Can the Gain-Bandwidth Product be increased by combining multiple Op Amps?

Thread Starter

Matjelo Naleli

Joined Oct 7, 2015
42
I am trying to build a high Q resonator at 15MHz using op amps. The trouble that I run into is that when the Gain-Bandwidth Product of the op amps is low the resonance shift dramatically and the Q-factor drops. In the attached picture I have shown a circuit diaram, the intended transfer function and the Bode diagrams for a case where GBP is 50MHz and where it's 20GHz. At GBP of 20GHz the frequency response is very close to what I expected/intended it to do (which is resonance at 15MHz).

So what I am really seeking help with is how one can improve GBP perhaps by combining several op amps (if it's something possible) or some other way. I would also appreciate any suggestions on some op amps with high GBP. By high in this case I mean hundreds MHz to a few GHz or higher.

Thanks
 

Attachments

shteii01

Joined Feb 19, 2010
4,644
You can use multiple stages, each stage providing small amplification. This way you keep a lot more of the bandwidth in each stage because each stage has only small gain. The downside is that you use a lot more parts and there are more inductances and capacitances due to all the wires and components.
 

WBahn

Joined Mar 31, 2012
29,976
So back of the envelope how many stages would be required for say 25 dB gain and a bandwidth of 15 MHz.?
That's gonna depend on the opamp selected. 25 dB is a signal gain of about 18, so at 15 MHz that's a GBP of 266 MHz.

Back of the envelope, if you want to do it with five stages then each stage needs a gain of 5 dB, or about 1.8. So opamps with a GBP of about 30 MHz would put you in the ballpark with very little margin. I don't have a good feel for how much margin you want in practice.

The TI THS3001 has a GBP of 420 MHz, so if everything else falls in line a single amp might be sufficient. I have very little experience (and none of it within the last couple decades) with current-feedback opamps, so I don't know what kind of gremlins are lurking in an application of this type.

Since this is well outside my normal playground, I'm probably missing something major.
 

GopherT

Joined Nov 23, 2012
8,009
That's gonna depend on the opamp selected. 25 dB is a signal gain of about 18, so at 15 MHz that's a GBP of 266 MHz.

Back of the envelope, if you want to do it with five stages then each stage needs a gain of 5 dB, or about 1.8. So opamps with a GBP of about 30 MHz would put you in the ballpark with very little margin. I don't have a good feel for how much margin you want in practice.

The TI THS3001 has a GBP of 420 MHz, so if everything else falls in line a single amp might be sufficient. I have very little experience (and none of it within the last couple decades) with current-feedback opamps, so I don't know what kind of gremlins are lurking in an application of this type.

Since this is well outside my normal playground, I'm probably missing something major.

If you are trying to help the OP, you are way off from his request of 20GHz GBP. If you're carrying on a side conversation to win a pissing contests, knock it off or start your own thread.
 

Marley

Joined Apr 4, 2016
502
Op-amp, high-Q filter, 15Mhz - can't be done! Think discrete inductors and capacitors together with simple gain stages.

Or even crystal or piezoelectric filters.
 

WBahn

Joined Mar 31, 2012
29,976
If you are trying to help the OP, you are way off from his request of 20GHz GBP. If you're carrying on a side conversation to win a pissing contests, knock it off or start your own thread.
More trolling, huh?

I just answered a question (one that you apparently didn't have a problem with).

I guess, according to you, providing a choice of op-amp that has a GBP of 420 MHz is "way off" his request of, "I would also appreciate any suggestions on some op amps with high GBP. By high in this case I mean hundreds MHz to a few GHz or higher."
 

GopherT

Joined Nov 23, 2012
8,009
More trolling, huh?

I just answered a question (one that you apparently didn't have a problem with).

I guess, according to you, providing a choice of op-amp that has a GBP of 420 MHz is "way off" his request of, "I would also appreciate any suggestions on some op amps with high GBP. By high in this case I mean hundreds MHz to a few GHz or higher."
No, I think #12 was heading in the right direction. We need the OP back to see what suits him/her. Until then, keep proposing as many sub-standard solutions as you want and creating more side-bar conversations - it doesn't muddy the conversation at all.
 

Papabravo

Joined Feb 24, 2006
21,157
If you are trying to help the OP, you are way off from his request of 20GHz GBP. If you're carrying on a side conversation to win a pissing contests, knock it off or start your own thread.
I don't quite see it that way. A statement was made that you could achieve a result if you strung enough lower capability devices together. I wanted to put a number on it to see if it was practical. Your mileage may vary. We all want to help the TS/OP, but the best way to help him with his problem remains unclear. I do think he needs to weigh in. It is this kind of thinking out loud that leads to amazing results. Point is nobody has to be perfect 100% of the time and have all the answers. Going down blind alleys can be a useful exercise, so lighten up will 'ya.
 
Top