Author Topic: GDT for RF Amp  (Read 1837 times)

0 Members and 1 Guest are viewing this topic.

Offline stcosoTopic starter

  • Regular Contributor
  • *
  • Posts: 107
  • Country: it
GDT for RF Amp
« on: May 09, 2017, 01:31:04 pm »
Hi guys.. I'm building a little RF Amp and i'm looking for some information To optimize one of the gdt that i'm gonna use.

So basically i've got an oscillator (about 14MHz), then a bunch of transistor and finally a totem pole Made with bd139/140. I'm trying to find some data about the output capacitance of these two transistor but with no luck.

Thank you in advance.
 

Offline BigBoss

  • Regular Contributor
  • *
  • Posts: 115
  • Country: fr
Re: GDT for RF Amp
« Reply #1 on: May 10, 2017, 12:47:59 am »
BD139/140 pairs are for Audio Frequencies, they won't work at 14MHz.
 

Offline mmagin

  • Frequent Contributor
  • **
  • Posts: 610
  • Country: us
Re: GDT for RF Amp
« Reply #2 on: May 10, 2017, 01:08:29 am »
Not sure thats entirely true. At least one datasheet specifies a Ft of 190 MHz: https://www.centralsemi.com/get_document.php?cmp=1&mergetype=pd&mergepath=pd&pdf_id=BD135_SERIES.PDF

That said, a complementary pair push-pull output is pretty uncommon for RF.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16607
  • Country: us
  • DavidH
Re: GDT for RF Amp
« Reply #3 on: May 10, 2017, 02:36:24 am »
I think of totem-pole output stages as using the same polarity of transistor with one common emitter and one common collector like with TTL.  An NPN BD139 and PNP BD140 would be used to make a complementary push-pull (1) output stage.

Function generators in that frequency range usually have complementary output stages with series termination and usually shunt feedback which is *not* want you want in an RF output.  It will work but the series termination halves the output power and efficiency.  Real RF amplifiers have a deliberately mismatched output impedance which lowers power loss in the output transistor raising efficiency.

If you really want a complementary push-pull output, then there is a design where the output resistor is a fraction of the output impedance.  Positive feedback from the output raises the effective output impedance so the output is matched but losses are still low.  It is commonly used with operational amplifiers to implement series termination when the output power needs to be maximized for a given output swing.  This shows the general idea and I think it could be made to work at 14MHz:

https://www.maximintegrated.com/en/app-notes/index.mvp/id/3390

RF amplifiers do not need that kind of complication and it would not work at higher frequencies.

(1) Be careful about old documentation.  In the past push-pull sometimes meant differential just like now differential sometimes means difference.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf