What characterizes a powerFET for audio use?

[Phil Allison]

"Shaun"

I used to work at an Industrial and Instrumentation repair and Calibration
shop that was licensed by Beckman to repair and calibrate there DMMs. In
the good ones, the RMS converted used a thermal method to determine RMS
Voltage and Current. That was 18 years ago and I don't remember any
details except that they used thermal measurement on a heated element. I
have compared the reading on the Beckman - Then Wavetek - now I don't know
if anyone is still making them, to a Fluke RMS DMM at 60 Hz and DC and the
readings matched.

** No fooling...........

I have always heard that if you really want to know the real value, you
have to buy an RMS meter.

** That sort of nonsense *would* be said by someone who either owned an RMS
meter or was selling them.

Thermal RMS meters are virtually obsolete these days and have been replaced
by cheap analogue RMS to DC computation ICs in many hand held DMMs OR by
digital sampling computation in most DSOs.

The hand held kind have a limited measurement bandwidth compared to the
latter.

Depends what your needs are.

As Clint Eastwood might have said -

" Man's gotta know the limitations of his test equipment ".




..... Phil

 

[dave]

On 01/02/2014 04:07 PM, Maynard A. Philbrook Jr. wrote:

In article <B4GdnYZ19saY4VjPnZ2dnUVZ_tudnZ2d@earthlink.com>,
ricketzz@earthlink.net says...

On 01/02/2014 04:05 AM, William Sommerwerck wrote:
"Jeff Liebermann" wrote in message
news:nsj9c95bm2ec1giaju7tj0gc43ne25s007@4ax.com...

http://www.eznec.com/Amateur/RMS_Power.pdf
On Pg 6 it says:
"The RMS value of power is not the equivalent heating power and,
in fact, it doesn't represent any useful physical quantity."
and:
"The RMS power is different than the average power, and therefore
isn?t the equivalent heating power. In fact, the RMS value of the
power doesn?t represent anything useful."

I remember a Popular Electronics quiz with questions about how a
voltmeter (at that time, a moving-coil device) would read, depending on
the waveform supplied. A key point was that the deflection was
proportional to the average current flowing through the coil, but the
meter was usually calibrated for the RMS value of a sinewave.

Why do you keep quoting notes from Llewelen? He wrote a modeling program
for wire antennas. We are talking about FETs.

I think AVG = RMS only if waveform is pure sinewave. Otherwise you must
use calorimeter or thermocouple to find true RMS.

AVG and RMS is not the same with sine waves..

Then they never are the same?

 

[dave]

On 01/02/2014 05:04 PM, Phil Allison wrote:

"Shaun"

I used to work at an Industrial and Instrumentation repair and Calibration
shop that was licensed by Beckman to repair and calibrate there DMMs. In
the good ones, the RMS converted used a thermal method to determine RMS
Voltage and Current. That was 18 years ago and I don't remember any
details except that they used thermal measurement on a heated element. I
have compared the reading on the Beckman - Then Wavetek - now I don't know
if anyone is still making them, to a Fluke RMS DMM at 60 Hz and DC and the
readings matched.

** No fooling...........

I have always heard that if you really want to know the real value, you
have to buy an RMS meter.

** That sort of nonsense *would* be said by someone who either owned an RMS
meter or was selling them.

Thermal RMS meters are virtually obsolete these days and have been replaced
by cheap analogue RMS to DC computation ICs in many hand held DMMs OR by
digital sampling computation in most DSOs.

The hand held kind have a limited measurement bandwidth compared to the
latter.

Depends what your needs are.

As Clint Eastwood might have said -

" Man's gotta know the limitations of his test equipment ".




.... Phil

In television you calibrate the output power metering by operating into
a full power liquid cooled dummy load and measuring the temperature of
the coolant.

Using computational methods on amplifier input parameters is not the
preferred way to determine power (an efficiency factor must be assumed)
and you must ask the FCC for special permission to use the "direct method".

Dave the Re-Re

 

[Michael A. Terrell]

William Sommerwerck wrote:

"Jeff Liebermann" wrote in message
news:quc9c95b8el5pqevbftgm87mqs5cf8akrh@4ax.com...

OK, I yield to a higher authority. Heating power is average power:
http://www.eznec.com/Amateur/RMS_Power.pdf
However that doesn't stop nearly everyone from specifying
RMS power.

Unfortunately, no.

So, why do people buy true-RMS voltmeters? If they don't measure
anything useful, why bother?

A fair question. Apparently, there was a time when AC power had a significant
percentage of harmonics (perhaps it still does), and engineers wanted to know
its "true" heating effect.

This is my favorite True RMS meter. Section three is a description
of how it works. The AC power line is anything but a clean sine wave.
Switching power supplies and older electronics with a single diode in
the power supply distort it, even more. The single diode causes a DC
component on your power, which confuses a simple analog meter. Three
phase is even worse, with all the harmonics on the neutral. There were a
lot of fires in office buildings and factories after the PC became
widespread. It used to be to code to use a smaller neutral on three
phase, than the tree phase lines. The harmonics caused the neutral to
overheat and in some cases, started fires.

I like this Fluke for many reasons, including the ability to read to
..01 dB variations in a signal. I used one to test the -3dB points on
video filters, up to 20 MHz. For anything higher I used a Boonton 9200
RF voltmeter.

Before someone smarts off, Video isn't just television. We produced
equipment with selectable bandwidth from DC>10 KHz to DC>40 MHz in 16
customer specified bandwidths. SDI digital video was 270 MHz, and HD is
even higher.

<http://www.ko4bb.com/Manuals/Fluke/Fluke_8920A_8921A_True_RMS_Voltmeter_Operation_and%20_Service_Manual_Oct78.pdf>


--
Anyone wanting to run for any political office in the US should have to
have a DD214, and a honorable discharge.

 

[Phil Allison]

"dave the autistic fool " <ricketzz@earthlink.net>

As Clint Eastwood might have said -

" Man's gotta know the limitations of his test equipment ".


In television you calibrate the output power metering by operating into a
full power liquid cooled dummy load and measuring the temperature of the
coolant.

** That would be for the actual transmitter on VHF or UHF - right ?

Same idea as using a glass of water and a thermometer test the power RF
power of a microwave oven.

Got SFA to do with the topic.



..... Phil

 

[Phil Allison]

"dave the pig ignorant autistic IDIOT "

The question was why is audio power spec is half that of DC on/off
capability.

** No it was not.

It is because the device stays on all the time and because the audio
waveform causes more heating.

** A maker's max dissipation spec for a semiconductor is not related to the
application.

It is only related to the device itself.

Normally, the figure given is for *ideal* operating conditions - like when
mounted on an infinite heatsink held at 20C.



..... Phil

 

[dave]

On 01/03/2014 07:33 PM, Phil Allison wrote:

"dave the pig ignorant autistic IDIOT"


The question was why is audio power spec is half that of DC on/off
capability.

** No it was not.


It is because the device stays on all the time and because the audio
waveform causes more heating.


** A maker's max dissipation spec for a semiconductor is not related to the
application.

It is only related to the device itself.

Normally, the figure given is for *ideal* operating conditions - like when
mounted on an infinite heatsink held at 20C.



.... Phil

Odd. I thought Maximum values were on the data sheet. Anyway, a MOSFET
used as a switch makes less heat so it can have a higher current rating
than one used class AB.

 

[Michael A. Terrell]

dave wrote:

On 01/03/2014 07:33 PM, Phil Allison wrote:
"dave the pig ignorant autistic IDIOT"


The question was why is audio power spec is half that of DC on/off
capability.

** No it was not.


It is because the device stays on all the time and because the audio
waveform causes more heating.


** A maker's max dissipation spec for a semiconductor is not related to the
application.

It is only related to the device itself.

Normally, the figure given is for *ideal* operating conditions - like when
mounted on an infinite heatsink held at 20C.



.... Phil

Odd. I thought Maximum values were on the data sheet. Anyway, a MOSFET
used as a switch makes less heat so it can have a higher current rating
than one used class AB.

'Safe Operating Area' is more important than individual maximums.



--
Anyone wanting to run for any political office in the US should have to
have a DD214, and a honorable discharge.

 

[Leif Neland]

Phil Allison skrev:

Same idea as using a glass of water and a thermometer test the power RF power
of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power
on.

Measure the temperature of the cold water, then measure the time for it
to boil.

Preferably in a microwave-safe plastic container.

I measured the power of my previous oven to 230W...

Leif

--
Husk křrelys bagpĺ, hvis din bilfabrikant har taget den idiotiske
beslutning at undlade det.

 

[mike]

On 1/4/2014 8:08 AM, Leif Neland wrote:

Phil Allison skrev:

Same idea as using a glass of water and a thermometer test the power
RF power of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power on.

Measure the temperature of the cold water, then measure the time for it
to boil.

Preferably in a microwave-safe plastic container.

I measured the power of my previous oven to 230W...

Leif

When I did similar tests, I found the order of magnitude was right,
but the result depended on the shape and volume of the mass of water
and position in the oven.

It's easy to imagine that the microwaves bounce around and most of the
energy ends up
in the water. Wonder how accurate that model?
Wonder what the "official" water configuration is when they determine
the spec?

 

[Shaun]

"mike" wrote in message news:la9cdd$qov$1@dont-email.me...

On 1/4/2014 8:08 AM, Leif Neland wrote:

Phil Allison skrev:

Same idea as using a glass of water and a thermometer test the power
RF power of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power on.

Measure the temperature of the cold water, then measure the time for it
to boil.

Preferably in a microwave-safe plastic container.

I measured the power of my previous oven to 230W...

Leif

When I did similar tests, I found the order of magnitude was right,
but the result depended on the shape and volume of the mass of water
and position in the oven.

It's easy to imagine that the microwaves bounce around and most of the
energy ends up
in the water. Wonder how accurate that model?
Wonder what the "official" water configuration is when they determine
the spec?

I've seen that method used before. I had an article about measuring
microwave oven power. You of course measure the increase in temperature
after running then microwave oven for a predetermine time at full power,
with a measured amount of water (distilled) in the direct center of the
oven. You could google it.

Shaun

 

[Phil Allison]

"Leif Neland"

Phil Allison
Same idea as using a glass of water and a thermometer test the power RF
power of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power on.

** Should be OK with a glass tube and red liquid type.

Measure the temperature of the cold water, then measure the time for it to
boil.

** Bad idea.

Takes far too long, when boiling first begins is not clear and lots of heat
is lost to the air and evaporation.


> I measured the power of my previous oven to 230W...

** Using half a litre in a plastic jug for two minutes, I got the answer to
within 10% with a 700W rated oven.

Having a K-type bead thermocouple and digital temp meter made the job easier
too.



..... Phil

 

[Shaun]

"Shaun" wrote in message newsk3yu.184114$Ks5.25080@fx08.iad...



"Phil Allison" wrote in message news:birqprFai6pU1@mid.individual.net...


"Leif Neland"

Phil Allison
Same idea as using a glass of water and a thermometer test the power RF
power of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power on.

** Should be OK with a glass tube and red liquid type.

Measure the temperature of the cold water, then measure the time for it to
boil.

** Bad idea.

Takes far too long, when boiling first begins is not clear and lots of heat
is lost to the air and evaporation.

I measured the power of my previous oven to 230W...

** Using half a litre in a plastic jug for two minutes, I got the answer to
within 10% with a 700W rated oven.

Having a K-type bead thermocouple and digital temp meter made the job
easier
too.

You don't run it till boiling! once you get close to boiling point a lot of
extra energy is required to raise it further and make it boil. What you do
is run a glass of cold distilled water measured (temp and volume) in a
container, you could use several stacked Styrofoam cups for insulation and
cover the top with Styrofoam so that the heat generated does not escape and
run the oven till the temperature increase 20 to 50 degrees or so, then
measure the temp, the information will have an equation to convert degrees
rise to microwave power. Google the method - I haven't looked it up lately.

Shaun



Here is the Method from RepairFAQ from Sam:



7.1) Testing the oven - the water heating test


The precise number of degrees a known quantity of water increases in
temperature for a known time and power level is a very accurate test of
the actual useful microwave power. A couple of minutes with a cup of
water and a thermometer will conclusively determine if your microwave
oven is weak or you are just less patient (or the manufacturer of your
frozen dinners has increased their weight - sure, fat chance of that!)

You can skip the heavy math below and jump right to the final result
if you like. However, for those who are interested:

* 1 Calorie (C) will raise the temperature of 1 gram (g) of liquid water
exactly 1 degree Centigrade (DegC) or 9/5 degree Fahrenheit (DegF).

* 1 Calorie is equal to 4.184 Joules (J) or 1 J = .239 C.

* 1 Watt (W) of power is 1 J/s or 1 KW is 1000 J/s.

* 1 cup is 8 ounces (oz) which is 8 x 28.35 g/oz = 226.8 g.

* 1 minute equals 60 s (but you know this!).

Therefore, in one minute, a 1 KW microwave oven will raise the temperature
of 1 cup of water by:

T(rise) = (60 s * 1000 J/s * .239C/J * (g * DegC)/C)/(226.8 g) = 63
DegC.

Or, if your prefer Fahrenheit: 114 DegF.

To account for estimated losses due to conduction, convection, and imperfect
power transfer, I suggest using temperature rises of 60 DegC and 109 DegF.

Therefore, a very simple test is to place a measured cup of water in the
microwave from the tap and measure its temperature before and after heating
for exactly 1 minute on HIGH. Scale the expected temperature rise by the
ratio of the microwave (not AC line) power of your oven compared to a 1 KW
unit.

Or, from a Litton microwave handbook:

Heat one Liter (L) of water on HIGH for 1 minute.

Oven power = temperature rise in DegC multiplied by 70.

Use a plastic container rather than a glass one to minimize the needed
energy loss to raise its temperature by conduction from the hot water.
There will be some losses due to convection but this should not be that
significant for these short tests.

(Note: if the water is boiling when it comes out - at 100 DegC or 212 DegF,
then the test is invalid - use colder water or a shorter time.)

The intermediate power levels can be tested as well. The heating effect of
a microwave oven is nearly linear. Thus, a cup of water should take nearly
roughly twice as long to heat a specific number of degrees on 50% power or
3.3 times as long on 30% power as on full power. However, for low power
tests, increasing the time to 2 minutes with 2 cups of water will result
in more accurate measurements due to the long period pulse width power
control use by microwave ovens which may have a cycle of up to 30 seconds.

Any significant discrepancy between your measurements and the specified
microwave power levels - say more than 10 % on HIGH - may indicate a
problem.
(Due to conduction and convection losses as well as the time required to
heat the filament of the magnetron for each on-cycle, the accuracies of
the intermediate power level measurements may be slightly lower).

Shaun

 

[dave]

On 01/04/2014 06:21 PM, Shaun wrote:

"Phil Allison" wrote in message news:birqprFai6pU1@mid.individual.net...


"Leif Neland"
Phil Allison
Same idea as using a glass of water and a thermometer test the power
RF power of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power on.

** Should be OK with a glass tube and red liquid type.


Measure the temperature of the cold water, then measure the time for
it to boil.

** Bad idea.

Takes far too long, when boiling first begins is not clear and lots of
heat
is lost to the air and evaporation.


I measured the power of my previous oven to 230W...

** Using half a litre in a plastic jug for two minutes, I got the
answer to
within 10% with a 700W rated oven.

Having a K-type bead thermocouple and digital temp meter made the job
easier
too.

You don't run it till boiling! once you get close to boiling point a
lot of extra energy is required to raise it further and make it boil.
What you do is run a glass of cold distilled water measured (temp and
volume) in a container, you could use several stacked Styrofoam cups for
insulation and cover the top with Styrofoam so that the heat generated
does not escape and run the oven till the temperature increase 20 to 50
degrees or so, then measure the temp, the information will have an
equation to convert degrees rise to microwave power. Google the method
- I haven't looked it up lately.

Shaun



.... Phil

That is not how a calorimeter works. The load must be enclosed in liquid
and must perfectly match the RF output so it absorbs all the energy.
Then the temperature should be a very accurate way to measure power. It
must be a closed, water cooled load.

http://electro-impulse.com/techinfo/calorimeters.htm

 

[Shaun]

"Phil Allison" wrote in message news:birqprFai6pU1@mid.individual.net...


"Leif Neland"

Phil Allison
Same idea as using a glass of water and a thermometer test the power RF
power of a microwave oven.

Just don't leave the thermometer in the microwave oven with the power on.

** Should be OK with a glass tube and red liquid type.

Measure the temperature of the cold water, then measure the time for it to
boil.

** Bad idea.

Takes far too long, when boiling first begins is not clear and lots of heat
is lost to the air and evaporation.

I measured the power of my previous oven to 230W...

** Using half a litre in a plastic jug for two minutes, I got the answer to
within 10% with a 700W rated oven.

Having a K-type bead thermocouple and digital temp meter made the job
easier
too.

You don't run it till boiling! once you get close to boiling point a lot of
extra energy is required to raise it further and make it boil. What you do
is run a glass of cold distilled water measured (temp and volume) in a
container, you could use several stacked Styrofoam cups for insulation and
cover the top with Styrofoam so that the heat generated does not escape and
run the oven till the temperature increase 20 to 50 degrees or so, then
measure the temp, the information will have an equation to convert degrees
rise to microwave power. Google the method - I haven't looked it up lately.

Shaun



..... Phil

 

[Jeff Urban]

I wanted to reply to this when you first posted it but I couldn't for
whatever reason. fucking thing. As you can see I am not posting from
google, which I no longer capitalize !!!! LOL

Anyway, what characterizes ANY part for the transmission of audio is
linearity. Even though the gate of a MOSFET is driven wildly different
than the base of a BPT, linearity of gain fro the very small signal to
the very large signal is the prime. We used to look at the hfe and HFE
gain curves of bipolars in the old days, not it is different.

Any kinds of spurious shit like oscillations would be no good in
switcxhers as well. the thing is, switchers somethimes have a gain curve
that is like, made to be on or off.

An audio transistor must operate in the analog range, that means the
gain curve should be as flat as possible through the operating current
range.

The same is true of a MOSFET.

 

[N_Cook]

On 05/01/2014 07:53, Jeff Urban wrote:

I wanted to reply to this when you first posted it but I couldn't for
whatever reason. fucking thing. As you can see I am not posting from
google, which I no longer capitalize !!!! LOL

Anyway, what characterizes ANY part for the transmission of audio is
linearity. Even though the gate of a MOSFET is driven wildly different
than the base of a BPT, linearity of gain fro the very small signal to
the very large signal is the prime. We used to look at the hfe and HFE
gain curves of bipolars in the old days, not it is different.

Any kinds of spurious shit like oscillations would be no good in
switcxhers as well. the thing is, switchers somethimes have a gain curve
that is like, made to be on or off.

An audio transistor must operate in the analog range, that means the
gain curve should be as flat as possible through the operating current
range.

The same is true of a MOSFET.

Ah at last , thanks, a pertinent reply to my original question. I was
wondering if it was a larger area of silicon so the heat can migrate out
of the die quicker. So from what you say a non-audio switcher mosfet
could be used for analogue but the power rating would have to be derated
and no other qualification for such use, anymore than usual precautions
you would use for a switcher situation. I suppose the amount of derating
would then depend on the type of use bass amp v GP audio amp, dance
music v classical music etc

 

[Phil Allison]

"Nutcase Kook is yet another pig ignorant pommy cunt"

Other than p channel in this case, same for BUZ901P nch
eg BUZ906P 200V, 8A ,datasheet says
"POWER MOSFETS FOR
AUDIO APPLICATIONS"
but also
"FEATURES ... (for use in)
HIGH SPEED SWITCHING ... "

** The Semelab app note makes it pretty clear there is a HUGE difference
between "switching" and audio ( ie lateral) power mosfets.

http://products.semelab-tt.com/pdf/ApplicationNoteAlfet.pdf

Would a powerFET designed solely for high speed switching use and 125W
rating be derated in power handling terms to only 50W say for linear 10 Hz
use. Or secondary oscillation liability if paralleled up devices? or some
other operational failing in audio use not found with smps say ?

** Yawnnnnnnnnnnnnnnn....

More brainless, fucking TROLLING !!!!!!!

FOAD you vile pommy cunthead.



..... Phil

Ah at last , thanks, a pertinent reply to my original question.

** There is no possible answer to a question as wrong headed and stupid as
that.

So you will never get one.

FOAD you vile pommy cunthead.




..... Phil

 

[N_Cook]

On 05/01/2014 07:53, Jeff Urban wrote:

I wanted to reply to this when you first posted it but I couldn't for
whatever reason. fucking thing. As you can see I am not posting from
google, which I no longer capitalize !!!! LOL

Anyway, what characterizes ANY part for the transmission of audio is
linearity. Even though the gate of a MOSFET is driven wildly different
than the base of a BPT, linearity of gain fro the very small signal to
the very large signal is the prime. We used to look at the hfe and HFE
gain curves of bipolars in the old days, not it is different.

Any kinds of spurious shit like oscillations would be no good in
switcxhers as well. the thing is, switchers somethimes have a gain curve
that is like, made to be on or off.

An audio transistor must operate in the analog range, that means the
gain curve should be as flat as possible through the operating current
range.

The same is true of a MOSFET.

or on rereading . If an audio application can tolerate a certain amount
of cross-over distortion and general harmonic distortion then there is
no difference in powerfet useage type , up to some power level where
these distortions become too apparent.

 

Shaun <stereobuff07@gmail.com> wrote:

When I did similar tests, I found the order of magnitude was right,
but the result depended on the shape and volume of the mass of water
and position in the oven. [...]

Wonder what the "official" water configuration is when they determine
the spec?

I don't know about a US Federal Trade Commission (or other agency, or
equivalent in other countries) test procedure; there probably is one but
I don't know it.

I know that older GE microwave ovens, in the little service information
sheet that was folded up inside the oven, gave directions on the test, a
GE part number for a beaker you were supposed to use, and I think a
third-party part number for the thermometer you were supposed to use to
measure the water temperature before and after. I think the water level
was marked on the beaker, and the service sheet said to put it right in
the middle of the oven. You didn't boil it, just heated it for a fixed
amount of time.

If I remember right, the criteria was something like "if the oven
produced between X and Y degrees temperature rise in the water, it's
within spec" - it didn't give you an answer in watts, just an acceptable
temperature range for that particular model oven.

Matt Roberds