Body
Hi Mick,
" They still have to connect to the output device through traces of a certain length, so how do they choose the trace length besides as short as possible?"
It's the impedance of the traces that matter. The traces must be made as transmission lines of same impedance. As long as everything is of same impedance the signal continues moving forward. When the impedance changes no matter how far away that's where the signal splits up moving in different directions. Changing the width of a trace will change its impedance and thus distort the signal. Even a 90 degree turn will reflect some of the signal and also make some of it radiate straight ahead.
The following article shows some of the things to consider when designing microstrip transmission lines:
'Evaluating Microstrip with Time Domain Reflectometry'
ftp://140.126.5.158/Public/Basic%20Measurement%20Equipments%20AVI/Educator%20Corner%206.0%20CD/Data/pdf/5968-0007E.pdf ftp://140.126.5.158/Public/Basic%20Measurement%20Equipments%20AVI/Educator%20Corner%206.0%20CD/Data/pdf/5968-0007E.pdf
I have an article somewhere by someone dissecting the coaxial cable of a high speed oscilloscope probe. Such cables for high frequencies use a resistive inner conductor for damping reflections.
PCB layout: http://learnemc.com/pcb-layout http://learnemc.com/pcb-layout
More from this site: http://learnemc.com/emc-tutorials http://learnemc.com/emc-tutorials
Regards
Ole
---In [email protected], <mkjekyll@...> wrote :
Ole,
Yes I understand this part about building according to these RF rules, just wondering if the theory is correct or if I am missing something from my topology.
Thanks for the links, I already bought one of these but before I found it I was contemplating what I asked you. I have no idea how they build an amp that is so wideband, how they come up with the trace lengths to satisfy VHF to Microwave other than most of it resides on a chip with a much shorter wavelength than the chip process dimensions. They still have to connect to the output device through traces of a certain length, so how do they choose the trace length besides as short as possible?
I am sure the capacitance of the PCB is also a factor that has to be taken into account. Can't just put a huge ground plane on one side like we do in audio.
On 3/26/2018 3:53 AM, onielsen@... mailto:onielsen@... [EVGRAY] wrote:
Hi Mick,
At those frequencies the energy is very radiant. Everything has to be made according to RF rules and treated as transmission lines. Else the signal doesn't follow the signal path as planned. Any change in impedance reflects or splits the signal. Read Ivor Catt on this even though he explains it using digital signals. If attaching a current buffer everything has to match. I haven't worked on such high frequencies. If trying this at least know the theory behind it or have the measuring instruments for this frequency range.
This site (http://nov79.com/amr/cont.htm http://nov79.com/amr/cont.htm) shows how to make a simple current amplifier by using matched bipolar transistors. The upper frequency is only limited by the upper frequency of the transistors. This can be 40MHz for an audio amplifier. This may cause oscillations at the max frequency if not done the correct way.
What about a small 3W broadband power amplifier module:
https://www.ebay.com/itm/1-1000MHz-3W-RF-Broadband-Power-Amplifier-Module-Radio-Transmission-FM-VHF-/183031453375 https://www.ebay.com/itm/1-1000MHz-3W-RF-Broadband-Power-Amplifier-Module-Radio-Transmission-FM-VHF-/183031453375
It only goes to 1GHz.
Google search for 'broadband power amplifier module':
https://www.google.dk/search?source=hp&ei=c8G4Wuv0LIXbwALZxJTQDw&q=broadband+power+amplifier+module&oq=broadband+power+ampli&gs_l=psy-ab.3.2.0l2j0i22i30k1l8.1878.9748.0.22828.21.20.0.1.1.0.104.1822.18j2.20.0....0...1c.1.64.psy-ab..0.21.1820...46j0i46k1j0i10k1.0.jIuIG0_77XA https://www.google.dk/search?source=hp&ei=c8G4Wuv0LIXbwALZxJTQDw&q=broadband+power+amplifier+module&oq=broadband+power+ampli&gs_l=psy-ab.3.2.0l2j0i22i30k1l8.1878.9748.0.22828.21.20.0.1.1.0.104.1822.18j2.20.0....0...1c.1.64.psy-ab..0.21.1820...46j0i46k1j0i10k1.0.jIuIG0_77XA
Or search 'rf power amplifier module'
Regards
Ole
---In [email protected] mailto:[email protected], <mkjekyll@...> mailto:mkjekyll@... wrote :
Ole,
What would be the negative issues trying to attach a FET current buffer
to an LNA in order to make a few watt wideband RF amplifier?
As I am an audio guy I am paralleling this to an audio amp seems simple
a voltage gain stage and then a current gain or buffer stage.
There are plenty of cheap LNA's that are very wideband like this
https://www.ebay.com/itm/New-1MHz-2GHz-64dB-Gain-LNA-RF-Broadband-Low-Noise-Amplifier-Module-HF-VHF-UHF/232103825103?hash=item360a774ecf:g:iCQAAOSwgu9ZZEBg https://www.ebay.com/itm/New-1MHz-2GHz-64dB-Gain-LNA-RF-Broadband-Low-Noise-Amplifier-Module-HF-VHF-UHF/232103825103?hash=item360a774ecf:g:iCQAAOSwgu9ZZEBg