MMIC (op-amp) vs transistor amplifier noise

[Synaps3]

What type of amplifier typically has more noise, a MMIC or a transistor amplifier? Most of the MMICs seem to have a noise figure around 3dB. What is the noise figure range for a transistor amplifier?

In other words, if I'm looking to design an amplifier with the lowest noise possible, should I use an MMIC or transistor type amplifier?

 

[betwixt]

MMICs are good as basic gain blocks but not optimized for any particular purpose. You will get lower noise figure from a properly designed transistor (BJT/FET or HEMT) where you can set the bias and control the impedance matching to suit your particular application. Do not rule out MMICs as low noise amplifiers though, their performance is impressive and if you need more than one stage of amplification to meet the gain you require, the cumulative noise of several transistor stages may come close to that of a MMIC.

Brian.

 

[BigBoss]

What type of amplifier typically has more noise, a MMIC or a transistor amplifier? Most of the MMICs seem to have a noise figure around 3dB. What is the noise figure range for a transistor amplifier?

In other words, if I'm looking to design an amplifier with the lowest noise possible, should I use an MMIC or transistor type amplifier?

The answer is short and simple..
A LNA which has been carefully designed and optimized with discrete components ( BJT or FET depending on the frequency) will definitely have lower noise compare to MMIC or RFIC.

 

[volker@muehlhaus]

The answer is short and simple..
A LNA which has been carefully designed and optimized with discrete components ( BJT or FET depending on the frequency) will definitely have lower noise compare to MMIC or RFIC.

Short and simple answers are not always the most accurate answers.

It is true that many MMIC gain blocks, or wideband LNA, have higher noise figure than optimized narrowband LNAs from discrete devices. However, if you look at more specific integrated LNAs, you can find NF=0.4dB devices that can compete with discrete devices.

**broken link removed**
**broken link removed**

 

[vfone]

Yes, the noise figure of an LNA is lower than noise of MMIC.
But MMICs usually are very broadband gain amplifiers, and their noise figure is the same, I mean is flat with frequency.

A single transistor LNA usually is a narrow band amplifier, and its noise figure is the same, narrow band.
For example a narrow band LNA with 0.5dB noise figure at 1GHz, could have 10dB noise figure at 1.5GHz. This is just an example.

 

[BigBoss]

Short and simple answers are not always the most accurate answers.

It is true that many MMIC gain blocks, or wideband LNA, have higher noise figure than optimized narrowband LNAs from discrete devices. However, if you look at more specific integrated LNAs, you can find NF=0.4dB devices that can compete with discrete devices.

**broken link removed**
**broken link removed**

NFmin=0.4dB@150MHz !!

I can obtain a NF=0.1-0.2dB at that frequency with a discrete simple RF transistor.Datasheets can be sometimes very misleaded.

 

[volker@muehlhaus]

I can obtain a NF=0.1-0.2dB at that frequency with a discrete simple RF transistor.

0.1dB at 150Mhz? I'm impressed. Can I get one, please?

About the best that I am aware of are specified with 0.25dB:
https://www.ssb.de/shop1/index.php?page=categorie&cat=263

- - - Updated - - -

Yes, the noise figure of an LNA is lower than noise of MMIC.

If you mean MMIC = wideband gain block, I agree.
If we read MMIC as Microwave Monolithic Integrated Circuit, it could be a single-chip LNA with decent performance.

 

[vfone]

Yes, I mean wide band blocks, as the ones from Minicircuits, which have pretty flat noise figure vs frequency.

BGB series from Infineon have better noise figure, but their NF varies with frequency even is matched for optimum NF.
I think Infineon BGBs use single stage amplifier, when Minicircuits use Darlington topology.