How to bias a MOSFET amp?

[Zak]

Tam/WB2TT wrote:

You are only scratching the surface. Check out the Harris DX series of high
power AM transmitters. It will blow your mind away. Basically, the
instantanous RF output power is synthesized by turning on 0 to 64 fairly low
power (~KW) modules. I don't know what the sampling frequency is, but
probably >20 KHz.. All modules are driven by a square wave signal at the
carrier frequency. There is no modulator.

There is also a fractional stage - a 64 stage AM modulation would sound
quite nasty, so an analogue signal is added to make up.

There is also a 'spare stage' dthat can be switched in if one of the
stages fails.

Nice stuff, but the only really interesting thing IMO is the output
combiner. The rest is just 'how do we make this digital'.


Thomas

 

[John Fields]

On Sat, 28 Aug 2004 19:28:03 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

On Sat, 28 Aug 2004 12:39:34 -0500, John Fields
jfields@austininstruments.com> wrote:

On Sat, 28 Aug 2004 18:07:06 +0100, Paul Burridge
pb@notthisbit.osiris1.co.uk> wrote:


Actually I've built *several* class C RF amps, John.

---
Intentionally???

Aha! very amusing. Most of them have been intentional, yes, but who
here can say they haven't ended up at some point with something they
hadn't bargained for?

However, I
wouldn't call any of them linear. You will be aware than linearity
starts to go out of the window when Class A slides into Class AB and
beyond.

---
Really? I'd _love_ to hear your explanation for why that "happens".

I won't bore you with explanations you're already well acquainted
with.

---
On the contrary, I'd like to hear why you think class AB or B isn't
(or can't be) linear, input-to-output.
---

But I'm still reeling from the revelation that you confused AM
with Class C. :-/

---
Go back and read it again in the context of "is that a smart thing to
do?" with your tongue-in-cheek detector energized.
---

I've heard a lot of amps that sounded pretty good at both low and high
volumes, and in between, and they've almost all had class AB outputs.

I'm sure you have. But even class A isn't perfect. The pitfalls of
large-signal handling and all that. Do you know of an active device
with a *perfectly* linear transconductance between say 0 and 20V? No?
I thought not...

---
We weren't talking about components with perfectly linear transfer
functions, _you_ were alluding to deterioration of input-to-output
linearity in systems using different driver biasing schemes. Or so I
thought, when you said:

"You will be aware than linearity starts to go out of the window when
Class A slides into Class AB and beyond."

And I'll repeat:

"I'd _love_ to hear your explanation for why that "happens".".


But never mind, now that Woodgate's cleared it up there's no need for
you to embarrass yourself further.

--
John Fields

 

[Active8]

On Sat, 28 Aug 2004 19:20:06 +0100, John Woodgate wrote:

<snip>

With a **tuned load**, the output power also depends more or less
linearly on the supply voltage, so amplitude modulation can be achieved
by varying the supply voltage.

Just an obsevation... when you enclose a word or words in double
asterisks, my reader no longer renders them in bold, thus removing
the intended emphasis. IIRC it's the double underscore enclosure
that causes my reader to underline the text.

We now return you to our regularly scheduled terminology debate.

--
Best Regards,
Mike

 

[John Woodgate]

I read in sci.electronics.design that Active8 <reply2group@ndbbm.net>
wrote (in <1e6iv8lnwr2yt$.dlg@news.individual.net&gt about 'How to bias a
MOSFET amp?', on Sat, 28 Aug 2004:

Just an obsevation... when you enclose a word or words in double
asterisks, my reader no longer renders them in bold, thus removing the
intended emphasis. IIRC it's the double underscore enclosure that causes
my reader to underline the text.

You should use a plain text reader, as Manitou intended.(;-)
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Rich Grise]

dbowey@aol.com (Dbowey) wrote in message

jfields posted:
On Sat, 28 Aug 2004 10:36:21 +0100, Paul Burridge
pb@notthisbit.osiris1.co.uk> wrote:
On Fri, 27 Aug 2004 17:43:34 -0700, John Larkin

Have you actually built a class C linear RF power amp? Tell us how it
works.

It depends on how you define "linear" basically. But the term is a
total misnomer in RF amp terminology and very misleading. I can't
understand how it got there. :-/
....
So, that should take the mystery out of why it's called a
"linear amplifier"


A class C amp is not a "linear amp" and has never been called that within the
big circle of those who work with RF. Its very non-linearity is what permits
plate modulation; it is a mixer/multiplier - pick your favorite term.

It's cool that you mention that - I saw a construction article for an HF-VHF
upconverter, with a couple of those VHF tubes similar to a 6146 or something
like that - about 150 watts output. The interesting thing was, the plate supply
for the converter was the RF from the "exciter", an ordinary 150 watt or so
transmitter. (actually, it was a transceiver, but they used a different
converter for the receiver channel. ;-) ) The LO went to the grids, and of
course, the plate tank was tuned to the output freq. Pretty easy to filter,
too, when the images are 28 MHz apart. ;-)

Cheers!
Rich

I believe confusion comes from some books describing the process as linear, and
explaining that as being because the sidebands are an "exact" replica of the
modulating voltage.

Don

 

[Dave VanHorn]

Dad just bought a 2kW FET amplifier. It's ex-industrial use, but has MRF
line transistors, and we expect no trouble putting it online for 160-10
meter use.
It's a dozen MRF-150s

Look up the specs (and the prices (OUCH)) on the MRF-154.

--
KC6ETE Dave's Engineering Page, www.dvanhorn.org
Microcontroller Consultant, specializing in Atmel AVR

 

[Tim Wescott]

Paul Burridge wrote:

On Fri, 27 Aug 2004 17:43:34 -0700, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:


Have you actually built a class C linear RF power amp? Tell us how it
works.


It depends on how you define "linear" basically. But the term is a
total misnomer in RF amp terminology and very misleading. I can't
understand how it got there. :-/

From basic systems theory:

A system S is linear if and only if for any two input signals x1 and x2
that generate the output signals y1 = S(x1) and y2 = S(x2), and for any
two real constants A1 and A2, the output signal y = S(A1*x1 + A2*x2) is
equal to A1*y1 + A2*y2.

This condition is approached with a properly adjusted RF linear
amplifier, even one operated class AB or B. It is _not_ approached with
a class C amplifier.

So if you define "linear" the way electronics engineering professionals
define "linear" a class A, AB or B amplifier can be made to act
linearly, more or less, and a class C amplifier cannot. So the term
isn't a misnomer, and its use is obvious.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Tam/WB2TT]

"Zak" <jute@zak.invalid> wrote in message
news:EK5Yc.6906$vH5.6515@amstwist00...

Tam/WB2TT wrote:

You are only scratching the surface. Check out the Harris DX series of
high
power AM transmitters. It will blow your mind away. Basically, the
instantanous RF output power is synthesized by turning on 0 to 64 fairly
low
power (~KW) modules. I don't know what the sampling frequency is, but
probably >20 KHz.. All modules are driven by a square wave signal at the
carrier frequency. There is no modulator.

There is also a fractional stage - a 64 stage AM modulation would sound
quite nasty, so an analogue signal is added to make up.

There is also a 'spare stage' dthat can be switched in if one of the
stages fails.

Nice stuff, but the only really interesting thing IMO is the output
combiner. The rest is just 'how do we make this digital'.


Thomas

Yeah, exactly what I was wondering about, how do they combine the outputs?
They imply that they are combining power, but I suspect they are adding
voltages. BTW, the modules are not all the same. As I recall fom the WLW web
site, 60 are 1-2 KW, 1 is 500W, 2 are 200W, and 1 is 100W. They claim
something like 90% efficiency from 3 phase 60~ in to RF out. Also, there
must be some huge bandpass filter on the output to suppress the sampling
frequency.

Tam

 

[legg]

On Sat, 28 Aug 2004 10:36:21 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

On Fri, 27 Aug 2004 17:43:34 -0700, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

Have you actually built a class C linear RF power amp? Tell us how it
works.

It depends on how you define "linear" basically. But the term is a
total misnomer in RF amp terminology and very misleading. I can't
understand how it got there. :-/

I'd always understood that it refered to an operating frequency range
with a wide bandwidth, therefore suited to FM or frequency hopping.

RL

 

[Ken Smith]

In article <1$so5AEWzMMBFw6N@jmwa.demon.co.uk>,
John Woodgate <noone@yuk.yuk> wrote:
[...]

With a **tuned load**, the output power also depends more or less
linearly on the supply voltage, so amplitude modulation can be achieved
by varying the supply voltage.

Small quibble:

you need to add "for a large enough drive" to the above. For small drive
levels the drive and the device gm control the output power.

--
--
kensmith@rahul.net forging knowledge

 

[Ken Smith]

In article <1$so5AEWzMMBFw6N@jmwa.demon.co.uk>,
John Woodgate <noone@yuk.yuk> wrote:
[...]

But with a **tuned load**, 'linearity' can be achieved even with Class C
biasing. This is why linearity in this case is defined as output power
being proportional to input power.

I think this is right:

If this isn't the slow modulation case, for a single device RF output
stage to work as a linear you need to bias it to about the point where gm
is 1/2 the "mid current" gm. This makes the stage class AB.

If you want to go towards class B or C there are things you can do in the
slow modulation case. Basically as the RF input increases, you move into
class C and as it decreases you move back towards AB. Moving to class C
lowers the output power for large signals a bit so the AB class's standing
current can be reduced by some amount.


--
--
kensmith@rahul.net forging knowledge

 

[Ken Smith]

In article <5z5Yc.3257$6o3.2610@newsread2.news.atl.earthlink.net>,
Ralph Mowery <rmowery28146@earthlink.net> wrote:
[...]

Any class ( A, B , C ) of amp can be plate modulated for AM. It is then
not really an amplifier.

I disagree with this. If the stage puts out more RF than it takes in, it
is an amplifier even if the purpose of the stage is to be a modulator.


--
--
kensmith@rahul.net forging knowledge

 

[John Woodgate]

I read in sci.electronics.design that Tim Wescott
<tim@wescottnospamdesign.com> wrote (in <10j2acioqmafqec@corp.supernews.
com&gt about 'How to bias a MOSFET amp?', on Sat, 28 Aug 2004:

A system S is linear if and only if for any two input signals x1 and x2
that generate the output signals y1 = S(x1) and y2 = S(x2), and for any
two real constants A1 and A2, the output signal y = S(A1*x1 + A2*x2) is
equal to A1*y1 + A2*y2.

This condition is approached with a properly adjusted RF linear
amplifier, even one operated class AB or B. It is _not_ approached with
a class C amplifier.

The point that the discussion has reached is that the term 'linear' IS
used in a different sense that includes Class C, whether we like it or
not.

So if you define "linear" the way electronics engineering professionals
define "linear" a class A, AB or B amplifier can be made to act
linearly, more or less, and a class C amplifier cannot. So the term
isn't a misnomer, and its use is obvious.

These are the 'electronic engineering professionals' who have 49
different definitions of 'level' in the International Electrotechnical
Vocabulary? (;-)
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Paul Burridge]

On Sun, 29 Aug 2004 00:48:51 +0000 (UTC), kensmith@green.rahul.net
(Ken Smith) wrote:

In article <5z5Yc.3257$6o3.2610@newsread2.news.atl.earthlink.net>,
Ralph Mowery <rmowery28146@earthlink.net> wrote:
[...]
Any class ( A, B , C ) of amp can be plate modulated for AM. It is then
not really an amplifier.

I disagree with this. If the stage puts out more RF than it takes in, it
is an amplifier

By that definition, it could also be an oscillator!
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[Paul Burridge]

On Sun, 29 Aug 2004 00:46:11 +0000 (UTC), kensmith@green.rahul.net
(Ken Smith) wrote:

If this isn't the slow modulation case, for a single device RF output
stage to work as a linear you need to bias it to about the point where gm
is 1/2 the "mid current" gm. This makes the stage class AB.

If you want to go towards class B or C there are things you can do in the
slow modulation case. Basically as the RF input increases, you move into
class C and as it decreases you move back towards AB. Moving to class C
lowers the output power for large signals a bit so the AB class's standing
current can be reduced by some amount.

Can't agree. If you're running any amount of standing current, then it
ain't class C.
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[Paul Burridge]

On Sun, 29 Aug 2004 02:10:52 GMT, legg <legg@nospam.magma.ca> wrote:

On Sat, 28 Aug 2004 10:36:21 +0100, Paul Burridge
pb@notthisbit.osiris1.co.uk> wrote:

On Fri, 27 Aug 2004 17:43:34 -0700, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

Have you actually built a class C linear RF power amp? Tell us how it
works.

It depends on how you define "linear" basically. But the term is a
total misnomer in RF amp terminology and very misleading. I can't
understand how it got there. :-/

I'd always understood that it refered to an operating frequency range
with a wide bandwidth, therefore suited to FM or frequency hopping.

Curious. I have a book (by an author whose name I can't recall off
hand) called "High-Linearity RF Amplifier Design" in which said author
claims that a linear RF amplifier is one in which the transmitted
signal, regardless of however non-linear it may be, is passed through
sufficient filtering to clean it up! Work that one out if you can!
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[John Larkin]

On Sun, 29 Aug 2004 12:49:46 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

Can't agree. If you're running any amount of standing current, then it
ain't class C.

Tube class-C amps often self-bias (via grid rectification) when
operating normally. Without drive, they often have high, sometimes
dangerous, idle currents.

John

 

[Paul Burridge]

On Sun, 29 Aug 2004 11:12:18 -0700, John Larkin
<jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

Tube class-C amps often self-bias (via grid rectification) when
operating normally. Without drive, they often have high, sometimes
dangerous, idle currents.

Toobz? I was talking of transistors. Many tranny designs for Class C
RF utilize self-bias, too, but there is still no quiescent current!
--

"What is now proved was once only imagin'd." - William Blake, 1793.

 

[John Larkin]

On Sun, 29 Aug 2004 19:34:57 +0100, Paul Burridge
<pb@notthisbit.osiris1.co.uk> wrote:

On Sun, 29 Aug 2004 11:12:18 -0700, John Larkin
jjlarkin@highlandSNIPtechTHISnologyPLEASE.com> wrote:

Tube class-C amps often self-bias (via grid rectification) when
operating normally. Without drive, they often have high, sometimes
dangerous, idle currents.

Toobz? I was talking of transistors. Many tranny designs for Class C
RF utilize self-bias, too, but there is still no quiescent current!

What makes an amp Class C is a low conduction angle while operating.
When it's not operating, is has no class at all.

John

 

[Ken Smith]

In article <7fg3j09sk1vlsi59ulbido2ctqi39evd96@4ax.com>,
Paul Burridge <pb@notthisbit.osiris1.co.uk> wrote:

On Sun, 29 Aug 2004 00:48:51 +0000 (UTC), kensmith@green.rahul.net
(Ken Smith) wrote:

In article <5z5Yc.3257$6o3.2610@newsread2.news.atl.earthlink.net>,
Ralph Mowery <rmowery28146@earthlink.net> wrote:
[...]
Any class ( A, B , C ) of amp can be plate modulated for AM. It is then
not really an amplifier.

I disagree with this. If the stage puts out more RF than it takes in, it
is an amplifier

By that definition, it could also be an oscillator!

If you are putting RF in and getting RF out at the same frequency, in any
reasonable case the circuit is acting as an amplifier. The circuit may
well oscillate when no input is applied. This sort of amplifier was very
common in the past and still is somewhat common.

A super-regen receiver is the most obvious example. Many tube based FM
receiver designs had a FM detector that would oscillate with no input
signal. The "burst lock oscillator" in a TV is in fact a very narrow pass
filter and amplifier when there is a burst to lock to. With no burst it
oscillates.


--
--
kensmith@rahul.net forging knowledge