• squeezing a field

    From john larkin@21:1/5 to All on Wed Oct 23 10:22:31 2024
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good
    cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Phil Hobbs@21:1/5 to john larkin on Wed Oct 23 18:14:57 2024
    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.




    Not a whole lot. In the near-field region, B obeys Laplace’s equation,
    which means among other things that the field falls off on the length scale
    of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs Principal Consultant
    ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics,
    Electro-optics, Photonics, Analog Electronics

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to pcdhSpamMeSenseless@electrooptical. on Wed Oct 23 11:33:23 2024
    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good
    cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133 >>
    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.




    Not a whole lot. In the near-field region, B obeys Laplace’s equation,
    which means among other things that the field falls off on the length scale >of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to john larkin on Thu Oct 24 21:07:06 2024
    On 24/10/2024 5:33 am, john larkin wrote:
    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good
    cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.




    Not a whole lot. In the near-field region, B obeys Laplace’s equation,
    which means among other things that the field falls off on the length scale >> of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    Somebody who knew what they were doing could model it in LTSpice. The
    adjacent metal-work is a poorly coupled shorted turn. Model your
    inductor as 47uH coil with 0.135 series resistance and 4pF of parallel capacitance, and model the metal as a coupled - perhaps 1nH single turn
    - with perhaps 1% coupling and maybe a milliohm of resistance. The
    difference between your two test situations is going to be mainly the
    coupling you get - though the loop resistance of the shorted turn is
    going to be lower in the second case.

    Assuming that it will work might be a touch optimistic.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From legg@21:1/5 to All on Thu Oct 24 10:37:56 2024
    On Wed, 23 Oct 2024 11:33:23 -0700, john larkin <jl@glen--canyon.com>
    wrote:

    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs ><pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good
    cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.




    Not a whole lot. In the near-field region, B obeys Laplace’s equation, >>which means among other things that the field falls off on the length scale >>of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a >multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    Depending on how closely packed the parts were, you might
    gat adjacent parts acting as pole pieces. So space >> than
    body hight.

    Assuming also that your metalwork is Al and not Fe.

    RL

    RL

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Lasse Langwadt@21:1/5 to Bill Sloman on Thu Oct 24 18:45:20 2024
    On 10/24/24 12:07, Bill Sloman wrote:
    On 24/10/2024 5:33 am, john larkin wrote:
    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs
    <pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the >>>> bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good >>>> cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.




    Not a whole lot.  In the near-field region, B obeys Laplace’s equation, >>> which means among other things that the field falls off on the length
    scale
    of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a
    multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    Somebody who knew what they were doing could model it in LTSpice. The adjacent metal-work is a poorly coupled shorted turn. Model your
    inductor as 47uH coil with 0.135 series resistance and 4pF of parallel capacitance, and model the metal as a coupled - perhaps 1nH single turn
    - with perhaps 1% coupling and maybe a milliohm of resistance.

    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to All on Thu Oct 24 10:42:34 2024
    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:

    On 10/24/24 12:07, Bill Sloman wrote:
    On 24/10/2024 5:33 am, john larkin wrote:
    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs
    <pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the >>>>> bottom of the multilayer board. There's a 0.2" gap between the bottom >>>>> of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good >>>>> cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder >>>>> what that will do to its electrical behavior.




    Not a whole lot.  In the near-field region, B obeys Laplace’s equation, >>>> which means among other things that the field falls off on the length
    scale
    of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a
    multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    Somebody who knew what they were doing could model it in LTSpice. The
    adjacent metal-work is a poorly coupled shorted turn. Model your
    inductor as 47uH coil with 0.135 series resistance and 4pF of parallel
    capacitance, and model the metal as a coupled - perhaps 1nH single turn
    - with perhaps 1% coupling and maybe a milliohm of resistance.

    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?





    It is for people who don't actually work with real parts.

    I wouldn't know how to use Spice to model mag field coupling into PCB
    copper planes, or into a big hunk of aluminum. And it's good to get
    away from a screen and do something real now and then. Drill holes,
    dremel, solder, measure, things like that.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to legg on Thu Oct 24 10:50:43 2024
    On Thu, 24 Oct 2024 10:37:56 -0400, legg <legg@nospam.magma.ca> wrote:

    On Wed, 23 Oct 2024 11:33:23 -0700, john larkin <jl@glen--canyon.com>
    wrote:

    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs >><pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the >>>> bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good >>>> cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.




    Not a whole lot. In the near-field region, B obeys Laplace’s equation, >>>which means among other things that the field falls off on the length scale >>>of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a >>multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    Depending on how closely packed the parts were, you might
    gat adjacent parts acting as pole pieces. So space >> than
    body hight.

    Assuming also that your metalwork is Al and not Fe.

    RL

    RL

    These products will usually be a 6-layer PCB in an extruded aluminum
    box with a fat bottom flange. My perferred inductor just fits on the
    bottom of the board, assuming the solder spaces it off the board maybe
    5 mils.

    https://www.dropbox.com/scl/fi/7tsg530psa2dkcdoarh8o/B-box_flange_3D_crop.jpg?rlkey=tne2fmmnw26s7xgivbloeqnu8&raw=1

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to john larkin on Thu Oct 24 21:56:59 2024
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use analogue computing to work out his designs i.e. He built prototypes and measured them.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to Liz Tuddenham on Thu Oct 24 16:23:50 2024
    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >analogue computing to work out his designs i.e. He built prototypes and >measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to Lasse Langwadt on Fri Oct 25 15:50:38 2024
    On 25/10/2024 3:45 am, Lasse Langwadt wrote:
    On 10/24/24 12:07, Bill Sloman wrote:
    On 24/10/2024 5:33 am, john larkin wrote:
    On Wed, 23 Oct 2024 18:14:57 -0000 (UTC), Phil Hobbs
    <pcdhSpamMeSenseless@electrooptical.net> wrote:

    john larkin <jl@glen--canyon.com> wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the >>>>> bottom of the multilayer board. There's a 0.2" gap between the bottom >>>>> of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good >>>>> cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder >>>>> what that will do to its electrical behavior.




    Not a whole lot.  In the near-field region, B obeys Laplace’s equation, >>>> which means among other things that the field falls off on the
    length scale
    of the gap, not of the whole inductor.

    Cheers

    Phil Hobbs

    You are right. A similar part is 47 uH in free air, 44.6 mounted on a
    multilayer board, and 42.1 squeezed between the board and a big chunk
    of aluminum.

    So it will work.

    Somebody who knew what they were doing could model it in LTSpice. The
    adjacent metal-work is a poorly coupled shorted turn. Model your
    inductor as 47uH coil with 0.135 series resistance and 4pF of parallel
    capacitance, and model the metal as a coupled - perhaps 1nH single
    turn - with perhaps 1% coupling and maybe a milliohm of resistance.

    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?

    Obviously not. My point was that John Larkin could use LTSpice to see
    what his numbers actually meant.

    Putting a chunk of metal next to a drum core inductor doesn't change the
    drum core inductor in any way - it just adds a shorted turn to the
    assembly, and messes up what a dumb inductance measuring device sees.

    In this case, that dumb measuring device is John Larkin, who doesn't
    want to think about what is actually going on, as opposed to his
    inductance meter, which can't.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to Liz Tuddenham on Fri Oct 25 16:08:08 2024
    On 25/10/2024 7:56 am, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?

    It is for people who don't actually work with real parts.

    Actually, it is extremely useful for people who work with real parts and
    who want to know exactly what is going on. You can see stuff that is
    very hard to measure on real parts.

    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use analogue computing to work out his designs i.e. He built prototypes and measured them.

    But he understood what he was measuring - a least most of the time.
    His revolutionary ideas about capacitor microphones and the patent
    application he made fell down when he found out about the Philips
    capacitative pressure gauges which had been exploiting the same
    principle for about a decade before.

    This played out in the pages of Wireless World, between the first and
    second parts of a two part article. I don't think he mentioned the
    Philips pressure gauges, but I've got a 1954 reference to them in my
    1970 Ph.D. thesis. They might not have been the prior art that he found,
    but they would have served.

    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator
    paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to john larkin on Fri Oct 25 09:37:19 2024
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >analogue computing to work out his designs i.e. He built prototypes and >measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between resistance, capacitance, frequency and time constant (put in two and the
    others appear, put in three and the error% appears). It also gives dB
    loss below or above the 'cutoff' frequency. Some years ago I also made
    some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for
    design work.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to Bill Sloman on Fri Oct 25 09:37:20 2024
    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:56 am, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?

    It is for people who don't actually work with real parts.

    Actually, it is extremely useful for people who work with real parts and
    who want to know exactly what is going on. You can see stuff that is
    very hard to measure on real parts.

    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use analogue computing to work out his designs i.e. He built prototypes and measured them.

    But he understood what he was measuring - a least most of the time.
    His revolutionary ideas about capacitor microphones and the patent application he made fell down when he found out about the Philips capacitative pressure gauges which had been exploiting the same
    principle for about a decade before.

    If anything, that enhances his reputation rather than diminishing it, A
    single inventor working on a project in his spare time compared with a laboratory-full of specialists working full-time on commercial projects
    (if I remember rightly) one of which was almost certainly partly
    financed by the Ford Motor Company.


    This played out in the pages of Wireless World, between the first and
    second parts of a two part article.

    The articles were in WW Nov/Dec 1963. At the end he refers to two Dutch papers:

    Philips Technical Review Vol9 Nr12 1947/48 pp357-363

    Omroep-technische Mededelingen Feb15 1961

    These are both describing to low-noise condenser microphones but he
    points out that they don't have some of the desirable features of his
    design.


    I don't think he mentioned the Philips pressure gauges,
    but I've got a 1954 reference to them in my
    1970 Ph.D. thesis

    That may have been the article on capacitive pressure guages for car
    engines in the Philips Technical Review. Their main problem was that
    the temperatures and pressures they were trying to measure gave a short diaphragm life if the diaphragm was thin enough to respond to the
    required frequency range with sufficient sensitivity.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to Liz Tuddenham on Fri Oct 25 23:42:52 2024
    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >>> analogue computing to work out his designs i.e. He built prototypes and >>> measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between resistance, capacitance, frequency and time constant (put in two and the others appear, put in three and the error% appears). It also gives dB
    loss below or above the 'cutoff' frequency. Some years ago I also made
    some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for
    design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to Liz Tuddenham on Fri Oct 25 23:39:04 2024
    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:56 am, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?

    It is for people who don't actually work with real parts.

    Actually, it is extremely useful for people who work with real parts and
    who want to know exactly what is going on. You can see stuff that is
    very hard to measure on real parts.

    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >>> analogue computing to work out his designs i.e. He built prototypes and >>> measured them.

    But he understood what he was measuring - a least most of the time.
    His revolutionary ideas about capacitor microphones and the patent
    application he made fell down when he found out about the Philips
    capacitative pressure gauges which had been exploiting the same
    principle for about a decade before.

    If anything, that enhances his reputation rather than diminishing it, A single inventor working on a project in his spare time compared with a laboratory-full of specialists working full-time on commercial projects
    (if I remember rightly) one of which was almost certainly partly
    financed by the Ford Motor Company.

    The Philip's capacitative pressure gauges were scientific instruments,
    not mass market products, and if I could find them as graduate student, somebody whose day job was at the Royal Radar Establishment in Malvern
    should have been able to do that too. My direct boss at George Kent in
    Luton - Colin Hunter - had done his electronic appenticeship at the RRE
    under Peter Baxandall. He wasn't any kind of single inventor working on
    stuff in his spare time, but rather a full time expert exploiting the
    resources his job gave him to follow his interests outside of his day job.

    This played out in the pages of Wireless World, between the first and
    second parts of a two part article.

    The articles were in WW Nov/Dec 1963. At the end he refers to two Dutch papers:

    Philips Technical Review Vol9 Nr12 1947/48 pp357-363

    Omroep-technische Mededelingen Feb15 1961

    These are both describing to low-noise condenser microphones but he
    points out that they don't have some of the desirable features of his
    design.

    Very likely. 1947/48 was pre-transistor. The planar process was invented
    around 1955, and Fairchild started selling cheap planar transistors
    around 1959. They revolutionised circuit design,and I latched onto that
    in 1965 as a graduate student in chemistry. By 1963 you could suddenly
    do a lot with planar transistors which hadn't been practical with
    earlier parts.

    I don't think he mentioned the Philips pressure gauges,
    >but I've got a 1954 reference to them in my
    1970 Ph.D. thesis

    That may have been the article on capacitive pressure gauges for car
    engines in the Philips Technical Review.

    It wasn't. The references in my Ph.D. thesis are to J.J.Opstelten and
    N. Warmholtz, App.Sci.Res.Hague B4 page 329 (1955)
    and J.J.Opstelten, N. Warmholtz and J.J. Zaalberg van Zelst ibid B6 page
    129 (1956).

    There may have been work on a product for the car market, but they
    weren't cheap enough for that.

    Their main problem was that
    the temperatures and pressures they were trying to measure gave a short diaphragm life if the diaphragm was thin enough to respond to the
    required frequency range with sufficient sensitivity.

    Etched silicon diaphragms with strain gauge sensors in the diaphragm was
    the mass market product, and they came a lot later.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to Liz Tuddenham on Fri Oct 25 08:25:56 2024
    On Fri, 25 Oct 2024 09:37:19 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use
    analogue computing to work out his designs i.e. He built prototypes and
    measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between >resistance, capacitance, frequency and time constant (put in two and the >others appear, put in three and the error% appears). It also gives dB
    loss below or above the 'cutoff' frequency. Some years ago I also made
    some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for
    design work.

    I wrote a PowerBasic program to find resistors in stock that can hit a
    target divider ratio and Thevenin impedance. That's in PowerBasic.

    It saves a lot of time and helps minimize the number of BOM items.

    I love Basic, but the code snobs hate it.

    I wrote some other apps, like resonance calculators, filter design,
    unit conversions. I could post a link.

    I also have scribbles about what can be done with one quad resistor
    pack.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to Bill Sloman on Fri Oct 25 17:32:21 2024
    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >>> analogue computing to work out his designs i.e. He built prototypes and >>> measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between resistance, capacitance, frequency and time constant (put in two and the others appear, put in three and the error% appears). It also gives dB loss below or above the 'cutoff' frequency. Some years ago I also made some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    I don't understand why you use the word 'admit'. I make my own tools to
    meet my particular requirements, there is nothing shameful about that.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to Liz Tuddenham on Fri Oct 25 14:59:01 2024
    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >> >>> analogue computing to work out his designs i.e. He built prototypes and >> >>> measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between
    resistance, capacitance, frequency and time constant (put in two and the >> > others appear, put in three and the error% appears). It also gives dB
    loss below or above the 'cutoff' frequency. Some years ago I also made
    some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for
    design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    I don't understand why you use the word 'admit'. I make my own tools to
    meet my particular requirements, there is nothing shameful about that.

    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.


    Have you used LT Spice? It's easy to learn and is great fun. I(t makes
    time for snacking and napping too.

    I even use it instead of a calculator for simple stuff like voltage
    dividers and RC time constants and things.

    I can fiddle LC filters to at least third order. Sometimes 5th order,
    until things diverge and explode.

    LT Spice schematics can be screen-shot for things like block diagram
    figures in manuals too.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to Liz Tuddenham on Sat Oct 26 12:13:54 2024
    On 26/10/2024 3:32 am, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >>>>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >>>>> analogue computing to work out his designs i.e. He built prototypes and >>>>> measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But >>>> it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between
    resistance, capacitance, frequency and time constant (put in two and the >>> others appear, put in three and the error% appears). It also gives dB
    loss below or above the 'cutoff' frequency. Some years ago I also made
    some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for
    design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    I don't understand why you use the word 'admit'. I make my own tools to
    meet my particular requirements, there is nothing shameful about that.

    There are more powerful tools out there. Audio doesn't necessarily need
    them, but John Larkin operates in a wider market, and there are tools he
    could use if he could be bothered to master them. His somewhat selective approach to the stuff he could have studied at Tulane might mean that
    he'd have to do quite a lot of work to master them.

    I had to write my own multi-parameter non-linear least squares curve
    fitting program (in fortran 4) when I was an undergraduate. Since then
    I've used an off-the-shelf program when I need to do that kind of job.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to john larkin on Sat Oct 26 12:17:26 2024
    On 26/10/2024 8:59 am, john larkin wrote:
    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid >>>>> (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk> >>>>>>> wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >>>>>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use >>>>>> analogue computing to work out his designs i.e. He built prototypes and >>>>>> measured them.

    You youngsters probably don't remember a time when there wasn't Spice. >>>>
    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first >>>>> PC sim program was Tatum labs ECA, which required a typed netlist. But >>>>> it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between >>>> resistance, capacitance, frequency and time constant (put in two and the >>>> others appear, put in three and the error% appears). It also gives dB >>>> loss below or above the 'cutoff' frequency. Some years ago I also made >>>> some lookup tables for combinations of 5% tolerance resistors in series >>>> and parallel.

    Those and a pocket calculator are still the only 'computing' I use for >>>> design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    I don't understand why you use the word 'admit'. I make my own tools to
    meet my particular requirements, there is nothing shameful about that.

    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.


    Have you used LT Spice? It's easy to learn and is great fun. It makes
    time for snacking and napping too.

    I even use it instead of a calculator for simple stuff like voltage
    dividers and RC time constants and things.

    I can fiddle LC filters to at least third order. Sometimes 5th order,
    until things diverge and explode.

    But Williams and Taylor would let you do it better, if you could follow
    their advice.

    LT Spice schematics can be screen-shot for things like block diagram
    figures in manuals too.

    There are other, better, schematic editors which you can also use that way.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to john larkin on Sat Oct 26 10:35:31 2024
    john larkin <jl@glen--canyon.com> wrote:

    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >> >>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed
    to use analogue computing to work out his designs i.e. He built
    prototypes and measured them.

    You youngsters probably don't remember a time when there wasn't
    Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist.
    But it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships
    between resistance, capacitance, frequency and time constant (put in
    two and the others appear, put in three and the error% appears). It
    also gives dB loss below or above the 'cutoff' frequency. Some years
    ago I also made some lookup tables for combinations of 5% tolerance
    resistors in series and parallel.

    Those and a pocket calculator are still the only 'computing' I use
    for design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    I don't understand why you use the word 'admit'. I make my own tools to >meet my particular requirements, there is nothing shameful about that.

    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.

    My reply to that had anatomical connotations. :-)


    Have you used LT Spice? It's easy to learn and is great fun. I(t makes
    time for snacking and napping too.

    I don't think they ever did a version that would run on a Mac G3 (OS
    8.6), which is my main workhorse.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Phil Hobbs@21:1/5 to Liz Tuddenham on Sat Oct 26 11:27:49 2024
    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid >>>>>> (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk> >>>>>>>> wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >>>>>>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed >>>>>>> to use analogue computing to work out his designs i.e. He built >>>>>>> prototypes and measured them.

    You youngsters probably don't remember a time when there wasn't
    Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first >>>>>> PC sim program was Tatum labs ECA, which required a typed netlist. >>>>>> But it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships
    between resistance, capacitance, frequency and time constant (put in >>>>> two and the others appear, put in three and the error% appears). It >>>>> also gives dB loss below or above the 'cutoff' frequency. Some years >>>>> ago I also made some lookup tables for combinations of 5% tolerance
    resistors in series and parallel.

    Those and a pocket calculator are still the only 'computing' I use
    for design work.

    No surprise there, though I am a bit surprised that you would admit it >>>> in public.

    I don't understand why you use the word 'admit'. I make my own tools to >>> meet my particular requirements, there is nothing shameful about that.

    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.

    My reply to that had anatomical connotations. :-)


    Have you used LT Spice? It's easy to learn and is great fun. I(t makes
    time for snacking and napping too.

    I don't think they ever did a version that would run on a Mac G3 (OS
    8.6), which is my main workhorse.



    I’ve been running LTspice using Wine on Linux for 15 years or so, no problems. Wine runs fine on both x86 and apple silicon, I’m told.

    Cheers

    Phil Hobbs

    --
    Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to Phil Hobbs on Sat Oct 26 12:41:05 2024
    Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid >>>>>> (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk> >>>>>>>> wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >>>>>>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed >>>>>>> to use analogue computing to work out his designs i.e. He built >>>>>>> prototypes and measured them.

    You youngsters probably don't remember a time when there wasn't
    Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first >>>>>> PC sim program was Tatum labs ECA, which required a typed netlist. >>>>>> But it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships
    between resistance, capacitance, frequency and time constant (put in >>>>> two and the others appear, put in three and the error% appears). It >>>>> also gives dB loss below or above the 'cutoff' frequency. Some years >>>>> ago I also made some lookup tables for combinations of 5% tolerance >>>>> resistors in series and parallel.

    Those and a pocket calculator are still the only 'computing' I use >>>>> for design work.

    No surprise there, though I am a bit surprised that you would admit it >>>> in public.

    I don't understand why you use the word 'admit'. I make my own tools to >>> meet my particular requirements, there is nothing shameful about that.

    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.

    My reply to that had anatomical connotations. :-)


    Have you used LT Spice? It's easy to learn and is great fun. I(t makes
    time for snacking and napping too.

    I don't think they ever did a version that would run on a Mac G3 (OS
    8.6), which is my main workhorse.



    I’ve been running LTspice using Wine on Linux for 15 years or so, no problems. Wine runs fine on both x86 and apple silicon, I’m told.

    A G3 won't run Linux or any of the OSX applications.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From legg@21:1/5 to Liz Tuddenham on Sat Oct 26 11:16:03 2024
    On Fri, 25 Oct 2024 09:37:19 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed to use
    analogue computing to work out his designs i.e. He built prototypes and
    measured them.

    You youngsters probably don't remember a time when there wasn't Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first
    PC sim program was Tatum labs ECA, which required a typed netlist. But
    it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships between >resistance, capacitance, frequency and time constant (put in two and the >others appear, put in three and the error% appears). It also gives dB
    loss below or above the 'cutoff' frequency. Some years ago I also made
    some lookup tables for combinations of 5% tolerance resistors in series
    and parallel.

    Those and a pocket calculator are still the only 'computing' I use for
    design work.

    For 5%, I'd just use the ratio of difference and starting value to
    multiply the parallel mate.

    My own manually tabulated book of tables for parallel values assumed
    an E48 or E96 (2 or 1%) rack of possibilities. Otherwise, why bother?

    Took a while to complete, and only aimed at E24-type variations.
    Still around here somewhere, though the dime-store binding is fubar.

    RL
    parts

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From legg@21:1/5 to All on Sat Oct 26 11:20:42 2024
    On Fri, 25 Oct 2024 16:08:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    <snip>
    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator
    paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    What's wrong with 'squegging' ? It's a simple word that covers
    a host of faults that all give the same approximate symptom . .

    RL

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Joe Gwinn@21:1/5 to Liz Tuddenham on Sat Oct 26 14:16:05 2024
    On Sat, 26 Oct 2024 12:41:05 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid >> >>>>>> (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed
    to use analogue computing to work out his designs i.e. He built
    prototypes and measured them.

    You youngsters probably don't remember a time when there wasn't
    Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first >> >>>>>> PC sim program was Tatum labs ECA, which required a typed netlist.
    But it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships
    between resistance, capacitance, frequency and time constant (put in >> >>>>> two and the others appear, put in three and the error% appears). It >> >>>>> also gives dB loss below or above the 'cutoff' frequency. Some years >> >>>>> ago I also made some lookup tables for combinations of 5% tolerance
    resistors in series and parallel.

    Those and a pocket calculator are still the only 'computing' I use
    for design work.

    No surprise there, though I am a bit surprised that you would admit it >> >>>> in public.

    I don't understand why you use the word 'admit'. I make my own tools to >> >>> meet my particular requirements, there is nothing shameful about that. >> >>
    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.

    My reply to that had anatomical connotations. :-)


    Have you used LT Spice? It's easy to learn and is great fun. I(t makes
    time for snacking and napping too.

    I don't think they ever did a version that would run on a Mac G3 (OS
    8.6), which is my main workhorse.



    I’ve been running LTspice using Wine on Linux for 15 years or so, no
    problems. Wine runs fine on both x86 and apple silicon, I’m told.

    A G3 won't run Linux or any of the OSX applications.

    Yeah. Linux will run on anything that will run Windows on Intel.
    Which modern Intel Macs will do. I have such a Mac, and run both
    MacOS and Windows on it. I bet that the new Apple silicon will run
    Windows in emulation as well.

    Joe Gwinn

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to Joe Gwinn on Sat Oct 26 19:19:57 2024
    Joe Gwinn <joegwinn@comcast.net> wrote:

    On Sat, 26 Oct 2024 12:41:05 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:
    [...]

    A G3 won't run Linux or any of the OSX applications.

    Yeah. Linux will run on anything that will run Windows on Intel.
    Which modern Intel Macs will do. I have such a Mac, and run both
    MacOS and Windows on it. I bet that the new Apple silicon will run
    Windows in emulation as well.

    This machine is a G3 running Mac OS 8.6, it is not Intel-based and
    cannot run Windows or Linux or OSX.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to Liz Tuddenham on Sat Oct 26 13:07:56 2024
    On Sat, 26 Oct 2024 10:35:31 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid >> >> >> (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk>
    wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that >> >> >>>>> doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed
    to use analogue computing to work out his designs i.e. He built
    prototypes and measured them.

    You youngsters probably don't remember a time when there wasn't
    Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first >> >> >> PC sim program was Tatum labs ECA, which required a typed netlist.
    But it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships
    between resistance, capacitance, frequency and time constant (put in
    two and the others appear, put in three and the error% appears). It >> >> > also gives dB loss below or above the 'cutoff' frequency. Some years >> >> > ago I also made some lookup tables for combinations of 5% tolerance
    resistors in series and parallel.

    Those and a pocket calculator are still the only 'computing' I use
    for design work.

    No surprise there, though I am a bit surprised that you would admit it
    in public.

    I don't understand why you use the word 'admit'. I make my own tools to
    meet my particular requirements, there is nothing shameful about that.

    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.

    My reply to that had anatomical connotations. :-)


    You refer to his fat head of course.



    Have you used LT Spice? It's easy to learn and is great fun. I(t makes
    time for snacking and napping too.

    I don't think they ever did a version that would run on a Mac G3 (OS
    8.6), which is my main workhorse.

    Pity. Two wonderful parts of my life (after Mo of course) are LT Spice
    and my reverse Polish calculators.

    Get a cheap used Windows laptop, $75 maybe, just to run LT Spice. And
    the Saturn PCB tools thing.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dennis@21:1/5 to Liz Tuddenham on Sat Oct 26 16:17:22 2024
    On 10/26/24 13:19, Liz Tuddenham wrote:
    Joe Gwinn <joegwinn@comcast.net> wrote:

    On Sat, 26 Oct 2024 12:41:05 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:
    [...]

    A G3 won't run Linux or any of the OSX applications.

    Yeah. Linux will run on anything that will run Windows on Intel.
    Which modern Intel Macs will do. I have such a Mac, and run both
    MacOS and Windows on it. I bet that the new Apple silicon will run
    Windows in emulation as well.

    This machine is a G3 running Mac OS 8.6, it is not Intel-based and
    cannot run Windows or Linux or OSX.


    The G3 uses the IBM POWER PC chip.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Joe Gwinn@21:1/5 to Liz Tuddenham on Sat Oct 26 17:30:13 2024
    On Sat, 26 Oct 2024 19:19:57 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Joe Gwinn <joegwinn@comcast.net> wrote:

    On Sat, 26 Oct 2024 12:41:05 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:
    [...]

    A G3 won't run Linux or any of the OSX applications.

    Yeah. Linux will run on anything that will run Windows on Intel.
    Which modern Intel Macs will do. I have such a Mac, and run both
    MacOS and Windows on it. I bet that the new Apple silicon will run
    Windows in emulation as well.

    This machine is a G3 running Mac OS 8.6, it is not Intel-based and
    cannot run Windows or Linux or OSX.

    Yes, I know.

    I'm suggesting that it may be time for a small upgrade, say from the
    stone age to the bronze age. Don't want to rush headlong into
    anything, so stop before the iron age.

    I also have a windows laptop bought specifically for lab stuff that
    MacOs will never do, not even from Win10 running on the iMac (which is
    too big to schlepp around my little lab anyway). Although I can run
    LTSpice on the iMac, I could run Spice on the lab laptop as well.

    Joe Gwinn

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Klaus Vestergaard Kragelund@21:1/5 to john larkin on Sun Oct 27 01:14:45 2024
    On 23-10-2024 19:22, john larkin wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.


    The proximity of the aluminum is probably close to the effects of having
    ground plane or not below the inductor.

    Steve Sandler has tested this:

    https://www.signalintegrityjournal.com/blogs/17-practical-emc/post/2694-dc-dc-converters-solid-return-plane-or-cutouts-under-switch-node-and-inductor

    Found very little effect.

    Similar test:

    https://www.monolithicpower.com/en/learning/resources/when-is-it-beneficial-to-place-a-copper-layer-beneath-dc-dc-power-supplies?srsltid=AfmBOoq_cYcCoGN57iR4TXaq9n4hlYK1VWLt5m6yYGpVBdRF6RF7L7hz

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From john larkin@21:1/5 to klauskvik@hotmail.com on Sat Oct 26 19:16:45 2024
    On Sun, 27 Oct 2024 01:14:45 +0200, Klaus Vestergaard Kragelund <klauskvik@hotmail.com> wrote:

    On 23-10-2024 19:22, john larkin wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good
    cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133 >>
    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.


    The proximity of the aluminum is probably close to the effects of having >ground plane or not below the inductor.

    Steve Sandler has tested this:

    https://www.signalintegrityjournal.com/blogs/17-practical-emc/post/2694-dc-dc-converters-solid-return-plane-or-cutouts-under-switch-node-and-inductor

    Found very little effect.

    Similar test:

    https://www.monolithicpower.com/en/learning/resources/when-is-it-beneficial-to-place-a-copper-layer-beneath-dc-dc-power-supplies?srsltid=AfmBOoq_cYcCoGN57iR4TXaq9n4hlYK1VWLt5m6yYGpVBdRF6RF7L7hz


    They seem mostly concerned with EMI, which isn't a concern for me now.
    I just wanted to be sure that the inductor would work in the switching regulator, sandwiched close between a multilayer PCB and the aluminum baseplate.

    My 48-to-5-volt switcher should be OK. The load current is low so I
    can use a lot of microhenries if needed. If I trust my AADE LC-meter,
    L drops roughly 20% when the inductor is squeezed between the
    conductive things.

    The four half-bridge power switchers are more concerning. I think
    we'll try to make room for four giant shielded inductors on the parts
    side of that section... move other things to the bottom of the board.

    Our policy is for PCB layer 2 to be a solid ground plane, and we very
    rarely chop holes in that.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to legg on Sun Oct 27 15:01:34 2024
    On 27/10/2024 2:20 am, legg wrote:
    On Fri, 25 Oct 2024 16:08:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    <snip>
    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator
    paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    What's wrong with 'squegging' ? It's a simple word that covers
    a host of faults that all give the same approximate symptom . .

    With the advantage of 65 years of hindsight, it looks as if what he was
    seeing was gain in bipolar transistors running in the inverted mode.

    "Squegging" was mostly used for weird oscillations in resonant circuits.

    Class-D oscillators built with MOSFet switches don't squeg. Class-D
    oscillators built with bipolar transistors in LTSpice don't squeg either
    - the Gummel-Poon transistor model doesn't model inverted mode behavior
    all that well.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From legg@21:1/5 to All on Sun Oct 27 12:43:18 2024
    On Sun, 27 Oct 2024 15:01:34 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 27/10/2024 2:20 am, legg wrote:
    On Fri, 25 Oct 2024 16:08:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    <snip>
    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator
    paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    What's wrong with 'squegging' ? It's a simple word that covers
    a host of faults that all give the same approximate symptom . .

    With the advantage of 65 years of hindsight, it looks as if what he was >seeing was gain in bipolar transistors running in the inverted mode.

    "Squegging" was mostly used for weird oscillations in resonant circuits.

    Class-D oscillators built with MOSFet switches don't squeg. Class-D >oscillators built with bipolar transistors in LTSpice don't squeg either
    - the Gummel-Poon transistor model doesn't model inverted mode behavior
    all that well.

    Squegging in any oscillatory circuit, driven or otherwise,
    describes widely varying amplitudes that typically approach
    self-quenching and can otherwise approach unintentional
    overstess in the 'wobulating' cycle.

    Not what the doctor ordered, or the designer anticipated.

    Only blocking oscillators do it on purpose.

    RL

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to legg on Mon Oct 28 21:33:41 2024
    On 28/10/2024 3:43 am, legg wrote:
    On Sun, 27 Oct 2024 15:01:34 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 27/10/2024 2:20 am, legg wrote:
    On Fri, 25 Oct 2024 16:08:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    <snip>
    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator
    paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    What's wrong with 'squegging' ? It's a simple word that covers
    a host of faults that all give the same approximate symptom . .

    With the advantage of 65 years of hindsight, it looks as if what he was
    seeing was gain in bipolar transistors running in the inverted mode.

    "Squegging" was mostly used for weird oscillations in resonant circuits.

    Class-D oscillators built with MOSFet switches don't squeg. Class-D
    oscillators built with bipolar transistors in LTSpice don't squeg either
    - the Gummel-Poon transistor model doesn't model inverted mode behavior
    all that well.

    Squegging in any oscillatory circuit, driven or otherwise,
    describes widely varying amplitudes that typically approach
    self-quenching and can otherwise approach unintentional
    overstess in the 'wobulating' cycle.

    Not what the doctor ordered, or the designer anticipated.

    Only blocking oscillators do it on purpose.

    What Baxandall was describing was a situation where you've built a
    class-D oscillator and used a feed inductor which has an appreciably
    higher inductance than the inverter transformer.

    If you simulate that in LTSpice, the voltage at the centre tap starts
    off climbing up to about twice the steady-state peak and drops below the
    rail during recovery, but this roller-coaster effect dies away. In real
    life it doesn't (if you are using bipolar transistor for your switches).

    My guess is that you could stop it by adding the right zener diode
    between the centre tap and ground - one that didn't ever conduct when
    the circuit was running smoothly, but would start conducting if the
    centre tap got much above the steady state peak. This stops the
    centre-tap ever getting below the rail at the bottom of the start-up
    roller coaster - or at least it does in LTSpice and would keep you away
    from the mode of operation where the switching transistors were
    operating in the inverted mode.

    Peter Baxandall invented the circuit before 1959, before Zener diodes
    were widely available.

    --
    Bill Sloman, Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From piglet@21:1/5 to Bill Sloman on Mon Oct 28 12:06:11 2024
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 28/10/2024 3:43 am, legg wrote:
    On Sun, 27 Oct 2024 15:01:34 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 27/10/2024 2:20 am, legg wrote:
    On Fri, 25 Oct 2024 16:08:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>> wrote:

    <snip>
    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator >>>>> paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    What's wrong with 'squegging' ? It's a simple word that covers
    a host of faults that all give the same approximate symptom . .

    With the advantage of 65 years of hindsight, it looks as if what he was
    seeing was gain in bipolar transistors running in the inverted mode.

    "Squegging" was mostly used for weird oscillations in resonant circuits. >>>
    Class-D oscillators built with MOSFet switches don't squeg. Class-D
    oscillators built with bipolar transistors in LTSpice don't squeg either >>> - the Gummel-Poon transistor model doesn't model inverted mode behavior
    all that well.

    Squegging in any oscillatory circuit, driven or otherwise,
    describes widely varying amplitudes that typically approach
    self-quenching and can otherwise approach unintentional
    overstess in the 'wobulating' cycle.

    Not what the doctor ordered, or the designer anticipated.

    Only blocking oscillators do it on purpose.

    What Baxandall was describing was a situation where you've built a
    class-D oscillator and used a feed inductor which has an appreciably
    higher inductance than the inverter transformer.

    If you simulate that in LTSpice, the voltage at the centre tap starts
    off climbing up to about twice the steady-state peak and drops below the
    rail during recovery, but this roller-coaster effect dies away. In real
    life it doesn't (if you are using bipolar transistor for your switches).

    My guess is that you could stop it by adding the right zener diode
    between the centre tap and ground - one that didn't ever conduct when
    the circuit was running smoothly, but would start conducting if the
    centre tap got much above the steady state peak. This stops the
    centre-tap ever getting below the rail at the bottom of the start-up
    roller coaster - or at least it does in LTSpice and would keep you away
    from the mode of operation where the switching transistors were
    operating in the inverted mode.

    Peter Baxandall invented the circuit before 1959, before Zener diodes
    were widely available.


    That topology is very critical around conduction overlap vs dead band and
    nano seconds matter.


    --
    piglet

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bill Sloman@21:1/5 to piglet on Mon Oct 28 23:41:39 2024
    On 28/10/2024 11:06 pm, piglet wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 28/10/2024 3:43 am, legg wrote:
    On Sun, 27 Oct 2024 15:01:34 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 27/10/2024 2:20 am, legg wrote:
    On Fri, 25 Oct 2024 16:08:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>> wrote:

    <snip>
    He was remarkably good, just not totally perfect.

    His footnote reference to "squegging" in the 1959 class-D oscillator >>>>>> paper is another minor drop-off. He can't be blamed for it, but a
    super-hero might have done better.

    What's wrong with 'squegging' ? It's a simple word that covers
    a host of faults that all give the same approximate symptom . .

    With the advantage of 65 years of hindsight, it looks as if what he was >>>> seeing was gain in bipolar transistors running in the inverted mode.

    "Squegging" was mostly used for weird oscillations in resonant circuits. >>>>
    Class-D oscillators built with MOSFet switches don't squeg. Class-D
    oscillators built with bipolar transistors in LTSpice don't squeg either >>>> - the Gummel-Poon transistor model doesn't model inverted mode behavior >>>> all that well.

    Squegging in any oscillatory circuit, driven or otherwise,
    describes widely varying amplitudes that typically approach
    self-quenching and can otherwise approach unintentional
    overstess in the 'wobulating' cycle.

    Not what the doctor ordered, or the designer anticipated.

    Only blocking oscillators do it on purpose.

    What Baxandall was describing was a situation where you've built a
    class-D oscillator and used a feed inductor which has an appreciably
    higher inductance than the inverter transformer.

    If you simulate that in LTSpice, the voltage at the centre tap starts
    off climbing up to about twice the steady-state peak and drops below the
    rail during recovery, but this roller-coaster effect dies away. In real
    life it doesn't (if you are using bipolar transistor for your switches).

    My guess is that you could stop it by adding the right zener diode
    between the centre tap and ground - one that didn't ever conduct when
    the circuit was running smoothly, but would start conducting if the
    centre tap got much above the steady state peak. This stops the
    centre-tap ever getting below the rail at the bottom of the start-up
    roller coaster - or at least it does in LTSpice and would keep you away
    from the mode of operation where the switching transistors were
    operating in the inverted mode.

    Peter Baxandall invented the circuit before 1959, before Zener diodes
    were widely available.

    That topology is very critical around conduction overlap vs dead band and nano seconds matter.

    If you drive the transistor bases with a centre tapped secondary (with
    many fewer turns) as Peter originally described, the nanoseconds look
    after themselves. Conduction overlap isn't a good idea but an handful of nanoseconds of underlap isn't a problem.

    --
    Bill Sloman Sydney

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Klaus Vestergaard Kragelund@21:1/5 to john larkin on Thu Oct 31 01:10:48 2024
    On 27-10-2024 03:16, john larkin wrote:
    On Sun, 27 Oct 2024 01:14:45 +0200, Klaus Vestergaard Kragelund <klauskvik@hotmail.com> wrote:

    On 23-10-2024 19:22, john larkin wrote:
    I'm designing a small PCB with essentially 5 sync buck switching
    regulators. Board space is tight so I want to put the inductors on the
    bottom of the multilayer board. There's a 0.2" gap between the bottom
    of the board and a big aluminum flange.

    Unshielded drum cores have the most energy storage per volume or
    dollars. They store energy in the universe instead of in ferrite. Good
    cooling too.

    Something like this just fits

    https://www.digikey.com/en/products/detail/bourns-inc/SRN8040TA-470M/6155133

    Its mag field lines will bounce off the PCB planes and the flange,
    change from the classic bar magnet pattern into a pancake . I wonder
    what that will do to its electrical behavior.


    The proximity of the aluminum is probably close to the effects of having
    ground plane or not below the inductor.

    Steve Sandler has tested this:

    https://www.signalintegrityjournal.com/blogs/17-practical-emc/post/2694-dc-dc-converters-solid-return-plane-or-cutouts-under-switch-node-and-inductor

    Found very little effect.

    Similar test:

    https://www.monolithicpower.com/en/learning/resources/when-is-it-beneficial-to-place-a-copper-layer-beneath-dc-dc-power-supplies?srsltid=AfmBOoq_cYcCoGN57iR4TXaq9n4hlYK1VWLt5m6yYGpVBdRF6RF7L7hz


    They seem mostly concerned with EMI, which isn't a concern for me now.
    I just wanted to be sure that the inductor would work in the switching regulator, sandwiched close between a multilayer PCB and the aluminum baseplate.

    My 48-to-5-volt switcher should be OK. The load current is low so I
    can use a lot of microhenries if needed. If I trust my AADE LC-meter,
    L drops roughly 20% when the inductor is squeezed between the
    conductive things.

    The four half-bridge power switchers are more concerning. I think
    we'll try to make room for four giant shielded inductors on the parts
    side of that section... move other things to the bottom of the board.

    Our policy is for PCB layer 2 to be a solid ground plane, and we very
    rarely chop holes in that.

    If you don't have it, I can recommend buying the Bode 100 analyzer. It's
    not terrible expensive, and the SW is great.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jasen Betts@21:1/5 to Joe Gwinn on Sat Nov 2 09:34:16 2024
    On 2024-10-26, Joe Gwinn <joegwinn@comcast.net> wrote:
    On Sat, 26 Oct 2024 12:41:05 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:

    Liz Tuddenham <liz@poppyrecords.invalid.invalid> wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Fri, 25 Oct 2024 17:32:21 +0100, liz@poppyrecords.invalid.invalid
    (Liz Tuddenham) wrote:

    Bill Sloman <bill.sloman@ieee.org> wrote:

    On 25/10/2024 7:37 pm, Liz Tuddenham wrote:
    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 21:56:59 +0100, liz@poppyrecords.invalid.invalid >>> >>>>>> (Liz Tuddenham) wrote:

    john larkin <jl@glen--canyon.com> wrote:

    On Thu, 24 Oct 2024 18:45:20 +0200, Lasse Langwadt <llc@fonz.dk> >>> >>>>>>>> wrote:
    [...]
    plugging numbers pulled out of thin air into LTSpice is better that
    doing the actual measurement?



    It is for people who don't actually work with real parts.


    Peter Baxandall (of tone control and QUAD amplifier fame) claimed >>> >>>>>>> to use analogue computing to work out his designs i.e. He built >>> >>>>>>> prototypes and measured them.

    You youngsters probably don't remember a time when there wasn't
    Spice.

    Hey!!! Who are you calling a youngster?!!! :-)

    I did some simulation in Basic-Plus, and it was a nuisance. My first >>> >>>>>> PC sim program was Tatum labs ECA, which required a typed netlist. >>> >>>>>> But it was pretty cool.

    I have a spreadsheet I wrote for calculating the relationships
    between resistance, capacitance, frequency and time constant (put in >>> >>>>> two and the others appear, put in three and the error% appears). It >>> >>>>> also gives dB loss below or above the 'cutoff' frequency. Some years >>> >>>>> ago I also made some lookup tables for combinations of 5% tolerance >>> >>>>> resistors in series and parallel.

    Those and a pocket calculator are still the only 'computing' I use >>> >>>>> for design work.

    No surprise there, though I am a bit surprised that you would admit it >>> >>>> in public.

    I don't understand why you use the word 'admit'. I make my own tools to
    meet my particular requirements, there is nothing shameful about that. >>> >>
    You have been admonished by Sloman. Confess your errors and beg
    forgiveness.

    My reply to that had anatomical connotations. :-)


    Have you used LT Spice? It's easy to learn and is great fun. I(t makes >>> >> time for snacking and napping too.

    I don't think they ever did a version that would run on a Mac G3 (OS
    8.6), which is my main workhorse.



    I’ve been running LTspice using Wine on Linux for 15 years or so, no >>> problems. Wine runs fine on both x86 and apple silicon, I’m told.

    A G3 won't run Linux or any of the OSX applications.

    Yeah. Linux will run on anything that will run Windows on Intel.
    Which modern Intel Macs will do. I have such a Mac, and run both
    MacOS and Windows on it. I bet that the new Apple silicon will run
    Windows in emulation as well.

    They stopped making G3s a few years before Apple announced moving to
    Intel CPUs. It's 32-bit power-pc based, I don't where you'd find a
    32-bit PPC linux these days. The last Debian to support 32-bit PPC
    was version 8 "Stretch" for which support ended in 2020.

    --
    Jasen.
    🇺🇦 Слава Україні

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)