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

VIP on the Bench

Expand Messages
  • joshbensadon
    Hi Guys, Over the weekend, I ve cleaned up my bench so I could play with the RCA VIP and my home brewed VELF. A friend online sent me the games from Game
    Message 1 of 13 , Jun 23, 2014
      Hi Guys,

      Over the weekend, I've cleaned up my bench so I could play with the RCA VIP and my home brewed VELF.  A friend online sent me the games from Game Manual I in .WAV format and I loaded them into the VIP, tried the game, then saved them to cassette tape.  At first, they would not load, the output from my laptop and PC sound card was only 3v p-p.  But after connecting them to the speaker output from the amplified speakers, I was able to get a 6v p-p signal which worked very well.  Since the laptop is my backup plan for VCF-MW, I will build a small inline amplifier for this purpose.  The output from my cassette is about 5v p-p at maximum volume.  The signal is distorted by clipping, but that does not matter, in fact, it just makes the signal look more squarish.  It's even written in the COSMAC manuals that you should be using a high volume for play back.

      From the bench!
      Cheers,
      Josh
    • Dave McGuire
      ... Much awesomeness. I must caution against taking the high volume thing as a rule of thumb, though. The reason amplifier clipping creates distortion is
      Message 2 of 13 , Jun 23, 2014
        On 06/23/2014 10:51 AM, joshbensadon wrote:
        > Over the weekend, I've cleaned up my bench so I could play with the RCA
        > VIP and my home brewed VELF. A friend online sent me the games from
        > Game Manual I in .WAV format and I loaded them into the VIP, tried the
        > game, then saved them to cassette tape. At first, they would not load,
        > the output from my laptop and PC sound card was only 3v p-p. But after
        > connecting them to the speaker output from the amplified speakers, I was
        > able to get a 6v p-p signal which worked very well. Since the laptop is
        > my backup plan for VCF-MW, I will build a small inline amplifier for
        > this purpose. The output from my cassette is about 5v p-p at maximum
        > volume. The signal is distorted by clipping, but that does not matter,
        > in fact, it just makes the signal look more squarish. It's even written
        > in the COSMAC manuals that you should be using a high volume for play back.

        Much awesomeness.

        I must caution against taking the "high volume" thing as a rule of
        thumb, though. The reason amplifier clipping creates distortion is
        because it "squares off" the peaks, as you noticed, which creates large
        amounts of harmonics. If the demodulator circuit in the cassette
        readback system isn't set up to filter those out, it will push the error
        rate sky high. Try increasing the volume incrementally and see where
        you start getting errors, and look for the "sweet spot".

        -Dave

        --
        Dave McGuire, AK4HZ/3
        New Kensington, PA
      • joshbensadon
        Dave, You re right to say high volume shouldn t be a rule of thumb , as you said, square waves are full of harmonics, and it could be worse, driving
        Message 3 of 13 , Jun 23, 2014
          Dave, You're right to say "high volume shouldn't be a rule of thumb",  as you said, square waves are full of harmonics, and it could be worse, driving transistors like that into full saturation takes away from their bandwidth as there's additional recovery time needed.  What is even worse is when the power supply rails start collapsing and you see negative slopes or "reflections" against the rails.  My experiments this weekend are however only for the COSMAC VIP, and I was just pointing out what is written in the manual regarding the likely hood that more volume is needed.  There was no sweet spot from the Laptop output... you would think 3V is enough to run the VIP tape input, but I was getting errors.  The VIP expects a loud volume,  it is also apparently dependent upon the polarity of the signal too (according to the manual), but my cassette is reversing the polarity but it doesn't seem to matter so much.

          What surprises me is how well the cassette read happens... even with all the wow and flutter of this old cassette player.  When I play it through the speakers, it sounds terrible!  the lead in should be a 2Khz signal but the pitch fluctuates terribly. 

          I've heard of people trying MP3 players in place of cassettes, but without much luck, perhaps the volume of those devices isn't adequate either?

          Cheers,
          Josh
        • Dave McGuire
          ... Very cool. It was not my intention to criticize what you had done; I was just pointing it out for others who may be following this. ... I wonder if
          Message 4 of 13 , Jun 23, 2014
            On 06/23/2014 09:36 PM, joshbensadon wrote:
            > Dave, You're right to say "high volume shouldn't be a rule of thumb",
            > as you said, square waves are full of harmonics, and it could be worse,
            > driving transistors like that into full saturation takes away from their
            > bandwidth as there's additional recovery time needed. What is even
            > worse is when the power supply rails start collapsing and you see
            > negative slopes or "reflections" against the rails. My experiments this
            > weekend are however only for the COSMAC VIP, and I was just pointing out
            > what is written in the manual regarding the likely hood that more volume
            > is needed. There was no sweet spot from the Laptop output... you would
            > think 3V is enough to run the VIP tape input, but I was getting errors.
            > The VIP expects a loud volume, it is also apparently dependent upon the
            > polarity of the signal too (according to the manual), but my cassette is
            > reversing the polarity but it doesn't seem to matter so much.

            Very cool. It was not my intention to criticize what you had done; I
            was just pointing it out for others who may be following this.

            > Wha t surprises me is how well the cassette read happens... even with
            > all the wow and flutter of this old cassette player. When I play it
            > through the speakers, it sounds terrible! the lead in should be a 2Khz
            > signal but the pitch fluctuates terribly.

            I wonder if there's a PLL in the read circuitry. The VIP is a pretty
            simple setup; I'd be surprised if there were a PLL in there...but is there?

            > I've heard of people trying MP3 players in place of cassettes, but
            > without much luck, perhaps the volume of those devices isn't adequate
            > either?

            It's not the volume in this case.

            Even trying this was probably a bad idea. People (again, not *you*)
            take far too simplistic a view of this sort of thing, even people who
            should know better. It's the same type of uninformed bumbling that
            causes people to think "JPEG means graphics" and they use that format
            for EVERY KIND of graphics, whether it's an appropriate use for the task
            at hand or not...and then, duh, they wonder where all those ugly
            artifacts around sharp color transitions came from. People today seem
            to have no problem at all with the idea of diving in and using a complex
            tool without being willing to take the time learn the first thing about
            how it is PROPERLY used.

            [end rant]

            The problem here is that the MPEG layer 3 algorithm, much like the
            JPEG compression algorithm does for continuous-tone images and human
            vision, exploits weaknesses in human hearing to remove data that is
            "less important" to HUMANS getting the gist of the program material.
            Put simply, it removes data that represents sound components that we
            wouldn't readily perceive anyway. Both algorithms (MP3 and JPEG) are
            "lossy" compressors, so the data run through them is irrecoverably
            changed by a compression/decompression cycle. A computer's cassette
            port demodulator has very different characteristics from the human
            hearing system, and it will "hear" sound VERY differently when pumped
            through an MP3 compression/decompression cycle.

            -Dave

            --
            Dave McGuire, AK4HZ/3
            New Kensington, PA
          • Dave McGuire
            On 06/23/2014 09:56 PM, Dave McGuire Mcguire@neurotica.com I ve heard of people trying MP3 players in place of cassettes, but ... Yes, it s bad form to
            Message 5 of 13 , Jun 23, 2014
              On 06/23/2014 09:56 PM, Dave McGuire Mcguire@... >> I've heard
              of people trying MP3 players in place of cassettes, but
              >> without much luck, perhaps the volume of those devices isn't adequate
              >> either?
              >
              > It's not the volume in this case.
              >
              > Even trying this was probably a bad idea. People (again, not *you*)
              > take far too simplistic a view of this sort of thing, even people who
              > should know better. It's the same type of uninformed bumbling that
              > causes people to think "JPEG means graphics" and they use that format
              > for EVERY KIND of graphics, whether it's an appropriate use for the task
              > at hand or not...and then, duh, they wonder where all those ugly
              > artifacts around sharp color transitions came from. People today seem
              > to have no problem at all with the idea of diving in and using a complex
              > tool without being willing to take the time learn the first thing about
              > how it is PROPERLY used.
              >
              > [end rant]

              Yes, it's bad form to reply to my own message, but I was distracted
              when I typed that and I thought I should add a clarification.

              Why I complained specifically about overly simplistic views leading to
              experiments with MP3 players and such is the assumption that "MP3
              player" simply means "sound storage device", and a "sound storage
              device" can store ANY sound. That makes perfect sense...until one
              delves beneath the surface and explores HOW MP3 players actually store
              sound. Or, more to the point, how they encode it for storage.

              Now, that said, for lower-capacity audio-based storage schemes the MP3
              approach can theoretically work. Any type of phase encoding is right
              out (which is why VoIP with standard voice CODECs generally don't work
              very well with high-speed modems...high-speed modems use various forms
              of phase encoding) but plain FSK *may* get through if it's slow enough.

              -Dave

              --
              Dave McGuire, AK4HZ/3
              New Kensington, PA
            • Paul Birkel
              Have read about these problems elsewhere. I m curious whether folks have specifically encoded MP3 at the max. available sampling rate or just accepted a
              Message 6 of 13 , Jun 23, 2014
                Have read about these problems elsewhere.  I'm curious whether folks have specifically encoded MP3 at the max. available sampling rate or just accepted a default (and what that was)?  Also wondering if a little RC-smoothing on the audio-out might be helpful?

                On Mon, Jun 23, 2014 at 10:27 PM, Dave McGuire Mcguire@... [midatlanticretro] <midatlanticretro@yahoogroups.com> wrote:
                 

                On 06/23/2014 09:56 PM, Dave McGuire Mcguire@... >> I've heard


                of people trying MP3 players in place of cassettes, but
                >> without much luck, perhaps the volume of those devices isn't adequate
                >> either?
                >
                > It's not the volume in this case.
                >
                > Even trying this was probably a bad idea. People (again, not *you*)
                > take far too simplistic a view of this sort of thing, even people who
                > should know better. It's the same type of uninformed bumbling that
                > causes people to think "JPEG means graphics" and they use that format
                > for EVERY KIND of graphics, whether it's an appropriate use for the task
                > at hand or not...and then, duh, they wonder where all those ugly
                > artifacts around sharp color transitions came from. People today seem
                > to have no problem at all with the idea of diving in and using a complex
                > tool without being willing to take the time learn the first thing about
                > how it is PROPERLY used.
                >
                > [end rant]

                Yes, it's bad form to reply to my own message, but I was distracted
                when I typed that and I thought I should add a clarification.

                Why I complained specifically about overly simplistic views leading to
                experiments with MP3 players and such is the assumption that "MP3
                player" simply means "sound storage device", and a "sound storage
                device" can store ANY sound. That makes perfect sense...until one
                delves beneath the surface and explores HOW MP3 players actually store
                sound. Or, more to the point, how they encode it for storage.

                Now, that said, for lower-capacity audio-based storage schemes the MP3
                approach can theoretically work. Any type of phase encoding is right
                out (which is why VoIP with standard voice CODECs generally don't work
                very well with high-speed modems...high-speed modems use various forms
                of phase encoding) but plain FSK *may* get through if it's slow enough.


                -Dave

                --
                Dave McGuire, AK4HZ/3
                New Kensington, PA


              • mwillega
                If you have trouble with MP3 files, use AIF encoding, which does not compress. If you still have trouble, which I wouldn t expect, and your music player or
                Message 7 of 13 , Jun 24, 2014
                  If you have trouble with MP3 files, use AIF encoding, which does not compress.  If you still have trouble,  which I wouldn't expect, and your music player or laptop software has a treble control, turn the treble down.  This acts a bit like a low pass filter.

                  My experience is that with AIF files and a digital music player or PC software, your ability to read in a file is vastly improved over the typical cassette player. 

                  I usually find that I'm wrong when I make blanket statements, but I expect that any cassette input circuit worth using, would have to have some kind of low or band pass filter.  Even the SCELBI's tape input circuit which dates to 1974, includes such a filter.

                  I wrote some software that converts 6502 binary files to Apple compatible AIF files and shaping the output into a basic square wave works fine.  My files are set to a 8kHz sample rate to save space.  This is for the 2k/1K Hz signal that the Apple expects, so each data point is recorded with only 4 or 8 data samples.

                  Between audio output shaping and input filtering, the input circuit shouldn't notice the high frequency components commonly found in a typical square wave.

                  regards,
                  Mike W.


                • jack99rubin
                  Here s a soundcard oscilloscope that you might find helpful - Soundcard Scope http://www.zeitnitz.de/Christian/scope_en
                  Message 8 of 13 , Jun 24, 2014
                    Here's a "soundcard oscilloscope" that you might find helpful  - Soundcard Scope

                     

                    It also includes a signal generator so you can test in both directions. BTW, I've always used .wav format in this situation and have good success with AIM, KIM, etc.

                    Jack
                  • joshbensadon
                    I think Mike has a good point there, the computer output of non compressed audio probably makes the best (cleanest) signal. I had no problem playing these
                    Message 9 of 13 , Jun 24, 2014
                      I think Mike has a good point there, the computer output of non compressed audio probably makes the best (cleanest) signal.  I had no problem playing these .wav files, once I was able to increase the voltage output to about 5v p-p. 
                      Dave, there is no PLL or other type of demodulator on the input, it's just one op amp.  I don't really understand why it didn't work at 3v p-p since the input is nothing more than a single op amp meant to square up the signal.  It does have a low pass filter effect, but it's not a proper filter, ie, it's not a T circuit, but what they did was just place a capacitor across the feedback resistor.  That almost makes it work like an integrator.  I guess the intent is to improve it's response time and have the output swing rail to rail based more on the slope of the incoming signal.  Which is why I'm confused to see it not working at the 3Vp-p level.  I think it's time I pop the cover off again and poke around with my scope and signal generator and get to the bottom of this.  There's also another capacitor on the input, but that's just a DC blocker or coupling cap.

                      The signal isn't very special, it's a 2Khz cycle for a logic 0 and 800hz cycle for logic 1.  The software times the low to high to low, which makes it very sensitive to phase.

                      I'm testing the tape recordings and finding flaky results now... I'm not a happy camper.  But let the fun begin :)

                      J.
                    • Dave McGuire
                      ... Ideally I d agree, but we both know the shortcuts that were taken in hobbyist-targeted designs in the 1970s. The question isn t whether that signal is
                      Message 10 of 13 , Jun 24, 2014
                        On 06/24/2014 06:49 AM, mike@... [midatlanticretro] wrote:
                        > If you have trouble with MP3 files, use AIF encoding, which does not
                        > compress. If you still have trouble, which I wouldn't expect, and your
                        > music player or laptop software has a treble control, turn the treble
                        > down. This acts a bit like a low pass filter.
                        >
                        > My experience is that with AIF files and a digital music player or PC
                        > software, your ability to read in a file is vastly improved over the
                        > typical cassette player.
                        >
                        > I usually find that I'm wrong when I make blanket statements, but I
                        > expect that any cassette input circuit worth using, would have to have
                        > some kind of low or band pass filter. Even the SCELBI's tape input
                        > circuit which dates to 1974, includes such a filter.
                        >
                        > I wrote some software that converts 6502 binary files to Apple
                        > compatible AIF files and shaping the output into a basic square wave
                        > works fine. My files are set to a 8kHz sample rate to save space. This
                        > is for the 2k/1K Hz signal that the Apple expects, so each data point is
                        > recorded with only 4 or 8 data samples.
                        >
                        > Between audio output shaping and input filtering, the input circuit
                        > shouldn't notice the high frequency components commonly found in a
                        > typical square wave.

                        Ideally I'd agree, but we both know the shortcuts that were taken in
                        hobbyist-targeted designs in the 1970s. The question isn't whether that
                        signal is filtered, it's whether it's filtered *enough*.

                        I agree with the recommendation of an uncompressed file format. At
                        the very least, a non-lossy compression is called for here.

                        -Dave

                        --
                        Dave McGuire, AK4HZ/3
                        New Kensington, PA
                      • Ray Sills
                        And, all things considered... even with an uncompressed file... they won t be very large anyway. The baud rate for the VIP FSK tape system is about 100 baud.
                        Message 11 of 13 , Jun 24, 2014
                          And, all things considered... even with an uncompressed file... they
                          won't be very large anyway. The baud rate for the VIP FSK tape system
                          is about 100 baud. So even a "full" 4K would only need about 40
                          seconds of audio. The leader tone is way longer than it needs to be.
                          Apparently, the simple system used for the vip is fairly tolerant of
                          wow and flutter. And, a file generated from the VIP output should be
                          clean as a whistle when imported into an audio application.

                          There are a few VIP audio clips in the Yahoo RCAVIP group files section.

                          73 de Ray
                          K2ULR

                          On Jun 24, 2014, at 9:26 PM, Dave McGuire Mcguire@...
                          [midatlanticretro] wrote:
                          >
                          > I agree with the recommendation of an uncompressed file format. At
                          > the very least, a non-lossy compression is called for here.
                          >
                          > -Dave
                          >
                          > --
                          > Dave McGuire, AK4HZ/3
                          > New Kensington, PA
                          >
                        • joshbensadon
                          Right, RCAVIP group! I forgot about them, I ll go check out what they have! Thanks Ray!
                          Message 12 of 13 , Jun 24, 2014
                            Right, RCAVIP group! I forgot about them, I'll go check out what they have!  Thanks Ray!
                          • joshbensadon
                            Hi Guys... Found 1 of the problems.... Looking through the source code for the VIP tape read/write.... I find the VIP writes to tape an 8 second leader of 0 s.
                            Message 13 of 13 , Jun 25, 2014
                              Hi Guys...

                              Found 1 of the problems....

                              Looking through the source code for the VIP tape read/write.... I find the VIP writes to tape an 8 second leader of 0's.  I thought I would be smart to shorten it a little, well... if you go under 2 seconds, it rejects the data that comes after because it continues to wait for a full 2 seconds of 0's before accepting data.  I have about 2 seconds of leader... but perhaps the voltage through the coupler hasn't "settled" yet.  I'll have to re-record this tape from the beginning.  Lesson learned, keep the whole leader!

                              Testing the input circuit, it works nicely with input signals as low as 3vp-p...  which is now confusing me, why didn't it work with my laptop?  It's late, I'll hook up the laptop tomorrow and scope it.

                              Cheers,
                              Josh
                            Your message has been successfully submitted and would be delivered to recipients shortly.