NMOS or NPN input OPAM for Bandgap reference.

Hi, Guys.

I am designing an OPAM for bandgap reference with BiCMOS process.

To reduce the OPAMP offset, I used the vertical NPN as the differential input pair.

Because the NPN transistor's collector current is about 60nA, I am afraid that the input current (Ib) of the NPN transistor would lead to a negative effect on the bandgap.

I hope the bandgap circuit will get high uniformity output voltage.

I am not sure if this is OK? How do I deal with it?

Thanks a lot!

if you are afraid of using it then you can go for NMOS

since your concern is primarily the base current

MOS has no or literally 0 gate current that will not affect your apparatus

i dono more about it so eagerly waiting for others to fill it in

Major source of error in MOS. pair is often due to the input-offset voltage of the opamp. One means of eliminating this source of error is to switch-capacitor amplifier. If you can implement this technique or some another then design your Bandgap Amplifier on MOS input pair, otherwise use BJT. As you understand disadvantage of BJT is finite input impedance.

KonstanU said:

Major source of error in MOS. pair is often due to the input-offset voltage of the opamp. One means of eliminating this source of error is to switch-capacitor amplifier. If you can implement this technique or some another then design your Bandgap Amplifier on MOS input pair, otherwise use BJT. As you understand disadvantage of BJT is finite input impedance.

Click to expand...

Dear KonstanU, thanks for your reply.

According to the area limitation, I can't eliminate the input-offset voltage of the opamp with switch-capacitor amplifier or chopper amplifier, etc.

Based on your suggestion, the BJT would be more suitable than the MOS input pair for this application.

Because the impedance connected to the base of the input BJT would be large, 1.2M ohm, I am not sure if it is OK?

Thanks.

The base current will add another term to the bandgap TC.
This may be detrimental or beneficial (usually detrimental
as it tends to add curvature). But I recommend you try
both, since the details are what it's all about.

dick_freebird said:

The base current will add another term to the bandgap TC.
This may be detrimental or beneficial (usually detrimental
as it tends to add curvature). But I recommend you try
both, since the details are what it's all about.

Click to expand...

Hi, thanks for help.

As you said, there will be another term (Ic_opamp_input/Beta)*R added to the output, which would lead to a negative effect on the TC.

In this case, I care more about the reference voltage uniformity than the TC.

Cause the NPN has better uniformity than the NMOS input pair.

I can't try both in this project. So we can get high uniformity with NPN input pair? Is this correct?

Thanks!

NPNs (at least in an "analog" process where they are a
valued element and processing cares about mismatch)
will tend to have a better natural matching in a smaller
area than a NMOSFET.

You can do a chopped and filtered bandgap with good
results, provided that you can tolerate a high output
impedance and some chop-frequency residue. In a
clocked circuit this is probably true. In a low noise
continuous-time amplifier based circuit, probably not.

Back when I did bipolar products we'd get 1mV-range
natural offsets out of NPN front end op amps, few
mV out of ones that used a PNP follower in front of
the gain stage. You need a lot of MOSFET area to
match that performance (in today's terms) - although
back in the day, it was > 10x10um emitters and maybe
it's not that far different after all.

Your foundry should provide you matching data for
both types of device, and just going by the front
end current density of interest and the geometry
it would take for the matching you want, you could
make a value judgment.