• Decades-Old DIY Sampler Mystery Solved

    From Lawrence D?Oliveiro@3:633/10 to All on Mon Nov 10 00:58:04 2025
    How?s this for a ?cold case? <https://www.tomshardware.com/pc-components/sound-cards/diy-amiga-sound-sampler-circuit-mystery-solved-32-years-later-magazine-instructions-had-key-component-spec-typos>:
    an issue of ?CU Amiga? magazine from 1993 had a project to let you
    build your own audio sampler -- only it never worked.

    One of those hackers who tried, and failed, to follow the project at
    the time recently revisited it, and delved deeper into trying to
    figure out why the completed project did nothing but crash his
    machine. He found a couple of errors in capacitor values, particularly
    one that threw a clock frequency way off -- instead of 30kHz, it was
    putting out only 287Hz.

    After fixing these problems, he finally had his working sampler!

    Worth noting that one of those typos was pointed out in the subsequent
    magazine issue, but it?s not clear that the other one was ever
    corrected. Without that fix, it?s hard to see how anyone could have successfully built the project. Did all those who failed simply chalk
    it up to their own (assumed) lack of ability?

    Imagine all those young, budding electronics engineers who were put
    off that career for life, because they couldn?t figure out how to get
    this thing working ...

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Theo@3:633/10 to All on Tue Nov 11 16:59:41 2025
    Lawrence D?Oliveiro <ldo@nz.invalid> wrote:
    How?s this for a ?cold case? <https://www.tomshardware.com/pc-components/sound-cards/diy-amiga-sound-sampler-circuit-mystery-solved-32-years-later-magazine-instructions-had-key-component-spec-typos>:
    an issue of ?CU Amiga? magazine from 1993 had a project to let you
    build your own audio sampler -- only it never worked.

    One of those hackers who tried, and failed, to follow the project at
    the time recently revisited it, and delved deeper into trying to
    figure out why the completed project did nothing but crash his
    machine. He found a couple of errors in capacitor values, particularly
    one that threw a clock frequency way off -- instead of 30kHz, it was
    putting out only 287Hz.

    After fixing these problems, he finally had his working sampler!

    Worth noting that one of those typos was pointed out in the subsequent magazine issue, but it?s not clear that the other one was ever
    corrected. Without that fix, it?s hard to see how anyone could have successfully built the project. Did all those who failed simply chalk
    it up to their own (assumed) lack of ability?

    The first typo was pretty obvious - 7uF is not a common capacitor value. Interestingly the datasheet has it (called Cref) as 4.7uF not 47uF as in the correction. I don't think it would have made a great difference - it's just
    a decoupling cap.

    For the clock, it appears to be set by C2 which looks like it should have
    been 470pF not 470nF.

    For C3 the datasheet says "[C1 it calls it] is chosen according
    to the clock frequency so that droop of the capacitor voltage is
    not significant during a conversion." but doesn't give any clues how to calculate it. At a guess 470nF is more likely than 470pF.

    Imagine all those young, budding electronics engineers who were put
    off that career for life, because they couldn?t figure out how to get
    this thing working ...

    I suspect the number of people who made circuits out of computer magazines
    was relatively small [I did!]. Probably more for dedicated electronics magazines which were all about publishing projects.

    Theo

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Theo@3:633/10 to All on Tue Nov 11 22:08:41 2025
    rbowman <bowman@montana.com> wrote:
    On 11 Nov 2025 16:59:41 +0000 (GMT), Theo wrote:

    I suspect the number of people who made circuits out of computer
    magazines was relatively small [I did!]. Probably more for dedicated electronics magazines which were all about publishing projects.

    I think that's true for wood working, needle crafts, boat building, and so forth. It's sort of a vicarious feeling of satisfaction as people think 'Yeah, I could do that!'.

    That and an interesting puzzle in the vein of 'C4 and R8 form a high pass filter into the base of Q3, which is biased via R7 and R9. Q5 ....'
    It's nice to try to understand how a circuit works even if you'll never
    build it, or have no use for the intended application.

    (there were lots of projects like 'DIY electronic ignition for your car' and
    I do wonder how many people actually built them in order to mod their cars)

    I'm just as bad. i see MCU projects in Make magazine and websites that I could do. Sometimes I even have all the components on hand but soemhow I never get roundtoit.

    At that time I did build things just to build them, without actually having much use for the end result. I tend to be a bit more circumspect about
    needing something nowadays.

    Theo

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Wed Nov 12 10:45:48 2025
    On 2025-11-11 23:08, Theo wrote:
    rbowman <bowman@montana.com> wrote:
    On 11 Nov 2025 16:59:41 +0000 (GMT), Theo wrote:

    I suspect the number of people who made circuits out of computer
    magazines was relatively small [I did!]. Probably more for dedicated
    electronics magazines which were all about publishing projects.

    I think that's true for wood working, needle crafts, boat building, and so >> forth. It's sort of a vicarious feeling of satisfaction as people think
    'Yeah, I could do that!'.

    That and an interesting puzzle in the vein of 'C4 and R8 form a high pass filter into the base of Q3, which is biased via R7 and R9. Q5 ....'
    It's nice to try to understand how a circuit works even if you'll never
    build it, or have no use for the intended application.

    (there were lots of projects like 'DIY electronic ignition for your car' and I do wonder how many people actually built them in order to mod their cars)

    I did, I built one such. It lasted something like 60 Km. But it was
    designed so that it was easy to remove and have the car as originally,
    so I just did that on the road and continued the trip. I supposed the
    back EMF from the car coil burned the electronics.

    I was very curious to find how modern cars solved the problem. A low
    voltage electronics, and one little coil per cylinder.


    I'm just as bad. i see MCU projects in Make magazine and websites that I
    could do. Sometimes I even have all the components on hand but soemhow I
    never get roundtoit.

    At that time I did build things just to build them, without actually having much use for the end result. I tend to be a bit more circumspect about needing something nowadays.

    Theo


    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Wed Nov 12 10:46:47 2025
    On 2025-11-10 01:58, Lawrence D?Oliveiro wrote:
    How?s this for a ?cold case? <https://www.tomshardware.com/pc-components/sound-cards/diy-amiga-sound-sampler-circuit-mystery-solved-32-years-later-magazine-instructions-had-key-component-spec-typos>:
    an issue of ?CU Amiga? magazine from 1993 had a project to let you
    build your own audio sampler -- only it never worked.

    One of those hackers who tried, and failed, to follow the project at
    the time recently revisited it, and delved deeper into trying to
    figure out why the completed project did nothing but crash his
    machine. He found a couple of errors in capacitor values, particularly
    one that threw a clock frequency way off -- instead of 30kHz, it was
    putting out only 287Hz.

    After fixing these problems, he finally had his working sampler!

    Worth noting that one of those typos was pointed out in the subsequent magazine issue, but it?s not clear that the other one was ever
    corrected. Without that fix, it?s hard to see how anyone could have successfully built the project. Did all those who failed simply chalk
    it up to their own (assumed) lack of ability?

    Imagine all those young, budding electronics engineers who were put
    off that career for life, because they couldn?t figure out how to get
    this thing working ...

    I wonder how many projects I made failed because of printing typos :-/

    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Wed Nov 12 10:52:01 2025
    On 2025-11-12 03:51, rbowman wrote:
    On 11 Nov 2025 22:08:41 +0000 (GMT), Theo wrote:

    (there were lots of projects like 'DIY electronic ignition for your car'
    and I do wonder how many people actually built them in order to mod
    their cars)

    I did :) The project specified a particular ferrite core that was unobtanium. A friend knew a guy who knew a guy... At the time IBM spun
    off a lot of little cottage industries to handle stuff that they didn't
    want in house, in some cases because of paranoia about being a monopoly. Anyway the guy in question was operating his little ferrite business out
    of his garage. I drove down to his place in the outskirts of Kingston. As
    i recall he gave me the cores as a sample to avoid paperwork. And I was on
    my way to a CDI!

    I've looked at enough of those projects that when I see some strange component or parts with values that aren't common I realize those were whatever the author had laying on his workbench at the time rather than
    the result of careful design.

    I still have a collection of ferrite toroids, air variable caps, and so
    forth from when I was messing around with radios. You can bet if I design something it will be using them. Good luck finding the exact part in 2025.

    Many coils and variable capacitors at the time were unobtanium :-(

    Several projects I actually built did not have the exact pieces. Some
    worked some not. Like one amplifier that required a couple of matched
    push pull transistors. Each transistor I got was a different type of equivalent in different packaging. One plastic, the other metallic.

    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Theo@3:633/10 to All on Wed Nov 12 13:14:07 2025
    Carlos E.R. <robin_listas@es.invalid> wrote:
    Many coils and variable capacitors at the time were unobtanium :-(

    Several projects I actually built did not have the exact pieces. Some
    worked some not. Like one amplifier that required a couple of matched
    push pull transistors. Each transistor I got was a different type of equivalent in different packaging. One plastic, the other metallic.

    I used to avoid projects with inductors because they were much harder to get hold of than resistors or capacitors. Often the magazine would say 'we arranged with XYZ supplier to stock the Coilcraft 1234-56789' but you just
    knew that would be trouble, especially if it was a past issue and the
    supplier had probably run out. Winding your own was similarly risky - there was usually a ferrite core that was similarly unobtainable.

    It was very 'cargo cult' - you need XXXX-XXXX part number inductor/core/transformer, instead of saying 'you need an inductor of XX uH that supports YY amps up to ZZ kHz' at which point you might have been able
    to find a replacement by specs locally.

    Theo

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From John Ames@3:633/10 to All on Wed Nov 12 11:00:02 2025
    On 12 Nov 2025 18:40:09 GMT
    rbowman <bowman@montana.com> wrote:

    I wonder how many projects I made failed because of printing typos
    :-/

    Those BASIC programs printed in magazines that required PUSHing half
    a page of hex values must have been fun. Years ago I had a TracFone
    where actuation required entering 24, iirc, hex values. Great fun.

    Compute's Gazette actually developed a checksum system for those pages
    upon pages of DATA statements so's readers could catch and correct
    typing errors on a line-by-line basis o_O


    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Scott Alfter@3:633/10 to All on Wed Nov 12 23:22:36 2025
    In article <20251112110002.00004ddd@gmail.com>,
    John Ames <commodorejohn@gmail.com> wrote:
    On 12 Nov 2025 18:40:09 GMT
    rbowman <bowman@montana.com> wrote:

    I wonder how many projects I made failed because of printing typos
    :-/

    Those BASIC programs printed in magazines that required PUSHing half
    a page of hex values must have been fun. Years ago I had a TracFone
    where actuation required entering 24, iirc, hex values. Great fun.

    Compute's Gazette actually developed a checksum system for those pages
    upon pages of DATA statements so's readers could catch and correct
    typing errors on a line-by-line basis o_O

    Nibble had a couple of checksum systems to cover both BASIC programs and
    binary blobs. They sold their first one, but the second one was offered as both a paid program that would dump checksums for a program and a free
    version, small enough to include in every issue, that would check each line
    as you typed it in.

    I took that free version and hacked it into something that would dump
    checksums for a program you'd already typed in. If I remember right, I
    tried submitting that back to them as a suggested improvement, but they
    weren't interested for some reason. :)

    Fun fact: a dump of checksums from the older of these systems shows up in
    The Terminator as part of the T-800's HUD-like display:

    https://www.youtube.com/watch?v=v0vbTHd04gE

    I found this explanation, which tracked down the particular issues of Nibble with code featured in the movie:

    https://www.theterminatorfans.com/the-terminator-vision-hud-source-code-explained/

    --
    _/_
    / v \ Scott Alfter (remove the obvious to send mail)
    (IIGS( https://alfter.us/ Top-posting!
    \_^_/ >What's the most annoying thing on Usenet?

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Thu Nov 13 18:47:46 2025
    On 2025-11-12 14:14, Theo wrote:
    Carlos E.R. <robin_listas@es.invalid> wrote:
    Many coils and variable capacitors at the time were unobtanium :-(

    Several projects I actually built did not have the exact pieces. Some
    worked some not. Like one amplifier that required a couple of matched
    push pull transistors. Each transistor I got was a different type of
    equivalent in different packaging. One plastic, the other metallic.

    I used to avoid projects with inductors because they were much harder to get hold of than resistors or capacitors. Often the magazine would say 'we arranged with XYZ supplier to stock the Coilcraft 1234-56789' but you just knew that would be trouble, especially if it was a past issue and the supplier had probably run out. Winding your own was similarly risky - there was usually a ferrite core that was similarly unobtainable.

    Yes. There was a local shop that had the copper, but ferrite cores were
    almost impossible.


    It was very 'cargo cult' - you need XXXX-XXXX part number inductor/core/transformer, instead of saying 'you need an inductor of XX uH that supports YY amps up to ZZ kHz' at which point you might have been able to find a replacement by specs locally.

    Maybe they did not know the specs themselves.

    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Thu Nov 13 18:52:05 2025
    On 2025-11-12 19:40, rbowman wrote:
    On Wed, 12 Nov 2025 10:46:47 +0100, Carlos E.R. wrote:

    I wonder how many projects I made failed because of printing typos :-/

    Those BASIC programs printed in magazines that required PUSHing half a
    page of hex values must have been fun. Years ago I had a TracFone where actuation required entering 24, iirc, hex values. Great fun.

    I think I did two of those, but they also came in ASM and I used that
    instead. I created some typo of my own. The one I remember was ted.com,
    a tiny text editor of 3 KB that would fit easily in floppies.

    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Thu Nov 13 19:03:06 2025
    On 2025-11-13 00:22, Scott Alfter wrote:
    In article <20251112110002.00004ddd@gmail.com>,
    John Ames <commodorejohn@gmail.com> wrote:
    On 12 Nov 2025 18:40:09 GMT
    rbowman <bowman@montana.com> wrote:

    I wonder how many projects I made failed because of printing typos
    :-/

    Those BASIC programs printed in magazines that required PUSHing half
    a page of hex values must have been fun. Years ago I had a TracFone
    where actuation required entering 24, iirc, hex values. Great fun.

    Compute's Gazette actually developed a checksum system for those pages
    upon pages of DATA statements so's readers could catch and correct
    typing errors on a line-by-line basis o_O

    Nibble had a couple of checksum systems to cover both BASIC programs and binary blobs. They sold their first one, but the second one was offered as both a paid program that would dump checksums for a program and a free version, small enough to include in every issue, that would check each line as you typed it in.

    I took that free version and hacked it into something that would dump checksums for a program you'd already typed in. If I remember right, I
    tried submitting that back to them as a suggested improvement, but they weren't interested for some reason. :)

    Fun fact: a dump of checksums from the older of these systems shows up in
    The Terminator as part of the T-800's HUD-like display:

    https://www.youtube.com/watch?v=v0vbTHd04gE

    I found this explanation, which tracked down the particular issues of Nibble with code featured in the movie:

    https://www.theterminatorfans.com/the-terminator-vision-hud-source-code-explained/


    Cute.

    I should not read these things, though, they ruin the magic :-D

    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From songbird@3:633/10 to All on Fri Nov 14 23:43:45 2025
    Lawrence D?Oliveiro wrote:
    ...
    Worth noting that one of those typos was pointed out in the subsequent magazine issue, but it?s not clear that the other one was ever
    corrected. Without that fix, it?s hard to see how anyone could have successfully built the project. Did all those who failed simply chalk
    it up to their own (assumed) lack of ability?

    Imagine all those young, budding electronics engineers who were put
    off that career for life, because they couldn?t figure out how to get
    this thing working ...

    amazingly i built a model railroad control system back in the
    mid-to later '70s from a series of articles in a magazine and it
    worked.

    there were step-by-step instructions and measurements you took
    at certain points to verify things worked. i had it up and
    running before i went away to college but the house it was all
    set up in was sold when i was away at college so i ended up with
    boxes of model railroad, parts for the engines, all the electronics
    and things all packed together in a rather jumbled mess. i don't
    know if the guy i sold it all to ever used it and some years later
    after he passed away his wife asked me to sell it for her and i
    had my brother do it so i never set it all up again.

    as a description is that it was a fixed voltage on the rails so
    you did not need to break the circuit into different parts and the
    computer imposed a signal on that voltage that was picked up by
    receivers in each engine you had going. so you needed a controller
    for each engine. at most i had three engines running at once but
    i think the overall capacity was more than that (enough that i
    doubt i'd have been able to run them all due to current limitations.


    songbird

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Carlos E.R.@3:633/10 to All on Sat Nov 15 14:54:34 2025
    On 2025-11-15 05:43, songbird wrote:
    Lawrence D?Oliveiro wrote:
    ...
    Worth noting that one of those typos was pointed out in the subsequent
    magazine issue, but it?s not clear that the other one was ever
    corrected. Without that fix, it?s hard to see how anyone could have
    successfully built the project. Did all those who failed simply chalk
    it up to their own (assumed) lack of ability?

    Imagine all those young, budding electronics engineers who were put
    off that career for life, because they couldn?t figure out how to get
    this thing working ...

    amazingly i built a model railroad control system back in the
    mid-to later '70s from a series of articles in a magazine and it
    worked.

    there were step-by-step instructions and measurements you took
    at certain points to verify things worked. i had it up and
    running before i went away to college but the house it was all
    set up in was sold when i was away at college so i ended up with
    boxes of model railroad, parts for the engines, all the electronics
    and things all packed together in a rather jumbled mess. i don't
    know if the guy i sold it all to ever used it and some years later
    after he passed away his wife asked me to sell it for her and i
    had my brother do it so i never set it all up again.

    as a description is that it was a fixed voltage on the rails so
    you did not need to break the circuit into different parts and the
    computer imposed a signal on that voltage that was picked up by
    receivers in each engine you had going. so you needed a controller
    for each engine. at most i had three engines running at once but
    i think the overall capacity was more than that (enough that i
    doubt i'd have been able to run them all due to current limitations.

    Nice!


    --
    Cheers, Carlos.
    ES??, EU??;

    --- PyGate Linux v1.5
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)