P
Phil Allison
Guest
DecadentLinux...@decadence.org wrote:
-------------------------------------
** ROTFL !!
...... Phil
-------------------------------------
** ROTFL !!
...... Phil
Navigation
Install the app
How to install the app on iOS
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
More options
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
P
Phil Allison
Guest
jla...@highlandsniptechnology.com wrote:
-------------------------------------------
** That is what VI limiters in a power amp do, vary the available current as the output voltage swings about in reverse proportion to the Vce appearing on upper and lower side BJTs at any moment.
The aim is to not allow excursions beyond the maker's SOA curve and hence device failure.
** So says a music hater.
** Another unsupported, blown out his fat arse opinion John gets to keep.
Yawnnnnnn ....
....... Phil
> lunatic fringe electronics
** Says it all, really.
-------------------------------------------
Looks to me that if one side swings close to a rail, and current
limits, it increases the current limit on the opposite side, namely
the one with a lot of voltage drop across it.
** That is what VI limiters in a power amp do, vary the available current as the output voltage swings about in reverse proportion to the Vce appearing on upper and lower side BJTs at any moment.
The aim is to not allow excursions beyond the maker's SOA curve and hence device failure.
Luckily, music is AC coupled so doesn't have really long duration
swings.
** So says a music hater.
That is really an ugly variation on the repeated-a-million times
standard audio power amp.
** Another unsupported, blown out his fat arse opinion John gets to keep.
Yawnnnnnn ....
....... Phil
> lunatic fringe electronics
** Says it all, really.
P
Phil Allison
Guest
piglet wrote:
---------------
** Having that -21V a boost rail make the amp work better and matches the benefits of the load side bootstrap on the class A stage.
TR107 is a current source so bucks any variation on the plus rail.
My first post said " the amp misbehaves terribly " - not it was merely a bit imperfect.
I have also posted that users could find the amp *unusable* though no faults had developed. Few comparable amps have the same problem.
It is not heat related & the amp does not stop working.
..... Phil
---------------
Another thing I am not sure about is the way the aux -ve rail is ac
bypassed not to ground but instead to the -80v rail bobbing up and down
with signal - could that invite motorboating. Like make C116 a higher
voltage part with anode returned to ground not -80V.
** Having that -21V a boost rail make the amp work better and matches the benefits of the load side bootstrap on the class A stage.
TR107 is a current source so bucks any variation on the plus rail.
My first post said " the amp misbehaves terribly " - not it was merely a bit imperfect.
I have also posted that users could find the amp *unusable* though no faults had developed. Few comparable amps have the same problem.
It is not heat related & the amp does not stop working.
..... Phil
P
Phil Allison
Guest
Phil Allison wrote:
-------------------------
** Here is a big clue:
The "misbehaviour" is an unintended result of the VI limiter.
TR113 & TR114 and their associated parts.
..... Phil
-------------------------
However, within the power stage is a flaw that makes the
amp misbehave terribly.
** Here is a big clue:
The "misbehaviour" is an unintended result of the VI limiter.
TR113 & TR114 and their associated parts.
..... Phil
P
piglet
Guest
On 23/10/2019 07:20, Phil Allison wrote:
Thanks for the clue but I'm still stumped. Does the C119 thru 122 combo
make the amplifier into a gigantic supersonic oscillator with certain
speaker loads?
piglet
Phil Allison wrote:
-------------------------
However, within the power stage is a flaw that makes the
amp misbehave terribly.
** Here is a big clue:
The "misbehaviour" is an unintended result of the VI limiter.
TR113 & TR114 and their associated parts.
.... Phil
Thanks for the clue but I'm still stumped. Does the C119 thru 122 combo
make the amplifier into a gigantic supersonic oscillator with certain
speaker loads?
piglet
A
Arie de Muynck
Guest
On 2019-10-22 18:05, jlarkin@highlandsniptechnology.com wrote:
No, the diodes takes care of that, it only acts on the side with the
lower voltage over the transistors (where the current flows). The other
side is not supposed to have much current, at least in audio designs.
Motor drivers do, I learned the hard way, they have a four-quadrant
load. The SOA computations there are more complex. Same for audio
designs with big transformer loads (like for electrostats).
Arie
On Tue, 22 Oct 2019 14:07:54 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 14:04, Arie de Muynck wrote:
D112/D113 seem upside down. This way they decrease the current limit
when there is less voltage over the output transistors, the reverse of a
SOA protection.
Correction --> they _increase_ the current limit when the voltage is
lower. Should be OK, unless wrongly dimensioned.
Looks to me that if one side swings close to a rail, and current
limits, it increases the current limit on the opposite side, namely
the one with a lot of voltage drop across it.
No, the diodes takes care of that, it only acts on the side with the
lower voltage over the transistors (where the current flows). The other
side is not supposed to have much current, at least in audio designs.
Motor drivers do, I learned the hard way, they have a four-quadrant
load. The SOA computations there are more complex. Same for audio
designs with big transformer loads (like for electrostats).
Arie
A
Arie de Muynck
Guest
On 2019-10-23 08:20, Phil Allison wrote:
Is that because of the SOA circuit _foldback_ effect when overload
occurs? Any clipping must have sound horrrible (maybe that's why people
still like soft clipping tube amps). Even some phase reversal might occur.
Hmmm... intriguing, but too busy now to start LTspice.
Arie
Phil Allison wrote:
However, within the power stage is a flaw that makes the
amp misbehave terribly.
** Here is a big clue:
The "misbehaviour" is an unintended result of the VI limiter.
TR113 & TR114 and their associated parts.
Is that because of the SOA circuit _foldback_ effect when overload
occurs? Any clipping must have sound horrrible (maybe that's why people
still like soft clipping tube amps). Even some phase reversal might occur.
Hmmm... intriguing, but too busy now to start LTspice.
Arie
P
Phil Allison
Guest
Arie de Muynck wrote:
---------------------
** You real sure about that ?
..... Phil
---------------------
No, the diodes takes care of that, it only acts on the side with the
lower voltage over the transistors (where the current flows). The other
side is not supposed to have much current, at least in audio designs.
** You real sure about that ?
..... Phil
W
Winfield Hill
Guest
Arie de Muynck wrote...
I never add capacitors to my current-limiting transistors;
you want it to turn off as fast as it turns on. When the
AC level drops you don't want continuing disabling action.
And as the polarity rapidly reverses at high frequencies,
you want the current-limiting off.
--
Thanks,
- Win
On 2019-10-23 08:20, Phil Allison wrote:
Phil Allison wrote:
However, within the power stage is a flaw that makes the
amp misbehave terribly.
** Here is a big clue:
The "misbehaviour" is an unintended result of the VI limiter.
TR113 & TR114 and their associated parts.
Is that because of the SOA circuit _foldback_ effect when
overload occurs? Any clipping must have sound horrrible
(maybe that's why people still like soft clipping tube amps).
Even some phase reversal might occur. Hmmm... intriguing,
but too busy now to start LTspice.
I never add capacitors to my current-limiting transistors;
you want it to turn off as fast as it turns on. When the
AC level drops you don't want continuing disabling action.
And as the polarity rapidly reverses at high frequencies,
you want the current-limiting off.
--
Thanks,
- Win
Guest
On Tue, 22 Oct 2019 14:50:15 -0700 (PDT), Phil Allison
<pallison49@gmail.com> wrote:
Sure, but that amp doesn't do it very well.
Do you use a DC coupled signal chain for your music?
In other words, you like this amp design.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
<pallison49@gmail.com> wrote:
jla...@highlandsniptechnology.com wrote:
-------------------------------------------
Looks to me that if one side swings close to a rail, and current
limits, it increases the current limit on the opposite side, namely
the one with a lot of voltage drop across it.
** That is what VI limiters in a power amp do, vary the available current as the output voltage swings about in reverse proportion to the Vce appearing on upper and lower side BJTs at any moment.
The aim is to not allow excursions beyond the maker's SOA curve and hence device failure.
Sure, but that amp doesn't do it very well.
Luckily, music is AC coupled so doesn't have really long duration
swings.
** So says a music hater.
Do you use a DC coupled signal chain for your music?
That is really an ugly variation on the repeated-a-million times
standard audio power amp.
** Another unsupported, blown out his fat arse opinion John gets to keep.
In other words, you like this amp design.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
Guest
** Here is a big clue:
The "misbehaviour" is an unintended result of the VI limiter.
TR113 & TR114 and their associated parts.
I'll take a swing..
The SOA transistor TR114 has unlimited authority to push current into the drivers TR108 TR109. Current through TR108/109 is limited only by R126 which is 68 Ohms and with >85 Volt supply that is >1 Amp. Probably more than TR108 and TR109 can withstand.
I guess that means with a loud enough audio drive signal and a low Z load that activates SOA protection, TR108 and / or TR109 may fail.
Mark
A
Allan Herriman
Guest
On Wed, 23 Oct 2019 04:30:09 -0700, Phil Allison wrote:
It ought to be fine as long as the amplifier is only connected to dummy
load resistors. Reactive loads (speakers?) might cause the current limit
to cut in earlier than expected.
Allan
Arie de Muynck wrote: ---------------------
No, the diodes takes care of that, it only acts on the side with the
lower voltage over the transistors (where the current flows). The other
side is not supposed to have much current, at least in audio designs.
** You real sure about that ?
.... Phil
It ought to be fine as long as the amplifier is only connected to dummy
load resistors. Reactive loads (speakers?) might cause the current limit
to cut in earlier than expected.
Allan
Guest
On Wed, 23 Oct 2019 10:11:04 +0200, Arie de Muynck
<no.spam@no.spam.org> wrote:
D112/113 increase the current limit of the transistors that are
swinging towards the rail, but don't decrease the opposite ones. That
would be OK if Iq is set low, but the Vbe multiplier isn't perfect. I
think this scheme over-stresses the not-active side on big swings. It
will NOT drive reactive loads well.
Speakers are largely ohmic, which makes life a lot simpler for audio
amps.
We're designing a class-D amp that will drive a big transformer, then
a wide range of loads, including various shunt-to-ground switching
regulators. We can tell from the data sheet of the TI class-D chip
that they didn't much consider non-speaker loads.
Our latest innovation is to put the heat sink and a fan on the bottom
of the board. Got to come up with a nice way to mount the fan to the
board. Brackets? Foam tape? Hot melt glue?
I like to do real SOA calculations, namely run a realtime model of
junction temperature and use that to limit or shut down. That
maximizes use of expensive silicon.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
<no.spam@no.spam.org> wrote:
On 2019-10-22 18:05, jlarkin@highlandsniptechnology.com wrote:
On Tue, 22 Oct 2019 14:07:54 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 14:04, Arie de Muynck wrote:
D112/D113 seem upside down. This way they decrease the current limit
when there is less voltage over the output transistors, the reverse of a
SOA protection.
Correction --> they _increase_ the current limit when the voltage is
lower. Should be OK, unless wrongly dimensioned.
Looks to me that if one side swings close to a rail, and current
limits, it increases the current limit on the opposite side, namely
the one with a lot of voltage drop across it.
No, the diodes takes care of that, it only acts on the side with the
lower voltage over the transistors (where the current flows). The other
side is not supposed to have much current, at least in audio designs.
D112/113 increase the current limit of the transistors that are
swinging towards the rail, but don't decrease the opposite ones. That
would be OK if Iq is set low, but the Vbe multiplier isn't perfect. I
think this scheme over-stresses the not-active side on big swings. It
will NOT drive reactive loads well.
Motor drivers do, I learned the hard way, they have a four-quadrant
load. The SOA computations there are more complex. Same for audio
designs with big transformer loads (like for electrostats).
Speakers are largely ohmic, which makes life a lot simpler for audio
amps.
We're designing a class-D amp that will drive a big transformer, then
a wide range of loads, including various shunt-to-ground switching
regulators. We can tell from the data sheet of the TI class-D chip
that they didn't much consider non-speaker loads.
Our latest innovation is to put the heat sink and a fan on the bottom
of the board. Got to come up with a nice way to mount the fan to the
board. Brackets? Foam tape? Hot melt glue?
I like to do real SOA calculations, namely run a realtime model of
junction temperature and use that to limit or shut down. That
maximizes use of expensive silicon.
--
John Larkin Highland Technology, Inc
lunatic fringe electronics
J
John Robertson
Guest
On 2019/10/23 7:55 a.m., jlarkin@highlandsniptechnology.com wrote:
combination of mounting hardware & brackets.
Hot melt or double sided tape can be problematic when exposed to the
real world...
John :-#)#
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
(604)872-5757 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."
If you want the fan to stay put, then I recommend securing it with someOn Wed, 23 Oct 2019 10:11:04 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 18:05, jlarkin@highlandsniptechnology.com wrote:
On Tue, 22 Oct 2019 14:07:54 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 14:04, Arie de Muynck wrote:
....
Our latest innovation is to put the heat sink and a fan on the bottom
of the board. Got to come up with a nice way to mount the fan to the
board. Brackets? Foam tape? Hot melt glue?
combination of mounting hardware & brackets.
Hot melt or double sided tape can be problematic when exposed to the
real world...
John :-#)#
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
(604)872-5757 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."
J
John Larkin
Guest
On Wed, 23 Oct 2019 11:14:48 -0700, John Robertson <spam@flippers.com>
wrote:
OK, grumble, I guess we don't want the fan falling off into the bottom
of the box.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
wrote:
On 2019/10/23 7:55 a.m., jlarkin@highlandsniptechnology.com wrote:
On Wed, 23 Oct 2019 10:11:04 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 18:05, jlarkin@highlandsniptechnology.com wrote:
On Tue, 22 Oct 2019 14:07:54 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 14:04, Arie de Muynck wrote:
...
Our latest innovation is to put the heat sink and a fan on the bottom
of the board. Got to come up with a nice way to mount the fan to the
board. Brackets? Foam tape? Hot melt glue?
If you want the fan to stay put, then I recommend securing it with some
combination of mounting hardware & brackets.
Hot melt or double sided tape can be problematic when exposed to the
real world...
OK, grumble, I guess we don't want the fan falling off into the bottom
of the box.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
P
Phil Allison
Guest
John Larkin wrote:
-------------------
** One and the same thing.
> Well, I do hate *some* music, as you probably do.
** Not a lot.
** Not true.
Most professional audio gear is well designed and built for the purpose intended and at very low cost cos the competition is *fierce*.
** Really, so the load's characteristics are irrelevant to your design ?
That's cool .....
....... Phil
-------------------
which makes life a lot simpler for audio amps.
** Says an audio hater and ignoramous.
I don't hate audio, I just don't care for music.
** One and the same thing.
> Well, I do hate *some* music, as you probably do.
** Not a lot.
But a lot of audio electronics is badly designed,
** Not true.
Most professional audio gear is well designed and built for the purpose intended and at very low cost cos the competition is *fierce*.
I like to do real SOA calculations,
** Nice if you have a known load.
No, I use hardware or software to compute mosfet junction temps, using
realtime measurement of voltages and currents and heatsink temp, and a
thermal model of the chip and heat sink. The computed Tj is realistic,
independent of signals and loads. It's a small cost addition that
radically improves the use of expensive fets and heat sinks and gets
max safe output in all quadrants.
** Really, so the load's characteristics are irrelevant to your design ?
That's cool .....
....... Phil
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
P
Phil Allison
Guest
jla...@highlandsniptechnology.com wrote:
---------------------------------------
** A clearer way to view this is to note that equal amounts of current are available in both polarities at zero output. Available current increases as the output swings towards one or other rail.
** Who's IQ is that then ?
** LOL - what a shameless non sequitur.
JL is drowning in his own bullshit here.
** If it's "non-active" how then does it become stressed ?
** Really? Why is that ?
JL in on the sniff of a scent here.
** Massive fallacy.
> which makes life a lot simpler for audio amps.
** Says an audio hater and ignoramous.
** Nice if you have a known load.
...... Phil
---------------------------------------
D112/113 increase the current limit of the transistors that are
swinging towards the rail, but don't decrease the opposite ones.
** A clearer way to view this is to note that equal amounts of current are available in both polarities at zero output. Available current increases as the output swings towards one or other rail.
That would be OK if Iq is set low,
** Who's IQ is that then ?
but the Vbe multiplier isn't perfect.
** LOL - what a shameless non sequitur.
JL is drowning in his own bullshit here.
I
think this scheme over-stresses the not-active side on big swings.
** If it's "non-active" how then does it become stressed ?
It will NOT drive reactive loads well.
** Really? Why is that ?
JL in on the sniff of a scent here.
Speakers are largely ohmic,
** Massive fallacy.
> which makes life a lot simpler for audio amps.
** Says an audio hater and ignoramous.
I like to do real SOA calculations,
** Nice if you have a known load.
...... Phil
J
John Larkin
Guest
On Wed, 23 Oct 2019 12:47:21 -0700 (PDT), Phil Allison
<pallison49@gmail.com> wrote:
I don't hate audio, I just don't care for music. Well, I do hate
*some* music, as you probably do. But a lot of audio electronics is
badly designed, as you have pointed out.
No, I use hardware or software to compute mosfet junction temps, using
realtime measurement of voltages and currents and heatsink temp, and a
thermal model of the chip and heat sink. The computed Tj is realistic,
independent of signals and loads. It's a small cost addition that
radically improves the use of expensive fets and heat sinks and gets
max safe output in all quadrants.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
<pallison49@gmail.com> wrote:
which makes life a lot simpler for audio amps.
** Says an audio hater and ignoramous.
I don't hate audio, I just don't care for music. Well, I do hate
*some* music, as you probably do. But a lot of audio electronics is
badly designed, as you have pointed out.
I like to do real SOA calculations,
** Nice if you have a known load.
No, I use hardware or software to compute mosfet junction temps, using
realtime measurement of voltages and currents and heatsink temp, and a
thermal model of the chip and heat sink. The computed Tj is realistic,
independent of signals and loads. It's a small cost addition that
radically improves the use of expensive fets and heat sinks and gets
max safe output in all quadrants.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
L
Lasse Langwadt Christense
Guest
onsdag den 23. oktober 2019 kl. 16.55.19 UTC+2 skrev jla...@highlandsniptechnology.com:
two holes?
https://www.amazon.com/Northbridge-Chipset-Aluminium-Heatsink-Printer/dp/B07RKM2JJ7
On Wed, 23 Oct 2019 10:11:04 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 18:05, jlarkin@highlandsniptechnology.com wrote:
On Tue, 22 Oct 2019 14:07:54 +0200, Arie de Muynck
no.spam@no.spam.org> wrote:
On 2019-10-22 14:04, Arie de Muynck wrote:
D112/D113 seem upside down. This way they decrease the current limit
when there is less voltage over the output transistors, the reverse of a
SOA protection.
Correction --> they _increase_ the current limit when the voltage is
lower. Should be OK, unless wrongly dimensioned.
Looks to me that if one side swings close to a rail, and current
limits, it increases the current limit on the opposite side, namely
the one with a lot of voltage drop across it.
No, the diodes takes care of that, it only acts on the side with the
lower voltage over the transistors (where the current flows). The other
side is not supposed to have much current, at least in audio designs.
D112/113 increase the current limit of the transistors that are
swinging towards the rail, but don't decrease the opposite ones. That
would be OK if Iq is set low, but the Vbe multiplier isn't perfect. I
think this scheme over-stresses the not-active side on big swings. It
will NOT drive reactive loads well.
Motor drivers do, I learned the hard way, they have a four-quadrant
load. The SOA computations there are more complex. Same for audio
designs with big transformer loads (like for electrostats).
Speakers are largely ohmic, which makes life a lot simpler for audio
amps.
We're designing a class-D amp that will drive a big transformer, then
a wide range of loads, including various shunt-to-ground switching
regulators. We can tell from the data sheet of the TI class-D chip
that they didn't much consider non-speaker loads.
Our latest innovation is to put the heat sink and a fan on the bottom
of the board. Got to come up with a nice way to mount the fan to the
board. Brackets? Foam tape? Hot melt glue?
two holes?
https://www.amazon.com/Northbridge-Chipset-Aluminium-Heatsink-Printer/dp/B07RKM2JJ7
J
John Larkin
Guest
On Wed, 23 Oct 2019 13:28:17 -0700 (PDT), Phil Allison
<pallison49@gmail.com> wrote:
The load should be and is irrevelent to the protection algorithm. I
compute a pretty good approximation of actual fet junction temp, and
shut down if that gets too high. No simple current limit or foldback
does that.
Fets, fans, heat sinks are expensive. May as well use them to their
safe limits.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
<pallison49@gmail.com> wrote:
John Larkin wrote:
-------------------
which makes life a lot simpler for audio amps.
** Says an audio hater and ignoramous.
I don't hate audio, I just don't care for music.
** One and the same thing.
Well, I do hate *some* music, as you probably do.
** Not a lot.
But a lot of audio electronics is badly designed,
** Not true.
Most professional audio gear is well designed and built for the purpose intended and at very low cost cos the competition is *fierce*.
I like to do real SOA calculations,
** Nice if you have a known load.
No, I use hardware or software to compute mosfet junction temps, using
realtime measurement of voltages and currents and heatsink temp, and a
thermal model of the chip and heat sink. The computed Tj is realistic,
independent of signals and loads. It's a small cost addition that
radically improves the use of expensive fets and heat sinks and gets
max safe output in all quadrants.
** Really, so the load's characteristics are irrelevant to your design ?
That's cool .....
The load should be and is irrevelent to the protection algorithm. I
compute a pretty good approximation of actual fet junction temp, and
shut down if that gets too high. No simple current limit or foldback
does that.
Fets, fans, heat sinks are expensive. May as well use them to their
safe limits.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com