I am designing a bandgap with ppm<10 and vref=2.5

[samuel]

I am designing a bandgap with ppm<10 and vref=2.5, anyone can give me some advice?

 

[palmeiras]

Hi Samuel.

Could you be more precise in which type of advice you are expecting?

For supply voltage equal to 2.5 volts.... the most part of topologies found in literature is suitable. Traditional circuit can be used.

Regarding the temperature performance, (lets consider for instance -40 to 125 degrees celsius) probably, you will need a trimming circuit.

10 ppm/C stands for around 2 mV of VREF variation over the whole temperature range. To achieve this specification for a 3-sigma requirement for your circuit, you will probably need a trim circuit.

I say another thing... Maybe, depending on the process and devices you have, to achieve such requirement (10 ppm/C) you may need a curvature compensation.
For this task, there are several works I can suggest you.

Tell me other doubts you have...

 

[samuel]

thanks,palmeiras,

my spec is followed:
supply voltage 6V;
Vref:2.5V
ppm:10
temperature range:-40~80

 

[palmeiras]

Hi Samuel,

As your temperature range is not so wide, maybe you dont need curvature compensation. You need to simulated it. Because this may depend on the process. Do a simple test: design a traditional first-order BGR with one amplifer, and check the temperature coefficient.
However, regarding the impact of fabrication process, you will need a trim resistor.

 

[samuel]

thanks,palmeiraca

can you give me some materials about bandgap?

 

[palmeiras]

Hi Samuel!

For general knowledge about Bandgap reference:

1) Design of Analog CMOS Integrated Circuits
Behzad Razavi

2) Cmos Analog Circuit Design
Phillip Allen / Douglas Holberg's

3) Analysis and Design of Analog Integrated Circuit
Paul Gray

Other references:

1) oevlsi.ece.cornell.edu/ece453/fa05/projects/Bandgap.ppt

2) **broken link removed**

3) **broken link removed**

4) **broken link removed**

5) **broken link removed**

It is not complicated to find general references about BGR. If you need another type of information, tell me.

Best wishes,

 

[samuel]

thank you very much.palmeiras

if Vref is 2.5V, is it possible to provide 5mA current? it is to say, load is 500oms.

 

[palmeiras]

Hi Samuel,

do you mean the reference circuit provides VREF = 1.2 V (that is common) and also supply 5 mA of current?

Well.... What usually I do is connect a buffer (amplifier in negative feedback configuration) in the output of the BGR circuit. In this way, the buffer provides the necessary current.

Because the most part of BGR circuits presented in the literature are not able to supply current. (There are some examples of BGR that realize this task).

For instance, take a look at the following paper:

`A 30 uW CMOS Bandgap reference featuring a 1.4 - 6 mA output driving current and a Miller-effect startup circuit`.
David. C.W. Ng, David K.K. Kwong, Ngai Wong.
Microelectronics Journal. 2009.

Let me know if this explanation is not clear.

Best regards,

 

[samuel]

can you provide this paper?

 

[palmeiras]

For sure... Monday I will send it to you.
Because I dont have this paper at home.

Talk to you later...

 

[samuel]

my specs:

voltage supply :6V;
ppm; <10;
output reference:2.5V;

load resisor :500ohms.

 

[palmeiras]

As your output reference must be 2.5 volts.... Maybe banba topology is a good option. You can choose the value of the output voltage. And also connect a buffer in the output amplifier, to provide current.
Think about it.

. H. Banba, H. Shiga, A. Umazawa, T. Miyaba, T. Tanzawa, S. Atsumi, and K. Sukui, "A CMOS bandgap reference circuit with sub-1-V operation." IEEE Journal of Solid-State Circuits, vol. 34, no. 5, pp. 670-674, 1999.

 

[samuel]

thanks.

can a buffer provide a driving of 5mA using CMOS process?

 

[palmeiras]

Yes.... it can provides; for sure.
If you have difficult to design the buffer to provides 5 mA, you can put some buffers in parallel to provide the required current - Once I did this to save design time. But for sure, it is possible.

 

[samuel]

thanks.

can you find this paper?
3 V 110 μW 3.1 ppm/°C curvature-compensated CMOS bandgap reference
期刊 Analog Integrated Circuits and Signal Processing
出版社 Springer Netherlands
ISSN 0925-1030 (Print) 1573-1979 (Online)
期 Volume 62, Number 2 / 2010年2月
DOI 10.1007/s10470-009-9333-7
页 113-119
学科分类 工程学
SpringerLink Date 2009年6月26日

 

[palmeiras]

I send a PVT message to you.
Best Regards,

 

[samuel]

I have design banba structure bandgap. the test result is as follows,

can anyone explain this result? It seems that positive temp coeffecients is too small.

 

[palmeiras]

Hi Samuel. Did this circuit work now?
It seems to be not properly designed. Could you add more information?

 

[samuel]

Hi Samuel. Did this circuit work now?
It seems to be not properly designed. Could you add more information?

OK, It does work, but this figure shows that bandgap voltage changes larger at low temperature than that at higher temperature.

 

[palmeiras]

Can you upload this figure?
How much is this difference? In terms of mV, uV? Small difference is expected.