The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> op amp topology
https://designers-guide.org/forum/YaBB.pl?num=1204619621

Message started by aaron_do on Mar 4^th, 2008, 12:33am

Title: op amp topology
Post by aaron_do on Mar 4^th, 2008, 12:33am

Hi all,

i need to design a simple differential op amp which is able to drive a resistance of about 2K with a unity gain bandwidth of at least 100 MHz and as low power as possible. Any suggestions on a topology?

thanks,
Aaron

Title: Re: op amp topology
Post by thechopper on Mar 5^th, 2008, 2:47am

Hi,

One simple approach would be to use a folded cascode amp to maximize GBW, stabilized with a Miller second stage. This one could be an AB output stage to save some power an at the same time to have enough current driving for your 2K load.
This would be at least a starting point.

Hope this helps

Tosei

Title: Re: op amp topology
Post by vivkr on Mar 5^th, 2008, 2:57am

Hi Aaron,

I would use some kind of a class-AB topology. There are several out there, some using 2 stages + an AB, some purely AB. If you do not need
very large gain and noise is not very critical, try out: "Lopez-Martin et al" from JSSC, May 2005. Of course, there are several other possible
solutions in AB, all very good. I just know these ones because I was playing around with them and was able to make these very simply. They also
don't need any overhead circuitry. If you look carefully, you will find ways to reduce the power consumption even further, particularly with
Topology # 1 in this paper where 2 branches are redundant. Just search around for more if you like.

I would keep it simple.

Regards
Vivek

Title: Re: op amp topology
Post by vivkr on Mar 6^th, 2008, 12:17am

Hi Aaron,

I was just trying something out and I think that maybe the opamps in that paper I mentioned aren't the best for you as the gain would be rather average
with them (40-50 dB). They are better suited for driving cap loads. If you need more gain, try using a 2-stage or so topology with a class-AB stage.
There are several good examples of those in the literature and even in textbooks. Push-pull from Gray & Meyer would do well at the cost of some headroom.

Regards
Vivek

Title: Re: op amp topology
Post by fran2k5 on Mar 7^th, 2008, 9:45am

Hi,

if you have enough supply you could try to boost the gain of the first stage with a telescopic configuration and use a second stage like in Miller opamps. If the gain is enough you could let the resulting OTA drive the 2K resistance w/o output stage.

Hope this helps.

Regards,

frank

Title: Re: op amp topology
Post by aaron_do on Mar 8^th, 2008, 4:10am

Thanks for all the replies.

There are a few problems. I actually need a fully differential op amp. And yeah it needs to drive a resistive load. voltage headroom is 1.5 V, so i don't know if that's enough to design a telescopic op amp. So far I've been going with a simple two stage op amp. The gain is only around 30 dB when driving a 2kohm load so I definately need to boost the gain. Its already using about 400 uA so i want to avoid increasing the power consumption...

I chose 2 kohm based on the noise requirement, but I think I may need to simply increase the load resistance if i can't boost the gain any other way.

cheers,
Aaron

Title: Re: op amp topology
Post by vivkr on Mar 10^th, 2008, 12:09am

aaron_do wrote on Mar 8^th, 2008, 4:10am:

Thanks for all the replies.

And yeah it needs to drive a resistive load. voltage headroom is 1.5 V,

I chose 2 kohm based on the noise requirement, but I think I may need to simply increase the load resistance if i can't boost the gain any other way.

Hi Aaron,

Why do you need to drive a resistive load, especially since you mention that you are able to choose it freely (based on noise requirements)? Are you making
the next stage as a resistive differential amp? Surely you have considered all the tradeoffs involved to size the R-load to a maximum.

If possible, I would recommend using a class-AB output stage with your 2-stage opamp. That might get you what you need.

Regards
Vivek

Title: Re: op amp topology
Post by Berti on Mar 10^th, 2008, 1:04am

Hi Aaron,

I agree with the other respones that a class-AB buffer will be a good solution. Because driving 2kOhm with an OTA you
just sacrifice gain. However, be aware of interference when using class-AB buffer.

Whether you can use a telescopic OTA or not depends on the requirements on signal-swing. I use telesopic
amplifiers at 1.2V supply (small output-swing).

Regards

Title: Re: op amp topology
Post by aaron_do on Mar 11^th, 2008, 6:31am

Hi,

I'm designing an active-RC filter. Increasing the 2 kohm resistor increases the noise while reducing the requirements on the opamp. Anyway i ended up designing two different opamps. The first stage opamp is very power hungry but provides good gain. That allows it to drive the 2kohm and reduce the effect of the noise of the following stages. So the following stages all use very low power op amps.

I think i will end up going with a telescopic op amp and i'll look into the output stage. Thanks for all the help,

cheers,
Aaron