The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Mixed-Signal Design >> bandgap error amplifier bandwidth
https://designers-guide.org/forum/YaBB.pl?num=1271875994

Message started by ycm on Apr 21st, 2010, 11:53am

Title: bandgap error amplifier bandwidth
Post by ycm on Apr 21st, 2010, 11:53am

if a bandgap is loaded with a buffer amplifier, what is the constrain to select the bandwidth of the error amplifier in the bandgap?

one thing I can think about is the power, lower bandwidth lower power,

what else?

Title: Re: bandgap error amplifier bandwidth
Post by Mayank on Apr 21st, 2010, 9:53pm

hii

Error Amp BW should be as low as possible to filter out noise BUT  sufficiently large to accomodate major supply ripple.
So, it mostly depends on your supply profile & noise specs.

--
Mayank

Title: Re: bandgap error amplifier bandwidth
Post by ycm on Apr 22nd, 2010, 8:30am

thanks,

if we put a simple RC filter at the output of the bandgap to solve the noise issue, the bandwidth then mainly depends on PSRR spec? what is the TYPICAL bandwidth requirement, 100kHz?


Title: Re: bandgap error amplifier bandwidth
Post by thechopper on Apr 22nd, 2010, 6:01pm

you are correct...
BW will be set by PSRR requirements if bandgap reference is filtered by any means different from the opamp BW...

Best
Tosei

Title: Re: bandgap error amplifier bandwidth
Post by RobG on Apr 26th, 2010, 7:22pm

It depends on your application, but I generally favor higher bandwidth than the signals of interest (load and power supply noise). If you have transient loads you probably want a high bandwidth so the output will recover quickly.

Also, as mentioned, you need the bandwidth for PSRR. Filtering with a large cap doesn't help.. the transfer function from the supply to the output is often non-linear so AC noise on the power supply can result in a DC shift in the output voltage.. so you want good AC supply rejection. In this case the shift is DC, so a large filter cap won't help.

rg


Title: Re: bandgap error amplifier bandwidth
Post by loose-electron on Apr 27th, 2010, 12:22pm

I would keep the bandwidth of the amplifier low and passively filter at the output.

passive filter for the enviropnment noise and power line junk


RobG wrote on Apr 26th, 2010, 7:22pm:
It depends on your application, but I generally favor higher bandwidth than the signals of interest (load and power supply noise). If you have transient loads you probably want a high bandwidth so the output will recover quickly.

Also, as mentioned, you need the bandwidth for PSRR. Filtering with a large cap doesn't help.. the transfer function from the supply to the output is often non-linear so AC noise on the power supply can result in a DC shift in the output voltage.. so you want good AC supply rejection. In this case the shift is DC, so a large filter cap won't help.

rg


Title: Re: bandgap error amplifier bandwidth
Post by thechopper on Apr 28th, 2010, 6:33pm


loose-electron wrote on Apr 27th, 2010, 12:22pm:
I would keep the bandwidth of the amplifier low and passively filter at the output.

passive filter for the enviropnment noise and power line junk


RobG wrote on Apr 26th, 2010, 7:22pm:
It depends on your application, but I generally favor higher bandwidth than the signals of interest (load and power supply noise). If you have transient loads you probably want a high bandwidth so the output will recover quickly.

Also, as mentioned, you need the bandwidth for PSRR. Filtering with a large cap doesn't help.. the transfer function from the supply to the output is often non-linear so AC noise on the power supply can result in a DC shift in the output voltage.. so you want good AC supply rejection. In this case the shift is DC, so a large filter cap won't help.

rg


Bob:

I agree with Jerry...passively filtering the BG output will prevent AC noise from getting into the opamp and thus will also prevent noise mixing due to circuit asymetries resulting in DC shifts as you suggested. This beyond the low BW for the opamp.
Also a passive RC in seres with the supply line (if acceptable) will have the same purpose on the noise coming thru this line into the opamp...

Best
Tosei

Title: Re: bandgap error amplifier bandwidth
Post by RobG on Apr 29th, 2010, 8:41am

Sure, if you have external components big enough to keep all the noise from getting to the opamp you can use a poor opamp, but just using a low bandwidth opamp will likely cause problems. Similarly, a passive filter on the output will have series resistance so you are going to have poor DC load rejection. If you want to use an external cap for filtering you can make a case for passive filtering, well if bond wire inductance don't get ya coming back in... but it is hard to do with on-chip caps as a general rule.

So... a poor opamp with external components or a good internal opamp... you need the BW to be way lower or way higher than the signals of interest... that is kind of what it boils down to.

As I mentioned before, the transfer function from the power supply to the opamp output is non-linear so noise on the power supply will cause a DC shift in the output voltage unless there is enough gain to reject it -- even if you have a very large external cap. Be careful... it depends on the application.


rg

The Designer's Guide Community Forum » Powered by YaBB 2.2.2!
YaBB © 2000-2008. All Rights Reserved.