Thanks for pointing out the naming error. The pad size issue will need to be tweaked in another version, but probably supplying larger sized capacitors will make it easier to live with until then.

                        No problem - anyone with the skills to solder this kit will soon sort it out. Hmm, I'd probably label the bags with the component numbers now I think of it.

                        >> I'd probably prefer a lower baud rate to make interfacing with lower-spec PICs easier (I'm considering adding a FLW generator). I do have a PICkit2 programmer.
                        >>
                        >
                        > I know I mentioned it here first, but now that I look harder most pics can do the speed easily enough, even a bottom line capable 16f631/16f690 can do upto 115.2k at 3.6864MHz, but it is trivial to make it pin selectable in future versions

                        I guess I need to up my game from the 16F628 and internal osc (I was going to use an eeprom to hold the data for the FLW) ;-)

                        Cheers

                        Derryck
                      • fixitsan2
                        ... I don t see why you can t use the 16F628 Derryck, it can do serial data upto 250k baud off the internal 4MHz oscillator, it s all in the datasheet. However
                        Message 11 of 17 , Dec 27, 2011
                        • 0 Attachment
                          --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
                          >

                          >
                          > I guess I need to up my game from the 16F628 and internal osc (I was going to use an eeprom to hold the data for the FLW) ;-)
                          >
                          > Cheers
                          >
                          > Derryck
                          >

                          I don't see why you can't use the 16F628 Derryck, it can do serial data upto 250k baud off the internal 4MHz oscillator, it's all in the datasheet.

                          However (!) one gotcha is that as I've already mentioned, the internal oscillators of early pics are more susceptible to temperature drift than the newer ones. If the baudrate is accurate at 25 Celsius at 5V don't expect it to keep working at 10 Celsius too.(or at 4.8V either) It might work, it might not. Results vary batch to batch

                          I can't see you having a problem though, with running the 16F628 from a 4MHz crystal.

                          Chris
                        • fixitsan2
                          ... Actually Derryck, according to Farnell, the 16F1518 (latest series) is cheaper than the 16F628. It has more than enough onboard memory to store over 4000
                          Message 12 of 17 , Dec 27, 2011
                          • 0 Attachment
                            --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:
                            >

                            >
                            > I don't see why you can't use the 16F628 Derryck, it can do serial data upto 250k baud off the internal 4MHz oscillator, it's all in the datasheet.


                            Actually Derryck, according to Farnell, the 16F1518 (latest series) is cheaper than the 16F628. It has more than enough onboard memory to store over 4000 four letter words as well as your program, and will cope with 38400baud without a crystal.

                            It would save you money because it is cheaper because you don't need another IC(eeprom), or a seperate socket, or a crystal. Unless you have 16F628's with nothing to do it saves you a lot of trouble and gains you a lot of simplicity and stability/reliability.

                            Chris
                          • Neil Breeden
                            For flying lead tubes simply cut each lead about 1/8 inch shorter as you go around the tube, this makes the ends of the leads like a little staircase, then
                            Message 13 of 17 , Dec 27, 2011
                            • 0 Attachment
                              For flying lead tubes simply cut each lead about 1/8 inch shorter as you go around the tube, this makes the ends of the leads like a little staircase, then insert the longest one, the next longest etc. Be sure to plan well so you leave enough length to have the tube sit at the desired height once all leads are in place. 

                              Sent from my iPad

                              On Dec 27, 2011, at 9:04, Derryck Croker <derryck@...> wrote:

                               

                              On 25 Dec 2011, at 10:36, fixitsan2 wrote:

                              >
                              >
                              > --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
                              >>
                              >> I'm still constructing the "top deck", haven't got as far as that yet! Those IV-14 wires are pesky ;-)
                              >
                              > Have you tried an IV-17 or IV-4 instead Derryck, might be easier ;)

                              Inflation ;-)

                              >>
                              >> Slight error on that PCB BTW, C6 is labelled C5 (which is also present on the controller board), and possibly its pads are slightly on the wider side of spacing.
                              >
                              > Heck. Thanks for pointing out the naming error. The pad size issue will need to be tweaked in another version, but probably supplying larger sized capacitors will make it easier to live with until then.

                              No problem - anyone with the skills to solder this kit will soon sort it out. Hmm, I'd probably label the bags with the component numbers now I think of it.

                              >> I'd probably prefer a lower baud rate to make interfacing with lower-spec PICs easier (I'm considering adding a FLW generator). I do have a PICkit2 programmer.
                              >>
                              >
                              > I know I mentioned it here first, but now that I look harder most pics can do the speed easily enough, even a bottom line capable 16f631/16f690 can do upto 115.2k at 3.6864MHz, but it is trivial to make it pin selectable in future versions

                              I guess I need to up my game from the 16F628 and internal osc (I was going to use an eeprom to hold the data for the FLW) ;-)

                              Cheers

                              Derryck

                            • shklaw75
                              Thanks Chris, sounds like there are more advantages than disadvantages with the faster speed. Simon
                              Message 14 of 17 , Dec 27, 2011
                              • 0 Attachment
                                Thanks Chris, sounds like there are more advantages than disadvantages with the faster speed.

                                Simon

                                --- In smartsockets@yahoogroups.com, "fixitsan2" <fixitsan@...> wrote:
                                >
                                > Thanks for the comment Simon.
                                > The reason I ask about it is because, as I pointed out already, the reliability of 38400 on the 18f25k22 is faultless. I've put it down to the excellent oscillator which is used in this device, much better than the older internal RC device.
                                >
                                > The Smartsocket doesn't communicate with anything but Smartsockets and hosts, so it is only the host which needs to be considered for compatibility. Any host which has enough memory to to make effctive use of the Smartsockets message display capabilities is probably already compatible.
                                >
                                > I was the one who suggested keeping the speed low first, but now that I've thought about it more I can't see any reason not to use higher speeds if they are available, which goes towards reducing latency on arrays using several SS's daisychained togather.
                                >
                                > I would be interested in hearing from someone who can't run at 38400 to say what the reason is ?
                                >
                                > Chris
                                >
                              • Derryck Croker
                                ... Started off doing that, but chickened out with thoughts of your last sentence came into view ;-) Cheers Derryck
                                Message 15 of 17 , Dec 28, 2011
                                • 0 Attachment
                                  On 27 Dec 2011, at 18:51, Neil Breeden wrote:

                                  >
                                  >
                                  > For flying lead tubes simply cut each lead about 1/8 inch shorter as you go around the tube, this makes the ends of the leads like a little staircase, then insert the longest one, the next longest etc. Be sure to plan well so you leave enough length to have the tube sit at the desired height once all leads are in place.

                                  Started off doing that, but chickened out with thoughts of your last sentence came into view ;-)

                                  Cheers

                                  Derryck
                                • Derryck Croker
                                  ... Plenty of poke there ;-) Thanks for the pointer! One silly question, when I fire up for the first time, and without connecting up to some kind of terminal
                                  Message 16 of 17 , Dec 28, 2011
                                  • 0 Attachment
                                    On 27 Dec 2011, at 18:40, fixitsan2 wrote:

                                    > 16F1518

                                    Plenty of poke there ;-)

                                    Thanks for the pointer!

                                    One silly question, when I fire up for the first time, and without connecting up to some kind of terminal to send commands from, will I expect to see some kind of greeting message?

                                    Cheers

                                    Derryck
                                  • fixitsan2
                                    ... no Derryck, the display is purposely blank. I figured that some people would want to power the display s down when there is no data to display, say if
                                    Message 17 of 17 , Dec 28, 2011
                                    • 0 Attachment
                                      --- In smartsockets@yahoogroups.com, Derryck Croker <derryck@...> wrote:
                                      >
                                      >
                                      > One silly question, when I fire up for the first time, and without connecting up to some kind of terminal to send commands from, will I expect to see some kind of greeting message?
                                      >
                                      > Cheers
                                      >
                                      > Derryck
                                      >

                                      no Derryck, the display is purposely blank. I figured that some people would want to power the display's down when there is no data to display, say if you're making an RSS feed reader or an SMS display terminal. Every time you receive a text message under those conditions the tubes will cycle through a welcome message first, probably being quite annoying.

                                      I have just added a test message routine. Pulling pin 1 low forces the code into routines which produces test patterns suitable for indicating shorts between segments as well as shorts between screens, which could possibly occur if there is a solder bridge on any tube driver pins. The test also confirms that the pic controls the tube driver correctly. I am going to add a pin which mirrors the state of pin 1, which will confirm that the code is actually when it's state flips in accordance with pin1's state.

                                      For the baudrate issues, I'm looking to implement auto baudrate detect which just seems like it will take all worries away in one fell swoop.

                                      Anybody looking to upgrade can do so, if you have the blank pics I sent out with your kit I can program them for you.

                                      I think the autobaud version will not be downwards compatible with previous versions, but I have to prove that yet. I will save the code with latest features seperately before adding autobaud.

                                      I have added a couple of GeekKlok fonts (all credit to the authour Ray Weisling) they are included only for demonstration purposes.

                                      More news soon !
                                      Chris
                                    Your message has been successfully submitted and would be delivered to recipients shortly.