HF design approach

L

lerameur

Guest
Hello to all,
I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.
I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.
At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.
Thanks

Ken
 
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken
Click to expand...

Aloso, should ibe using high or low Rds mosfet?

Ken
 
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken
Click to expand...

I heard vacuum tube where expensive and did not perform better and also are very expensive

Ken
 
On Thursday, February 6, 2014 8:50:46 PM UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken
Click to expand...

A few watts so ~100mA of current? If money is no object you could look at the Apex opamps.
http://www.apexanalog.com/

If you want to make your own, then an opamp driving a push-pull would be OK.. I don't know about using Fet's in the push pull. That sounds like more cross over distortion. I might try some bjt's... (easier for me to bias.) Will a bit of cross over distortion be bad?

Re: Rds of mosfet. Since your load is someting like 400 ohms (what is the load?) I don't think the Rds of the mosfet will matter that much.

George H.
 
"lerameur"

Hello to all,
I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V)
a couple of watts.
I am a newbie to HF frequencies , but I would like to know what is the best
amplifier technique for these frequencies.
At first I wanted a pushed and pull mosfet driven by a mosfet driver, but
the more I am looking the more i'm confused. I've seen some expensive
specialized mosfet at over 400$.... other circuits with inductors and RF
type transformers. which path is the way to go in my case. I am a bit lost
in this sea of different circuits.


** This is the way it is really done:

http://www.thermionic.org/tx807s.jpg





.... Phil
 
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken
Click to expand...

Hopefully no cross over distortion. I want to replicate Richard Vialle experiment about resonating a coil over a tube.
http://jnaudin.free.fr/rvproject/html/ugentest27.htm#TESTV132

... keep in mind..The question here is not if this project (richard vialle) falls into sanity or not... but to have a 40V amplifer .

Regards,
K
 
On Fri, 7 Feb 2014, lerameur wrote:

On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken

I heard vacuum tube where expensive and did not perform better and also
are very expensive
Click to expand...

Not much was ever done with broadband tube amplifiers. They existed, but
not out in public view much.

Transistors, with their lower impedance, are easier to make broadband.

That said, you've not revealed what this is for. The application may
reflect on what you need, rather than just assuming "an amplifier that
goes from here to there".

Michael
 
On Friday, February 7, 2014 2:11:12 PM UTC-5, lerameur wrote:
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,
I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.
I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks
Ken
Hopefully no cross over distortion. I want to replicate Richard Vialle experiment about resonating a coil over a tube.

http://jnaudin.free.fr/rvproject/html/ugentest27.htm#TESTV132

.. keep in mind..The question here is not if this project (richard vialle) falls into sanity or not... but to have a 40V amplifer .

Regards,
K
Click to expand...

Well there's a circuit there, start by copying it.
(I don't know much about fets at 5MHz.)
I might want to put a choke in the power supply line.
I didn't translate the french. Does he bias the fet on with the sig. gen.?
I was going to say if you don't want cross over distortion, and can handle the power waste then class A is an easy option.
That circuit only needs +40V.. I was thinking +/- 40.
In which case its easier to wrap an opamp around the pass element.


When I think of these circuits (opamp->class A) I always imagine feeding back the current, by just hanging a sense resistor to ground. (or high side if I must.)
But what if I wanted to feedback the load voltage? (What do I sense and where's the pass element?)

George H.
 
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken
Click to expand...

Hi,

This is a lot of detail. Actually, I am only doing an experiment, so I do not want to spend a lot of money. My load is about 3uH and 550pF. Actually I want an amplifier with no crossover distortion. So my question is, what is the best economical way to do an amplifier with no crossover distortion.?

Ken
 
On Thu, 06 Feb 2014 17:50:46 -0800, lerameur wrote:

Hello to all,
I am designing an amplifier for frequencies between 500khz to 5Mhz.
(+/-40V) a couple of watts.
I am a newbie to HF frequencies , but I would like to know what is the
best amplifier technique for these frequencies.
At first I wanted a pushed and pull mosfet driven by a mosfet driver,
but the more I am looking the more i'm confused. I've seen some
expensive specialized mosfet at over 400$.... other circuits with
inductors and RF type transformers. which path is the way to go in my
case. I am a bit lost in this sea of different circuits.
Thanks

Ken
Click to expand...

There's not enough information in your question.

Your stated voltage and power implies load equivalent parallel
resistances in the neighborhood of 1k-ohm.

How reactive are your loads? Is there any significant inductance?
Capacitance higher than about 30pF? If the answer is "yes" then that
complicates your task.

How much can you expect your load to change? Do you need to drive a
variety of loads, or just one? If just one, do the load characteristics
vary much from unit to unit, or over temperature or humidity or pressure
or as a function of age or in some other manner? Varying loads make the
design task harder.

How cost sensitive is your application? Are you building ten systems for
medical or military? Or are you building a product that needs to ship
100,000 a year and sell for less than the price of a hamburger? Few
systems points you towards a $$$ off the shelf solution. Lots of systems
points you toward $$$$ spent in custom circuit design with low parts cost.

How broadband is your signal? Are you going to be using 500kHz to 5MHz
at the same time, or is this a radio-like application where the signal is
always some narrow slice of the operating bandwidth at any given time?
Broadband is harder. Narrowband is easier.

How efficient do you need to be? Can you afford to burn up 20 watts to
get "a few" to the load? Or is every wasted erg a tragedy? Efficiency
will cost you here, in terms of parts cost and engineering time as well.

Every single one of the answers that you can give affects what circuit is
"best". As an example, I used to build low-end military stuff (which
overlapped with high-end civilian law enforcement and search-and-
rescue). We usually expected to build 100 units a year of any given
product, with a lifetime for any one board of around five years. So
there wasn't much opportunity for payback if you flogged the design and
took $50 out of the parts cost. For us, when faced with "+/- 40V, 5MHz
top end, a few watts", the first place we looked was the Apex
Semiconductor site. Apex built hybrid op-amps that were higher voltage
or higher power than you could get in a monolithic. They built really
good products that nearly always worked as promised, and they charged up
the wazoo.

If your application fits the "small volume expensive" mold, then find out
who makes those parts now (Apex got bought by someone and I haven't kept
track of who). Then look through the catalog and chances are there'll be
an amplifier of sufficient voltage capacity and gain-bandwidth product to
satisfy your needs.

If your application is for 100,000 a month, then talk to Jim Thompson
about building you a custom IC that you can pair with a couple of $0.10
off-board MOSFETs. He'll charge you an arm an a leg for the work, and
the fab he points you to will charge that much, too, but when you
amortize the cost over your production volume it'll be fingernail
pairings.

If your application is somewhere in the middle -- well, you need
something in the middle, ranging from an all-transistor custom design, to
a really fast op-amp driving a couple of outboard transistors as a final
amp, to just about anything.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
 
On Sat, 8 Feb 2014, lerameur wrote:

On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken

Hi,

This is a lot of detail. Actually, I am only doing an experiment, so I
do not want to spend a lot of money. My load is about 3uH and 550pF.
Actually I want an amplifier with no crossover distortion. So my
question is, what is the best economical way to do an amplifier with no
crossover distortion.?

What are you amplifying? Do you actually need a broadband amplifier?
Click to expand...

Class C has no crossover distortion, but it's not a linear amplifier, and
counts on the output network to clean things up. If the signal is
modulated, then a class c amplifier is not the way to go.


Michael
 
On Sat, 08 Feb 2014 12:07:09 -0800, lerameur wrote:

On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz.
(+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the
best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver,
but the more I am looking the more i'm confused. I've seen some
expensive specialized mosfet at over 400$.... other circuits with
inductors and RF type transformers. which path is the way to go in my
case. I am a bit lost in this sea of different circuits.

Thanks



Ken

Hi,

This is a lot of detail. Actually, I am only doing an experiment, so I
do not want to spend a lot of money. My load is about 3uH and 550pF.
Actually I want an amplifier with no crossover distortion. So my
question is, what is the best economical way to do an amplifier with no
crossover distortion.?
Click to expand...

Whose post are you replying to?

Are the inductance and capacitance in parallel or series? Is there any
resistance involved?

550pF has an impedance of around 60 ohms at 5MHz. 3uH has an impedance
of around 9 ohms at 500kHz. If they're in parallel and you're driving
them with a sine wave, your problem isn't "a few watts", your problem is
driving four amps peak into a reactive load (at 500kHz), and staying
stable while looking into a honkin' big capacitor (at 5MHz).

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
 
Michael Black wrote:
On Fri, 7 Feb 2014, lerameur wrote:

On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken

I heard vacuum tube where expensive and did not perform better and also
are very expensive


Not much was ever done with broadband tube amplifiers. They existed, but
not out in public view much.
Click to expand...

Distributed amplifier. I've seen 16 6146 tubes in one for a DC to 5
MHz video modulator, plus another 6146 for the input amp.


Transistors, with their lower impedance, are easier to make broadband.

That said, you've not revealed what this is for. The application may
reflect on what you need, rather than just assuming "an amplifier that
goes from here to there".

Michael
Click to expand...

--
Anyone wanting to run for any political office in the US should have to
have a DD214, and a honorable discharge.
 
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:
Hello to all,

I am designing an amplifier for frequencies between 500khz to 5Mhz. (+/-40V) a couple of watts.

I am a newbie to HF frequencies , but I would like to know what is the best amplifier technique for these frequencies.

At first I wanted a pushed and pull mosfet driven by a mosfet driver, but the more I am looking the more i'm confused. I've seen some expensive specialized mosfet at over 400$.... other circuits with inductors and RF type transformers. which path is the way to go in my case. I am a bit lost in this sea of different circuits.

Thanks



Ken
Click to expand...

Hello,
I mentioned the goal of this in an earlier post, if you would like to read on it here is the link :http://richard-vialle.info/ or http://richard-vialle.info/~richardv/index.php/articles/systemes-exprmt/14-reproduction-autogenerateur-second-gen-jln

I would like no cross over since the circuit will be going into a resonant mode

ken
 
On Sat, 8 Feb 2014 12:07:09 -0800 (PST), lerameur
<lerameur101@gmail.com> wrote:

This is a lot of detail. Actually, I am only doing an experiment, so I do n=
ot want to spend a lot of money. My load is about 3uH and 550pF. Actually I=
want an amplifier with no crossover distortion. So my question is, what is=
the best economical way to do an amplifier with no crossover distortion.?

Ken
Click to expand...

Just out of curiosity, why the focus on "no crossover
distortion" for an RF amp? Crossover distortion is
typically a more-or-less fixed amount, which becomes a
significant percentage only as the overall signal level
approaches zero. In audio work, it's important because you
can have very quiet passages; if you only listened to loud
music you would never detect it.

So I'm curious about what sort of RF experiment would need
low crossover distortion. If it's AM, I guess it could be
significant at the modulation peaks (over 99%, say), when
the carrier is driven near zero. But you probably don't
want to run that close to zero anyway, since even with a
perfect amp if you break through zero you'll distort the
demodulated signal.

I suspect that crossover distortion in the carrier (being at
harmonics of the carrier frequency) would never be
detectable in the demodulated signal.

So, what's the experiment (if it's not a proprietary
secret)?

Best regards,



Bob Masta

DAQARTA v7.50
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Sound Level Meter
Frequency Counter, Pitch Track, Pitch-to-MIDI
FREE Signal Generator, DaqMusiq generator
Science with your sound card!
 
On Saturday, February 8, 2014 7:12:13 PM UTC-5, Tim Wescott wrote:
On Sat, 08 Feb 2014 12:07:09 -0800, lerameur wrote:
Whose post are you replying to?

Are the inductance and capacitance in parallel or series? Is there any
resistance involved?
Click to expand...

It's a parallel resonance, (at least if he's copying the circuit on the top of the web page he posted.)
http://jnaudin.free.fr/rvproject/html/ugentest27.htm#TESTV132

George H.

550pF has an impedance of around 60 ohms at 5MHz. 3uH has an impedance
of around 9 ohms at 500kHz. If they're in parallel and you're driving
them with a sine wave, your problem isn't "a few watts", your problem is
driving four amps peak into a reactive load (at 500kHz), and staying
stable while looking into a honkin' big capacitor (at 5MHz).

--

Tim Wescott

Control system and signal processing consulting

www.wescottdesign.com
Click to expand...
 
On Sunday, February 9, 2014 1:48:43 PM UTC-5, lerameur wrote:
On Thursday, 6 February 2014 20:50:46 UTC-5, lerameur wrote:

I would like no cross over since the circuit will be going into a resonant mode

Hmm well if it's an LC resonace (with a Q >10 or so) then you don't really care that much about distortion... The resonance sorta kills everything but the fundamental frequency.
Click to expand...

George H. (try sending a square wave into an resonance.)
ken
Click to expand...
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top