The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> OTA replica bias
https://designers-guide.org/forum/YaBB.pl?num=1208272841

Message started by Berti on Apr 15^th, 2008, 8:20am

Title: OTA replica bias
Post by Berti on Apr 15^th, 2008, 8:20am

Hi all,

Today I discovered that I can save a lot of power when I scale down the current in the replica-bias
of my power-hungry OTA by 1000 ;) (see a very simplified figure in the attachment).
However, where I work it is good practice to scale the current only by 8 to 12 and further current multiplication
is done in a seperate bias circuit.
Why should I not scale this current directly by 1000 (assuming it wouldn't be a problem in the layout)????

Cheers

Title: Re: OTA replica bias
Post by jiesteve on Apr 15^th, 2008, 9:03am

I can think of 2 issues:

1. The matching will probably not be good with the high M factor (1:1000)
2. The current mirror will be very, very slow. Any sort of noise that couples into the gate node of the current mirror will take a long time to settle, which could impact the overall settling of the amp.

Title: Re: OTA replica bias
Post by thechopper on Apr 15^th, 2008, 6:57pm

Hi Berti,

I agree with what jiesteve stated. In addition, regarding #2, not only the current mirror will be very slow, but also if you happen to copy such bias current from a bandgap reference circuit you might be generating potential stability problems (due to the large load cap) for such circuit if the mirror is directly copied from such reference.

Regards
Tosei

Title: Re: OTA replica bias
Post by Berti on Apr 15^th, 2008, 10:30pm

Hi Tosei and Steve,

Thank you for the answers. Matching is probably not so important, but I agree that the slow speed might be
a problem. Also thanks for the remark concerning the bandgap stability.

Regards