0
\$\begingroup\$

In the process of designing a broadband RF amplifier, I've come across some commercial boards, claiming gain flatnesses of +-1db from DC to 1Ghz. Those boards are ridiculously simple, having only one or two transistors and, in some of them, almost no matching on either side.

See for example this one:

https://www.dhgate.com/product/1mhz-to-2000mhz-64db-gain-nf-1-8-rf-low-noise/425922302.html#ENhp-en

It doesn't have input bias, no matching, no nothing, beside DC power and chokes.

Despite the simplicity, the specs claim 64db of absolute flatness from DC to 1Gz, and they even show a graph of S21 showing it.

How is it possible to archive such impressive performances with those simple circuits?

Are the active devices internally matched and biased?

Or are all those boards nothing but chinese fakery.

Thank you!

\$\endgroup\$
  • \$\begingroup\$ The active devices aren't simple transistors. They are complex chips with a lot of stuff squeezed into a 3 pin housing that looks like a transistor. That said, I wouldn't trust anything bought from those folks. The page says it's a voltage regulator drive IC, and says it's used in a laptop - and that it has an LCD. So, buyer beware. \$\endgroup\$ – JRE Nov 11 '18 at 21:56
  • \$\begingroup\$ Looking closely at the S21 plots, it's pretty obvious these don't have anywhere close to 1dB gain flatness. \$\endgroup\$ – isdi Nov 11 '18 at 22:31
  • \$\begingroup\$ if you want to compare, a mini-circuits CMA-62+ has a gain flatness spec of 1.4dB over 50MHz to 4GHz, but flatness is only one of many trade-offs in RF design. \$\endgroup\$ – isdi Nov 11 '18 at 22:37
  • \$\begingroup\$ Thanks @JRE, that makes sense now. Some of the active devices seemed to me like a BJT in a macro package, while other boards mount SOT89 devices. A search in Digikey yields lots of RFICs in transistor-like packages, which I had never seen before. \$\endgroup\$ – Angel Nov 11 '18 at 22:58
  • \$\begingroup\$ @isdi you're right, It's non even close to the Minicircuit's specs, but I thought the devices may be simple FETs or BJTs and, being so, the claimed performance would be impossible to archieve with such minimalistic arrangements. Thanks! \$\endgroup\$ – Angel Nov 11 '18 at 23:04
0
\$\begingroup\$

They don't even mention other specs such as P1dB and OIP3.

Moreover, look at the S11 and S22. If I am correct, they're close to 0dB. (The reference line is the first line and the vertical resolution is 5dB/div and 2dB/div respectively).

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.