Need suggestion of suitable amplifier - RF Cafe Forums

jom

Post subject: Need suggestion of suitable amplifier Posted: Wed Aug 22, 2007 5:14 pm

Captain

Joined: Fri Jan 26, 2007 2:40 am

Posts: 14

Hello

I need to use an amplifier that can handle the following:

Output: 12 dBm

Gain: 20 to 25 dB

Frequency Range: 1 MHz to 30 MHz

Harmonic distortion: > -50 dBc

It's the last spec I'm having trouble with. We've looked at Minicircuit amps which give high compression points (20 dBm or so) but the harmonic distortion is -30 dBc which is not good enough. I also noticed these Mini amps do quite a bit better when you set the DC current higher but that is NOT what they are specified at.

Any alternative ideas?

Thanks

jom

Top

jom

Post subject: Posted: Wed Aug 22, 2007 5:16 pm

Captain

Joined: Fri Jan 26, 2007 2:40 am

Posts: 14

Forgot one more thing: The power supply can be +8 or +15 V.

Thanks

jom

Top

IR

Post subject: Posted: Sat Aug 25, 2007 3:41 pm

Site Admin

Joined: Mon Jun 27, 2005 2:02 pm

Posts: 373

Location: Germany

Hi Jom,

Here is a gain block - SBF-5089 from Sirenza Microdevices:

http://www.sirenza.com/documents/produc ... SBF-5089(Z)_Green_Datasheet.pdf

It will meet all the requirements you listed.

Good luck!!

- IR

Top

jom

Post subject: Posted: Sun Aug 26, 2007 11:19 pm

Captain

Joined: Fri Jan 26, 2007 2:40 am

Posts: 14

Thanks for that!

This amp seems very similar to the Mini-circuits amp I'm using. Do you have any experience with it?

It seems the Mini amp is a bit current starved for some reason, especially at low frequencies. I'm wondering if this might be the case with this Sirenza amp. Neither amp gives any kind of harmonic distortion data. On my Mini-circuits amp I adjusted the bias voltage (thinking it was voltage limited) and noticed a marked improvement in the harmonic distortion. However, as I had raised the voltage the current went up also so I adjusted the bias resistor (to keep the same current) while raising the voltage and the harmonic distortion got bad again..this all with just a 10 dBm output signal (with -15 dBm in). So I'm thinking it's a bit current starved. However, raising the current in the device to above it's rated value isn't a very good idea.

Any ideas?

Thanks again!

jom

Top

IR

Post subject: Posted: Mon Aug 27, 2007 6:00 am

Site Admin

Joined: Mon Jun 27, 2005 2:02 pm

Posts: 373

Location: Germany

I have no experience with this amplifier. Usually manufacturers proivde IP3 and IP2 parameters and for more information you have to contact them.

To see the real non-linear behavior of the active device you need a non-linear model (Like SPICE model).

In general, an amplifier with better linearity will give you better harmonic distortion supression.

Top

Tony Kurlovich

Post subject: Posted: Tue Aug 28, 2007 4:08 pm

Captain

Joined: Thu Oct 19, 2006 6:02 pm

Posts: 7

Given a flat amplifier response the H3 will be lower than the IM3 by 9.54db (20log3). This is because the coefficient for the third order component is common to both.

Given a flat or decreasing gain with frequency, IM3-9.54db is a reasonable upper bound for H3.

When doing a 2 tone test there will be 4 spurs at around 3 times the frequency. The larger ones are not 3A or 3B. The are 2A+B AND 2B+A. The 3A and 3B components are the ones that are down by 9.5db.

Top

jom

Post subject: Posted: Fri Sep 07, 2007 5:12 pm

Captain

Joined: Fri Jan 26, 2007 2:40 am

Posts: 14

After looking around a bit this seems to be a difficult problem according to several vendors such as RFMD etc. Let me restate the problem. I need a +10 dBm to +12 dBm output with Harmonic Distortion less than -50dBc. The supply voltage can be a bit higher than what I said before according to our customer 15V maybe. Also:

Gain: 20 to 25 dB

Frequency Range: 1 MHz to 30 MHz

After talking with apps engineer at several places it seems this kind of low HD spec is a long-sought after. Anyone know the basics of what is REALISTIC when it comes to harmonic distortion in this range?

Thanks

jom

Top

IR

Post subject: Posted: Sat Sep 08, 2007 5:11 pm

Site Admin

Joined: Mon Jun 27, 2005 2:02 pm

Posts: 373

Location: Germany

Hi jom,

My suggestion is that if it hard for you to find such an amplifier, and if it is possible from your system point of view, then design a notch filter for the 3rd harmonics. Should be a notch made of single L and C which will short the 3rd harmonic frequency to GND.

Thanks to Tony Kurlovich for his post.

Top

jom

Post subject: Posted: Mon Sep 10, 2007 1:44 am

Captain

Joined: Fri Jan 26, 2007 2:40 am

Posts: 14

Thanks for that..

However, it seems I wasn't clear. The problem here is that this is a wide bandwidth problem. Meaning the output can be anywhere from 1.5 MHz to 30 MHz so a filter really can't be used, at least nothing below 30 MHz anyways.

Also, another thing is that it's not only the 3rd HD...in fact the 2nd HD seems to be more of a problem. A "balanced" approach could solve this but so far I've had trouble getting a good 180 degree shift for that. And, of course, this does nothing for the 3rd HD.

jom

Top

Peter Raynald

Post subject: Posted: Wed Oct 03, 2007 12:35 pm

Captain

Joined: Tue Sep 07, 2004 3:09 pm

Posts: 11

Would feedback be an option?

I think this is a linearisation technique that could be quite suitable for your frequency range.

Posted  11/12/2012