RF (7GHz) Transistor 2X Amplifier Design (or up to 10X preferably) Advice Needed

[Mechatrommer]

i have 2 intentions...

1) amplify RF signal at 2X gain or so from KC901V 7GHz VNA for S21 respond plot. i believe PORT2 from where the signal out is 50 ohm impedance. the PORT1 to where the signal in is also 50 ohm impedance.
2) generic RF LNA amplifier to RF transceiver such as RTC6705 chip or similar.

so i'm thinking of buying rf transistor such as...
https://www.ebay.com.my/itm/6pcs-ATF-35376-TRANSISTOR-Ga-As-PHEMT-FET-18GHz-HP/141587446135?hash=item20f744f977:g:HasAAOSwBLlU70XF
https://www.ebay.com.my/itm/50pcs-BFP420-TRANSISTOR-RF-25GHz-160mW-4-5V-SOT343-NPN-INFINEON/141587458325?hash=item20f7452915:g:80wAAOSweW5U70qu

instead of ready made module because more expensive. since this is RF in GHz range, i suspect i cannot simply follow the circuit design transistor amplifier for lower frequency like audio. i found guide on how to do it but its a bit overwhelming and math intensive for it simplest task...

so i need advice on simple schematic like such as how this module is made...
https://www.ebay.com.my/itm/10K-2GHZ-LNA-broadband-RF-amplifier-gain-26dB-1-5G-FM-HF-VHF-UHF-Ham-Radio/202075029166?hash=item2f0c9c3eae:g:fkcAAOSwiKlZ2Dl-
this one looks simpler and easier to build. any schematic for it? and any advice on what properties to look for when choosing rf transistor. i'm hoping simpler 2X gain transistor amplifier from few KHz or MHz up to 7GHz range, that can accept 50 ohm signal source and output 50 ohm impedance signal. thanks.

[David Hess]

What you are probably looking for and what your last Ebay link shows is a MMIC (monolithic microwave integrated circuit) amplifier.  They are inexpensive and do exactly what you want.  You have an add an input and output coupling capacitor and an output bias network.

https://www.minicircuits.com/app/AN60-010.pdf

[TheUnnamedNewbie]

What do you mean with 2X gain? Two times? So 3 dB?
Do you need it to be wideband, or is narrowband ok?

It is not easy to just quickly make a amplifier at 7 GHz. You will likely want at least some basic simulations to ensure stability. I've tried to build amplifiers at 2.4 GHz that ended up being shitty oscillators instead, because at the time I thought I didn't need simulations.

I suggest buying a off-the-shelf module.

[Mechatrommer]

yes 3dB, times 2. my head doesnt work well yet in dB... if few KHz to 7GHz range is considered wideband, then yes, preferably wideband. i may not know what i'm talking about. but i can see some smd inductors and resistors in the linked ebay module above, i need to know how is that done... another reason to build my own is for learning so in the future i may design my own circuit. thank david for the link i will try to digest the pdf. the linked ebay module pcb seems to simple that i think not necessating simulation, just a ground plane, and shortest possible tracks, or possibly impedance controlled, no? as i said i may not know what i'm talking about. david, yes i may need the mmic, but since the gain i want is 2X (3dB) so  may need something like > 14GHz GBW MMIC, correct?

[ThomasDK]

i may not know what i'm talking about. but i can see some smd inductors and resistors in the linked ebay module above, i need to know how is that done...

http://www.rfwireless-world.com/Terminology/mmic.html

[David Hess]

I always think of times factors as voltages and dB as powers making 2X be 6dB.

[TheUnnamedNewbie]

I always think of times factors as voltages and dB as powers making 2X be 6dB.

When dealing with RF doesn't one usually talk about power instead of voltage or current? Regardless, both are "right" given it was not specified. I guess this is a good example showing why understanding specifications is fundamental to a design.


OP, don't take this the wrong way, but I get the impression that this project may be over your head.

yes 3dB, times 2. my head doesnt work well yet in dB... if few KHz to 7GHz range is considered wideband, then yes, preferably wideband.

That is quite wideband. You have the additional advantage that you can calibrate out the phase and amplitude response of the amplifier, but still, making a amplifier that runs from DC (a few KHz is DC when compared to 7 GHz) to 7 GHz is not easy.

Since the gain i want is 2X (3dB) so  may need something like > 14GHz GBW MMIC, correct?

The concept of GBW-system bandwidth tradeoff only applies to amplifiers where we use feedback. I don't know of a lot of amplifiers you can buy off the shelf where you use feedback at these frequencies.

[David Hess]

I always think of times factors as voltages and dB as powers making 2X be 6dB.

When dealing with RF doesn't one usually talk about power instead of voltage or current? Regardless, both are "right" given it was not specified. I guess this is a good example showing why understanding specifications is fundamental to a design.

I am well aware of the difference and brought it up because the specification is ambiguous.  How many RF applications require just 3dB or 6dB of gain?  Gain before power dividers some to mind.

I actually have a whole set of RF attenuators intended for calibration which are marked x2, x2.5, x5, x10, etc. based on voltages instead of powers.  What I have never seen is a "x2 attenuator" or "x2 amplifier" which is 3dB and not 6dB.  Video amplifiers follow the same rules because they often drive double terminated transmission lines.

[TheUnnamedNewbie]

I always think of times factors as voltages and dB as powers making 2X be 6dB.

When dealing with RF doesn't one usually talk about power instead of voltage or current? Regardless, both are "right" given it was not specified. I guess this is a good example showing why understanding specifications is fundamental to a design.

If it was your intention to be patronizing while plausibly denying it, then you failed.

This was not at all my intention. I was merely trying to illustrate why I said 3 dB. I have never seen people refer to 2X in RF terms (but I am young so it's just me not having seen many things).

[David Hess]

I have never seen people refer to 2X in RF terms (but I am young so it's just me not having seen many things).

I have seen 2X used in time domain applications like pulse generators and oscilloscopes and when describing amplifiers intended to drive double terminated transmission lines like with video and wideband operational amplifiers for instance.  In these cases, it has the same meaning as 6dB power.

[KJDS]

Just pick the most suitable one from here

http://www.analog.com/en/products/rf-microwave/rf-amplifiers.html

Designing a transistor based LF to 7GHz amp is a project that's about a similar scale to a final year university project, if you don't mind putting in a hundred or so hours of effort in, and have access to a decent RF circuit simulator you may get something working, otherwise you'll either have horrible performance or an oscillator.

[David Hess]

Designing a transistor based LF to 7GHz amp is a project that's about a similar scale to a final year university project, if you don't mind putting in a hundred or so hours of effort in, and have access to a decent RF circuit simulator you may get something working, otherwise you'll either have horrible performance or an oscillator.

But designing and fabricating a MMIC based LF to 7GHz amplifier is much more practical.

Maybe Mechatrommer was not aware of MMIC based solutions.

[Mechatrommer]

i'm aware of it except the high frequency effect to the pcb and circuit that i'm worry about. unintentionally building an oscillator is one of it mentioned here. let alone try to ensure the IO is 50 ohm impedance throughout the bandwidth. but reading some mmic's datasheet it seems like readily providing 50 ohm impedance input and output such as HMC311LP3. but i'm not sure how thats possible since the MMIC requiring dc block capacitor at both IO. anyway (finger crossed) i'm still learning and looking for some MMIC option and pcb kit from ebay, comparing datasheets and all. seller rfextra sells several option rf related, so i'm thinking to buy hmc311 or hmc480 or MGF4919G and some pcb kit from him. he also got some rf attenuators that interest me, i hope he can provide cheaper shipping cost discount. i hope i will not spend hundreds of hours with this option trying to build what i want to do..

[David Hess]

Maintaining a 50 ohm environment is relatively easy and using a lossy printed circuit board material like FR4 actually helps prevent unintended oscillation as long as input and output lines are kept separated and proper decoupling with a ground plane is present.  I am sure if you look you can find a layout and parts list for using MMICs.  Since you do not need all of the available gain, resistive pads added to the input and output will improve stability.

With a single transistor amplifier, you have to worry about matching for broadband operation which is much more complicated.