Welcome Notice

Register Log in

Which N Or P Power MOSFETs ?...

P

Phil Hobbs

Guest
On 11/22/20 6:43 PM, John S wrote:
On 11/22/2020 3:34 PM, Phil Hobbs wrote:
On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5,
jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban
wrote:
Many know I set out to build a really good amp. Well after
all that I found it can\'t work. The drawing was on the bench
and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs.
This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either
?N channel or P channel. The difference it the power supply.
Which is easier to design, but what if the better choice need
negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably
fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right
now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend
to blow
up at some fraction of their rated power dissipation; bipolars
do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply
with stepped taps
on the transformer.) How about a linear amp (inner loop) with
some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\"
or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah.  The load is important.  R\'s are easy.
Walking around thinking, I don\'t really like the  multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff.  (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure.  It\'s called a \"plate modulator\". ;)

Cheers

Phil Hobbs


Where would be the delay line in a plate modulator?
\"I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope.\"

Teasing friends is a bit of a way of life round here. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
S

server

Guest
On Sun, 22 Nov 2020 16:54:01 -0800, boB <boB@K7IQ.com> wrote:

On Sun, 22 Nov 2020 10:41:03 -0800, jlarkin@highlandsniptechnology.com
wrote:

On Sun, 22 Nov 2020 13:09:47 -0500, bitrex <user@example.net> wrote:

On 11/21/2020 9:50 PM, Bill Sloman wrote:
On Sunday, November 22, 2020 at 3:23:20 AM UTC+11, jla...@highlandsniptechnology.com wrote:
On Sat, 21 Nov 2020 03:00:50 -0500, bitrex <us...@example.net> wrote:

On 11/21/2020 12:44 AM, jla...@highlandsniptechnology.com wrote:
On Fri, 20 Nov 2020 16:27:03 -0800 (PST), Phil Allison
palli...@gmail.com> wrote:

some bit brain fuckhead wrote:
=============================

Yes also because all the graphs and safe operating area curves for
modern switching MOSFETs are of little use when they\'re used in a linear
application.


** FFS post *only* on topics you know something about.

That will narrow the field right down to SFA.

snip

The SOA curve in most switching device datasheets tells you e.g. that
you can push 20 amps at DC with the MOSFET saturated and the case at 25
degrees. Amazing stuff. I try to run all linear power amps pushing 20
amps with the case at 25 degrees personally

We never believed the SOAR curves, so we just pushed them until they
exploded.

That probably means that John Larkin never understood what the Safe Operating Area curves were trying to tell him. They do present a plot of safe drain currents against drain-to-source voltage which does tell you something about operating the parts outside the saturation region, despite Bitrex\'s claim to the contrary,

The data sheet for Phil Allison\'s favourite N-channel part

http://www.irf.com/product-info/datasheets/data/irfp240.pdf

shows just such a curve at figure 8.

http://www.irf.com/product-info/datasheets/data/irfp9240.pdf

shows a very similar datasheet for the P-channel part Phil reccommends. Again, the SOA curve is a figure 8.

snipped John boasting about something some customer managed to extract from him


The claim I\'m making is that those constant power curves at the right
assume the temperature is uniform across the MOSFET die, as it\'s likely
to be in a switching application where it\'s either fully off or each
pulse is putting it solidly into the ohmic region.


Makes no sense. Fig 8 clearly has data at 200 volts and 5 amps.

I don\'t see any assumptions about temperature uniformity. The SOAR
curve just specs what is safe.

(Not that I unreservedly believe them. Processes change more often
than data sheets.)


Oh My ! Process changes are changes that I have to worry about. Or
used to have to at least. Back in the 1990s, in one of our power
inverters, we used a certain Motorola (On Semi) FET in the circuit.

They worked fine untill they did a die shrink on that part.

Inverters have to surge and we rely on the large die area for
transient thermal transfer. They said \"but the RdsOn is the same !\"

Well... They did not understand evidently how that can hurt the
thermal impedance in the application.

As I remember, we changed to Harris Semicondutor (then Fairchild and
On Semi again I think) and all was well.

Never had a problem like that again luckily.
I did a half-bridge motor driver and used two Moto fets. During
braking, the substrate diodes conducted some. They had the ideal
diffusion profile of a step-recovery diode, made enormous spikes, and
blew out their own gates.

Does anyone know if that\'s still ever a problem? My new class-D amp
has schottly diodes across each fet, just in case, but I may not need
them. May well stuff them; they don\'t cost much.




--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.
 
T

Tauno Voipio

Guest
On 23.11.20 1.43, John S wrote:
On 11/22/2020 3:34 PM, Phil Hobbs wrote:
On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5,
jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban
wrote:
Many know I set out to build a really good amp. Well after
all that I found it can\'t work. The drawing was on the bench
and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs.
This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either
?N channel or P channel. The difference it the power supply.
Which is easier to design, but what if the better choice need
negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably
fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right
now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend
to blow
up at some fraction of their rated power dissipation; bipolars
do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply
with stepped taps
on the transformer.) How about a linear amp (inner loop) with
some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\"
or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah.  The load is important.  R\'s are easy.
Walking around thinking, I don\'t really like the  multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff.  (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure.  It\'s called a \"plate modulator\". ;)

Cheers

Phil Hobbs


Where would be the delay line in a plate modulator?
Not needed for straight classic AM, but an EER amplifier handles
more complex modulations by handling the phase and amplitude
components separately. You may need delays to match the component
timings at the final. The plate modulator is just the amplitude
handler.

--

-TV
 
S

server

Guest
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Cheers

Phil Hobbs
Aren\'t plates those red things in tubes?

The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.



--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.
 
T

Tauno Voipio

Guest
On 23.11.20 17.42, jlarkin@highlandsniptechnology.com wrote:
Aren\'t plates those red things in tubes?
Not usually - except in a ham contest.

There were a few tubes designed to run plates red at nominal
dissipation. I recall Philips transmitting tube QE 08/200 which
specified \'Anodes red hot, remperature 850 K).


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.
You do not need anything more exotic than a cellular phone base station.

--

-TV

(Less than 10 miles from Nokia headquarters and factory).
 
P

Phil Hobbs

Guest
On 11/23/20 10:42 AM, jlarkin@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?
If you\'re hanging round with Joerg, yes.

The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.
Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
S

server

Guest
On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jlarkin@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs
Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.



--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.
 
L

Lasse Langwadt Christensen

Guest
tirsdag den 24. november 2020 kl. 00.57.43 UTC+1 skrev jla...@highlandsniptechnology.com:
On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
pcdhSpamM...@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jla...@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamM...@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs
Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.
that\'s all software-defined radio any way so you can control it all digitally
no need for analog delays
 
B

boB

Guest
On Sun, 22 Nov 2020 17:55:37 -0800, jlarkin@highlandsniptechnology.com
wrote:

On Sun, 22 Nov 2020 16:54:01 -0800, boB <boB@K7IQ.com> wrote:

On Sun, 22 Nov 2020 10:41:03 -0800, jlarkin@highlandsniptechnology.com
wrote:

On Sun, 22 Nov 2020 13:09:47 -0500, bitrex <user@example.net> wrote:

On 11/21/2020 9:50 PM, Bill Sloman wrote:
On Sunday, November 22, 2020 at 3:23:20 AM UTC+11, jla...@highlandsniptechnology.com wrote:
On Sat, 21 Nov 2020 03:00:50 -0500, bitrex <us...@example.net> wrote:

On 11/21/2020 12:44 AM, jla...@highlandsniptechnology.com wrote:
On Fri, 20 Nov 2020 16:27:03 -0800 (PST), Phil Allison
palli...@gmail.com> wrote:

some bit brain fuckhead wrote:
=============================

Yes also because all the graphs and safe operating area curves for
modern switching MOSFETs are of little use when they\'re used in a linear
application.


** FFS post *only* on topics you know something about.

That will narrow the field right down to SFA.

snip

The SOA curve in most switching device datasheets tells you e.g. that
you can push 20 amps at DC with the MOSFET saturated and the case at 25
degrees. Amazing stuff. I try to run all linear power amps pushing 20
amps with the case at 25 degrees personally

We never believed the SOAR curves, so we just pushed them until they
exploded.

That probably means that John Larkin never understood what the Safe Operating Area curves were trying to tell him. They do present a plot of safe drain currents against drain-to-source voltage which does tell you something about operating the parts outside the saturation region, despite Bitrex\'s claim to the contrary,

The data sheet for Phil Allison\'s favourite N-channel part

http://www.irf.com/product-info/datasheets/data/irfp240.pdf

shows just such a curve at figure 8.

http://www.irf.com/product-info/datasheets/data/irfp9240.pdf

shows a very similar datasheet for the P-channel part Phil reccommends. Again, the SOA curve is a figure 8.

snipped John boasting about something some customer managed to extract from him


The claim I\'m making is that those constant power curves at the right
assume the temperature is uniform across the MOSFET die, as it\'s likely
to be in a switching application where it\'s either fully off or each
pulse is putting it solidly into the ohmic region.


Makes no sense. Fig 8 clearly has data at 200 volts and 5 amps.

I don\'t see any assumptions about temperature uniformity. The SOAR
curve just specs what is safe.

(Not that I unreservedly believe them. Processes change more often
than data sheets.)


Oh My ! Process changes are changes that I have to worry about. Or
used to have to at least. Back in the 1990s, in one of our power
inverters, we used a certain Motorola (On Semi) FET in the circuit.

They worked fine untill they did a die shrink on that part.

Inverters have to surge and we rely on the large die area for
transient thermal transfer. They said \"but the RdsOn is the same !\"

Well... They did not understand evidently how that can hurt the
thermal impedance in the application.

As I remember, we changed to Harris Semicondutor (then Fairchild and
On Semi again I think) and all was well.

Never had a problem like that again luckily.







I did a half-bridge motor driver and used two Moto fets. During
braking, the substrate diodes conducted some. They had the ideal
diffusion profile of a step-recovery diode, made enormous spikes, and
blew out their own gates.

Does anyone know if that\'s still ever a problem? My new class-D amp
has schottly diodes across each fet, just in case, but I may not need
them. May well stuff them; they don\'t cost much.
I find tht the diode recovery problems are way worse with higher
voltage applications... i.e. Low voltage, less than around 200V or
150V, regular Silicon FET\'s intrinsic diode isn\'t too bad. Better as
you go lower in voltage. That high V and fast reverse swing kind of
rips the current through both high and low FETs apart.. So to speak.

And the higher voltage Schottkys with low forward voltage (> ~200V)
aren\'t available at real high V so doesn\'t do much good, if any, in
parallel with the HV FETs... And, the HV Schottkys have a positive
temperature-voltage coeffiient so they don\'t share or over-share the
reverse current as they heat up. Damn. Can\'t have anything !

My experience as I remember anyway.

What was your motor driver voltage ? That might complicate things
just being a motor drive. I don\'t know
 
B

boB

Guest
On Mon, 23 Nov 2020 15:57:31 -0800, jlarkin@highlandsniptechnology.com
wrote:

On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jlarkin@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs

Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.
I think that Space-X has a lot of those smart people. Would be great
to be a bug on the wall watching them talk about their work.
 
S

server

Guest
On Mon, 23 Nov 2020 19:28:06 -0800, boB <boB@K7IQ.com> wrote:

On Sun, 22 Nov 2020 17:55:37 -0800, jlarkin@highlandsniptechnology.com
wrote:

On Sun, 22 Nov 2020 16:54:01 -0800, boB <boB@K7IQ.com> wrote:

On Sun, 22 Nov 2020 10:41:03 -0800, jlarkin@highlandsniptechnology.com
wrote:

On Sun, 22 Nov 2020 13:09:47 -0500, bitrex <user@example.net> wrote:

On 11/21/2020 9:50 PM, Bill Sloman wrote:
On Sunday, November 22, 2020 at 3:23:20 AM UTC+11, jla...@highlandsniptechnology.com wrote:
On Sat, 21 Nov 2020 03:00:50 -0500, bitrex <us...@example.net> wrote:

On 11/21/2020 12:44 AM, jla...@highlandsniptechnology.com wrote:
On Fri, 20 Nov 2020 16:27:03 -0800 (PST), Phil Allison
palli...@gmail.com> wrote:

some bit brain fuckhead wrote:
=============================

Yes also because all the graphs and safe operating area curves for
modern switching MOSFETs are of little use when they\'re used in a linear
application.


** FFS post *only* on topics you know something about.

That will narrow the field right down to SFA.

snip

The SOA curve in most switching device datasheets tells you e.g. that
you can push 20 amps at DC with the MOSFET saturated and the case at 25
degrees. Amazing stuff. I try to run all linear power amps pushing 20
amps with the case at 25 degrees personally

We never believed the SOAR curves, so we just pushed them until they
exploded.

That probably means that John Larkin never understood what the Safe Operating Area curves were trying to tell him. They do present a plot of safe drain currents against drain-to-source voltage which does tell you something about operating the parts outside the saturation region, despite Bitrex\'s claim to the contrary,

The data sheet for Phil Allison\'s favourite N-channel part

http://www.irf.com/product-info/datasheets/data/irfp240.pdf

shows just such a curve at figure 8.

http://www.irf.com/product-info/datasheets/data/irfp9240.pdf

shows a very similar datasheet for the P-channel part Phil reccommends. Again, the SOA curve is a figure 8.

snipped John boasting about something some customer managed to extract from him


The claim I\'m making is that those constant power curves at the right
assume the temperature is uniform across the MOSFET die, as it\'s likely
to be in a switching application where it\'s either fully off or each
pulse is putting it solidly into the ohmic region.


Makes no sense. Fig 8 clearly has data at 200 volts and 5 amps.

I don\'t see any assumptions about temperature uniformity. The SOAR
curve just specs what is safe.

(Not that I unreservedly believe them. Processes change more often
than data sheets.)


Oh My ! Process changes are changes that I have to worry about. Or
used to have to at least. Back in the 1990s, in one of our power
inverters, we used a certain Motorola (On Semi) FET in the circuit.

They worked fine untill they did a die shrink on that part.

Inverters have to surge and we rely on the large die area for
transient thermal transfer. They said \"but the RdsOn is the same !\"

Well... They did not understand evidently how that can hurt the
thermal impedance in the application.

As I remember, we changed to Harris Semicondutor (then Fairchild and
On Semi again I think) and all was well.

Never had a problem like that again luckily.







I did a half-bridge motor driver and used two Moto fets. During
braking, the substrate diodes conducted some. They had the ideal
diffusion profile of a step-recovery diode, made enormous spikes, and
blew out their own gates.

Does anyone know if that\'s still ever a problem? My new class-D amp
has schottly diodes across each fet, just in case, but I may not need
them. May well stuff them; they don\'t cost much.


I find tht the diode recovery problems are way worse with higher
voltage applications... i.e. Low voltage, less than around 200V or
150V, regular Silicon FET\'s intrinsic diode isn\'t too bad. Better as
you go lower in voltage. That high V and fast reverse swing kind of
rips the current through both high and low FETs apart.. So to speak.

And the higher voltage Schottkys with low forward voltage (> ~200V)
aren\'t available at real high V so doesn\'t do much good, if any, in
parallel with the HV FETs... And, the HV Schottkys have a positive
temperature-voltage coeffiient so they don\'t share or over-share the
reverse current as they heat up. Damn. Can\'t have anything !

My experience as I remember anyway.

What was your motor driver voltage ? That might complicate things
just being a motor drive. I don\'t know
It was a 24 volt supply, and two fets in half-bridge, driving a motor
whose other end was grounded. We reversed with a real DPDT switch. It
was sort of a charity project.

Higher-voltage diodes tend to be pin structures, and often have a
doping profile that makes them drift step-recovery (Grehkov) diodes. I
guess a fet substrate diode could do that to. Maybe those moto mosfets
were oddballs.

My current project is a full h-bridge, running from 48 volts, driving
a transformer. The schottkies across the fets might increase
efficiency a tiny bit too.



--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.
 
P

Phil Hobbs

Guest
On 11/23/20 6:57 PM, jlarkin@highlandsniptechnology.com wrote:
On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jlarkin@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs

Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.
Nah, they don\'t pay enough for that.

The idea of working on a project that large gives me the screaming
heebie-jeebies. An optical table, an interesting problem, one or two
colleagues, and somebody to sign purchase requests without giving me
agita is the ticket. ;)

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
S

server

Guest
On Tue, 24 Nov 2020 17:42:14 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 6:57 PM, jlarkin@highlandsniptechnology.com wrote:
On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jlarkin@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs

Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.

Nah, they don\'t pay enough for that.

The idea of working on a project that large gives me the screaming
heebie-jeebies. An optical table, an interesting problem, one or two
colleagues, and somebody to sign purchase requests without giving me
agita is the ticket. ;)

Cheers

Phil Hobbs
We lost an interesting redesign to an in-house team specially
assembled to do it, two groups totalling about 20 engineers on two
continents. Might have taken us 4 months. Three years later, theirs
still isn\'t done. At some point, more engineers and especially more
management turns the progress rate back down. Approaching a man-year
per square inch.

It\'s sad how they have to keep buying the old stuff from us.



--

John Larkin Highland Technology, Inc

The best designs are necessarily accidental.
 
P

Phil Hobbs

Guest
On 11/24/20 10:43 PM, jlarkin@highlandsniptechnology.com wrote:
On Tue, 24 Nov 2020 17:42:14 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 6:57 PM, jlarkin@highlandsniptechnology.com wrote:
On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jlarkin@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
ggherold@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs

Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.

Nah, they don\'t pay enough for that.

The idea of working on a project that large gives me the screaming
heebie-jeebies. An optical table, an interesting problem, one or two
colleagues, and somebody to sign purchase requests without giving me
agita is the ticket. ;)


We lost an interesting redesign to an in-house team specially
assembled to do it, two groups totalling about 20 engineers on two
continents. Might have taken us 4 months. Three years later, theirs
still isn\'t done. At some point, more engineers and especially more
management turns the progress rate back down. Approaching a man-year
per square inch.

It\'s sad how they have to keep buying the old stuff from us.
;)

From our point of view the main problem working with large groups at
clients is that they\'re generally seriously underdamped. We\'ve learned
to nitpick every specification document they send us--not so much so
we can understand what they want, but so _they_ all have to understand it.

Awhile back we lost a potentially good customer by being too polite
about their specifications document for an ultraquiet laser. Turned out
that the guy who wrote it (and illustrated it profusely) thought we were
replacing one vendor\'s lasers, whereas the rest of their team wanted us
to replace their other vendor\'s laser. They just hadn\'t read their own
document carefully. :(

We actually fished that one out, and built greatly improved replacements
for _both_ vendors\' lasers within the specified budget, but then their
project lead left the company and the project got cancelled. On the
plus side, we now have this super quiet laser system that we can sell to
other folks. :)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
B

Bill Sloman

Guest
On Wednesday, November 25, 2020 at 2:43:32 PM UTC+11, jla...@highlandsniptechnology.com wrote:
On Tue, 24 Nov 2020 17:42:14 -0500, Phil Hobbs
pcdhSpamM...@electrooptical.net> wrote:

On 11/23/20 6:57 PM, jla...@highlandsniptechnology.com wrote:
On Mon, 23 Nov 2020 15:17:31 -0500, Phil Hobbs
pcdhSpamM...@electrooptical.net> wrote:

On 11/23/20 10:42 AM, jla...@highlandsniptechnology.com wrote:
On Sun, 22 Nov 2020 16:34:04 -0500, Phil Hobbs
pcdhSpamM...@electrooptical.net> wrote:

On 11/19/20 7:42 PM, John Larkin wrote:
On Thu, 19 Nov 2020 15:25:29 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 4:51:32 PM UTC-5, John Larkin wrote:
On Thu, 19 Nov 2020 09:12:04 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 11:51:58 AM UTC-5, piglet wrote:
On 19/11/2020 15:52, George Herold wrote:
On Thursday, November 19, 2020 at 10:37:06 AM UTC-5, jla...@highlandsniptechnology.com wrote:
On Thu, 19 Nov 2020 05:31:20 -0800 (PST), George Herold
gghe...@gmail.com> wrote:

On Thursday, November 19, 2020 at 7:25:48 AM UTC-5, Jeff Urban wrote:
Many know I set out to build a really good amp. Well after all that I found it can\'t work. The drawing was on the bench and I saw the problem immediately, at a glance. Damn.

But I did find it. Now it needs power MOSFETs for outputs. This looks pretty much non-negotiable.

I forget which is which but they are all the same. Al either ?N channel or P channel. The difference it the power supply. Which is easier to design, but what if the better choice need negative, ? Then I draw it upside down, so what ?

So which is better or more linear or whatever ?

We are in the 140V/18A range.
Is this a linear amp? If so that\'s a lot of heat.
Do you have \"Art of Electronics\". (The 2nd ed. is probably fairly cheap now.)

George H.
The problem is not well specified. \"Good amp\" is not very clear.

Class-D amps are simple and efficient. I\'m designing one right now.

Most mosfets are designed for switching and don\'t take kindly to
linear operation, way out there on their SOAR curve. They tend to blow
up at some fraction of their rated power dissipation; bipolars do that
too.

We learned about that.

https://www.dropbox.com/s/4nxm7m2q3j3buvc/ExFets.jpg?raw=1



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard
Here\'s a crazy idea... ( idea stolen from a linear power supply with stepped taps
on the transformer.) How about a linear amp (inner loop) with some switched
power supply rails... ? It would probably be ugly.
George H.

Yep, it\'s been done already. Goes by weird names like class \"H\" or \"G\"
or whatever marketing thinks sounds cool.

piglet
Thanks piglet, you can tell I\'m an audio expert. :^)
Not much when searching for class G/H but this seems good.
https://sound-au.com/articles/class-g.htm

GH
We don\'t know if the OP wants to drive motors or speakers or rail
guns.

Yeah. The load is important. R\'s are easy.
Walking around thinking, I don\'t really like the multi-tapped
G-amp anymore than the two tap A/B amp. (+/-)
When driving weird loads cross-over distortion.. hic-ups is a concern.
I\'m thinking the 3 tap class G thing has three times as many
cross-overs... depending on the amplitude.

So how about a class D amp doing a (relatively) slow power rail.
(say 1 ms)
and a class A amp inside doing the fast stuff. (1 us?)
I\'m thinking single sided.
That wouldn\'t work for a short pulse.

George H.


A real class-D amp is better than any human can tell, assuming we are
talking audio here. That\'s certainly a reasonable way to get
kilowatts.

I think there is a class of RF power amps that modulate the power
supplies of the final, to track the RF envelope. Delay lines are
involved I recall.

Sure. It\'s called a \"plate modulator\". ;)

Aren\'t plates those red things in tubes?

If you\'re hanging round with Joerg, yes.


The thing I was remembering was about using an ultra-linear amp to
output a complex constellation, maybe with multiple carriers, where
distortion would be expensive. So they delayed the signal long enough
to have the power supplies get ready for the rare but big amplitude
peaks.

It probably has a name. That sort of thing sells those million-dollar
oscilloscopes.

Dunno. Stuff like that is enough to give small children nightmares.

Cheers

Phil Hobbs

Yeah, I was thinking about the engineering required for 5G and such.
That must suck up most of the smartest people on this planet.

Nah, they don\'t pay enough for that.

The idea of working on a project that large gives me the screaming
heebie-jeebies. An optical table, an interesting problem, one or two
colleagues, and somebody to sign purchase requests without giving me
agita is the ticket. ;)

We lost an interesting redesign to an in-house team specially
assembled to do it, two groups totalling about 20 engineers on two
continents. Might have taken us 4 months. Three years later, theirs
still isn\'t done. At some point, more engineers and especially more
management turns the progress rate back down. Approaching a man-year
per square inch.
It\'s a fairly well known problem

https://en.wikipedia.org/wiki/The_Mythical_Man-Month

talks about software engineering, but it reflects a general problem - the more people you\'ve got working on a project, the more time gets spent on making all the contributions fit together (or on complaining that they can\'t be made to fit).

> It\'s sad how they have to keep buying the old stuff from us.

They have my deepest sympathies. One of the problems of replacing a bought-in solution is that the existence of that solution gives the design team tunnel vision.

If the problem is tackled from scratch, there\'s a fair chance that somebody will comes up with a rather different solution that wouldn\'t have evolved from up-grading the existing solution.

John\'s four month development time presumably reflects doing such an evolution. Throwing twenty engineers at the problem suggests the the customer was more ambitious, but if you can\'t find the right engineers having twenty engineers squabbling about the right way to go can waste a lot of time.

--
Bill Sloman, Sydney
 
B

Bill Sloman

Guest
On Saturday, November 28, 2020 at 2:27:16 AM UTC+11, John Larkin wrote:
On Thu, 26 Nov 2020 12:46:19 -0500, legg <le...@nospam.magma.ca> wrote:

On Sat, 21 Nov 2020 08:10:50 -0800, jla...@highlandsniptechnology.com
wrote:

The best designs are necessarily accidental.

If it\'s an accident, it\'s a discovery, not a design.

If a basic concept comes from a rarely or never explored place in the
solution space, it\'s a discovery. Deliberate, organized searches for
ideas tend to miss those places.
At least those searches organised by John Larkin may. Not that he\'s ever reported doing such a search.

> If it involves parts soldered to a board that people buy, it\'s a design.

If it\'s documented, you can tell people it\'s a design, even if you got there by a process of incremental modification, which is actually evolution, rather than design.

> The combination can be good.

If accident is the only way that you can get away from the familiar corner of the solution space, it\'s better than nothing. Design involves actually looking for alternate solutions, and usually means ripping up quite a few of them. Doing it before you\'ve soldered parts to boards is cheaper.

--
Bill Sloman, Sydney
 
W

whit3rd

Guest
On Friday, November 27, 2020 at 7:27:16 AM UTC-8, John Larkin wrote:

If a basic concept comes from a rarely or never explored place in the
solution space, it\'s a discovery. Deliberate, organized searches for
ideas tend to miss those places.
Huh? Deliberate, organized searches do NOT miss items in the
searched space. It\'s just that some folk miss possibilities
because of predispositions. This is a variant on the
<https://en.wikipedia.org/wiki/XY_problem>
classic XY scenario.

Deliberate, organized searches work just fine. Beware, though,
attempting to find a cheap D-to-A solution by searching IC merchants\'
offerings for \"DAC\" because that misses the R-2R ladder. It\'s not an IC...
 
P

Phil Hobbs

Guest
On 11/28/20 1:22 AM, whit3rd wrote:
On Friday, November 27, 2020 at 7:27:16 AM UTC-8, John Larkin wrote:

If a basic concept comes from a rarely or never explored place in the
solution space, it\'s a discovery. Deliberate, organized searches for
ideas tend to miss those places.

Huh? Deliberate, organized searches do NOT miss items in the
searched space. It\'s just that some folk miss possibilities
because of predispositions. This is a variant on the
https://en.wikipedia.org/wiki/XY_problem
classic XY scenario.
Looking for ideas like looking for your car keys? Riiiiiggghhhhtttt.
Deliberate, organized searches work just fine. Beware, though,
attempting to find a cheap D-to-A solution by searching IC merchants\'
offerings for \"DAC\" because that misses the R-2R ladder. It\'s not an IC...
Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
J

John Larkin

Guest
On Sat, 28 Nov 2020 12:47:26 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 11/28/20 1:22 AM, whit3rd wrote:
On Friday, November 27, 2020 at 7:27:16 AM UTC-8, John Larkin wrote:

If a basic concept comes from a rarely or never explored place in the
solution space, it\'s a discovery. Deliberate, organized searches for
ideas tend to miss those places.

Huh? Deliberate, organized searches do NOT miss items in the
searched space. It\'s just that some folk miss possibilities
because of predispositions. This is a variant on the
https://en.wikipedia.org/wiki/XY_problem
classic XY scenario.

Looking for ideas like looking for your car keys? Riiiiiggghhhhtttt.

Deliberate, organized searches work just fine. Beware, though,
attempting to find a cheap D-to-A solution by searching IC merchants\'
offerings for \"DAC\" because that misses the R-2R ladder. It\'s not an IC...


Cheers

Phil Hobbs
Inventing things and finding bugs in things are similar mental
processes. Both require letting your brain diffuse far and wide, and
not trusting data sheets or experimental observations or what
everybody knows. Both can be done in background, like when you are
asleep.

But you don\'t find novel ideas (or car keys) if you keep looking
incrementally in all the usual places.

I once found my car keys underwater, under a layer of ice, in the
parking lot at Sugar Bowl. That took a while, and the electronics in
the fob was destroyed. That\'s about a $270 or so fix for an Audi.


--

John Larkin Highland Technology, Inc trk

The cork popped merrily, and Lord Peter rose to his feet.
\"Bunter\", he said, \"I give you a toast. The triumph of Instinct over Reason\"
 
B

Bill Sloman

Guest
On Sunday, November 29, 2020 at 6:57:17 AM UTC+11, John Larkin wrote:
On Sat, 28 Nov 2020 12:47:26 -0500, Phil Hobbs
pcdhSpamM...@electrooptical.net> wrote:

On 11/28/20 1:22 AM, whit3rd wrote:
On Friday, November 27, 2020 at 7:27:16 AM UTC-8, John Larkin wrote:

If a basic concept comes from a rarely or never explored place in the
solution space, it\'s a discovery. Deliberate, organized searches for
ideas tend to miss those places.

Huh? Deliberate, organized searches do NOT miss items in the
searched space. It\'s just that some folk miss possibilities
because of predispositions. This is a variant on the
https://en.wikipedia.org/wiki/XY_problem
classic XY scenario.

Looking for ideas like looking for your car keys? Riiiiiggghhhhtttt.

Deliberate, organized searches work just fine. Beware, though,
attempting to find a cheap D-to-A solution by searching IC merchants\'
offerings for \"DAC\" because that misses the R-2R ladder. It\'s not an IC...

Inventing things and finding bugs in things are similar mental
processes. Both require letting your brain diffuse far and wide, and
not trusting data sheets or experimental observations or what
everybody knows.
This may be true for John Larkin. I\'ve inverted stuff that has got patented, and it doesn\'t line up with my experience.

> Both can be done in background, like when you are asleep.

Once you\'ve put in the hard work to get the information (lots of it) into your head, your sub-conscious will work on it in background.
My father had 25 patents, and claimed that his best ideas came to him when he was shaving. He was well aware that he\'d been sleeping on the problem, and that the answer that \"came to him\" had been sitting in a queue.

But you don\'t find novel ideas (or car keys) if you keep looking
incrementally in all the usual places.
Obviously not, unless \"incrementally\" means progressively increasing the area searched - as opposed to going back over areas already searched, time and time again.

<snip>

--
Bill Sloman, Sydney
 
Toggle Sidebar

Welcome to EDABoard.com

Sponsor

Top