Continue to Site

# Explain to me the PSRR in bandgap circuits

Status
Not open for further replies.

#### hrkhari

##### Full Member level 4
Bandgap Reference

Hi Guys:

I had designed a BGR reference circuit, the temperature variation of -20°C- 85°C gives a reference voltage variation of 2mV, and a 20% supply voltage variation gives a around 0.5mV, but the thing is my PSRR is quite screwed up, giving only around 30-40dB of supression. The journals do suggest PSRR in the range of 60-100dB of supression.

Upon consultation, I was informed that PSRR determines the immunity of the reference voltage to power supply variation, but I'm still skeptical on it. If it is true than my power supply sweep variation gives somewhat good reference voltage variation. Can anyone pls guide me in explaining the true means of PSRR and verify the validilty of the previous claim. Suggestion on methods of improving PSRR is also welcomed. Thanks in advance

Rgds

Bandgap Reference

PSSR = rejection of the power variation and indeed has to be greater as possible and has the bigest importance in mixed A/D circuits.

Re: Bandgap Reference

Let me try this:
Code:
                           20%
-53.6dB =  20 * log --------------
0.5mV/1.2V

that is PSRR ~ DC, and
giving only around 30-40dB of supression.
would be done by AC analysis.

But, AC result should approach DC result at very low frequency,
Suggest to check simulation file in detail.

Bandgap Reference

hi jcpu,
I think the code you gave out is just an average value of PSRR. the best way is to performe Ac analysis,righ?

Re: Bandgap Reference

psrr can be improved by improving the gain which can be improved through cascoding.,instead of current source use the cascode current source,it will reduce the variations.

Re: Bandgap Reference

hello hrkhari!

let's observe the AC simulation from 1Hz to 100Hz,
see if low frequency output voltage in the order of a few mV?

Re: Bandgap Reference

hi Hkhari
one thing u can do to increase the psrr is to cascode ur tail of opamp so that ur psrr will increase.
and other thing u can do is to increse the Ro by reducing the lambda ( means increasing the w and l and keeping the w/l ratio same by doing this ur R0=1/lambda*id will increas and psrr will increase.
hope this help
manish

Re: Bandgap Reference

other than this i suggest you to go thru the paper
"A General Relationship Between Amplifier Parameters, And Its Application to PSRR Improvement"
E Siickingcr, J Goette - IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1991 - ieeexplore.ieee.org

Re: Bandgap Reference

"A General Relationship Between Amplifier Parameters, And Its Application to PSRR Improvement"
E Siickingcr, J Goette - IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1991 - ieeexplore.ieee.org

thanks

Re: Bandgap Reference

jc2 said:

"A General Relationship Between Amplifier Parameters, And Its Application to PSRR Improvement"
E Siickingcr, J Goette - IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS, 1991 - ieeexplore.ieee.org

thanks

A general relationship between amplifier parameters, and itsapplication to PSRR improvement

Sackinger, E. Goette, J. Guggenbuhl, W.
Electron. Lab., Swiss Federal Inst. of Technol., Zurich;

This paper appears in: Circuits and Systems, IEEE Transactions on
Publication Date: Oct 1991
Volume: 38, Issue: 10
On page(s): 1173-1181

Re: Bandgap Reference

hrkhari said:
Can anyone pls guide me in explaining the true means of PSRR and verify the validilty of the previous claim.
Rgds
the PSRR is the rejection that happen at the o/p due to imposed noise on the supply
while what you have calculated "and a 20% supply voltage variation gives a around 0.5mV" is called the power supply regulation i.e if the supply voltage is drifted how the output voltage is changed this is far from the PSRR in meaning.
but i think they may be related by somehow i am not sure

Re: Bandgap Reference

ramy_maia said:
hrkhari said:
Can anyone pls guide me in explaining the true means of PSRR and verify the validilty of the previous claim.
Rgds
the PSRR is the rejection that happen at the o/p due to imposed noise on the supply
while what you have calculated "and a 20% supply voltage variation gives a around 0.5mV" is called the power supply regulation i.e if the supply voltage is drifted how the output voltage is changed this is far from the PSRR in meaning.
but i think they may be related by somehow i am not sure

Hi,

Regulation is the term used with voltage changes due to changes in load. PSRR means how well ouput voltage rejects changes in power supply voltage. Actually what jcpu wrote is correct. PSRR is generally associated with small signals. The value given by jcpu is for a large signal. And my gut feeling is that a small signal PSRR can't be worse than large signal one.

Status
Not open for further replies.