The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
https://designers-guide.org/forum/YaBB.pl?num=1332169795

Message started by mixed_signal on Mar 19^th, 2012, 8:09am

Title: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by mixed_signal on Mar 19^th, 2012, 8:09am

Hi,

Has anyone taped out a base pinch resistor in IBM 130nm cmrf8sf? I am using it for process compensation in temperature sensor. What precautions have to be taken for that?

How is this process modelled in weak inversion?

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by mixed_signal on Mar 19^th, 2012, 8:22am

Hi,
If i use a pnp transistor then to what potential the emitter should be tied? How can I access the two base terminals?

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by mixed_signal on Mar 22^nd, 2012, 5:20pm

Is there anyone who can help?

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by RobG on Mar 23^rd, 2012, 1:50pm

I didn't realize people used pinched resistors any more. I expect the lack of response is because pinched resistors are rarely used. They used to be poorly controlled so people only used them if absolutely necessary. Better options include high sheet poly, nwell, or a long PMOS device. I haven't even seen a MOS process that characterized them... I have only used them in bipolar processes.

If you are using a pnp emitter to pinch a base you would want to tie that emitter to the lowest supply, and also the collector. The amount of reverse bias will probably affect the resistance.

Does this particular resistance offer some sort of process correlation that you need?

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by mixed_signal on Mar 23^rd, 2012, 8:52pm

Thank you for your reply...Yes!! In fact base pinch resistance has approx. 68% correlation with Is variation in a BJT which can be used for process compensation in temperature sensors.

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by RobG on Mar 24^th, 2012, 8:18am

mixed_signal wrote on Mar 23^rd, 2012, 8:52pm:

Thank you for your reply...Yes!! In fact base pinch resistance has approx. 68% correlation with Is variation in a BJT which can be used for process compensation in temperature sensors.

I wondered about that when I looked into something that would counter IS variations many years ago. Do you use the pinched resistor to determine the current in the BJT? Doesn't the Nwell also have similar correlation?

The Nwell resistances became so well controlled I didn't bother pursuing it given the other problems with Nwells (including increased curvature of the BG), although I guess I set myself up for variations in poly resistors.

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by Lex on Mar 26^th, 2012, 4:35am

RobG wrote on Mar 24^th, 2012, 8:18am:

My guess is that if you talk about the physical sizing of the resistor you're probably not far off with N-well only resistor. I guess the difference is in the actual concentration level in the N-well. With the P+ implants of the emitter you probably change the concentration of the N-well too, which you cover when you take the pinched resistor model.

For me it's also a while ago that I've seen those devices. I just checked some of the (MOS) processes I use currently, but no pinched resistor in there.

Nice topic btw ;)

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by loose-electron on Mar 26^th, 2012, 2:40pm

might want to re-direct the question a little.

You seem to want to do some form of process compensation.
Tell us what you are trying to do there and maybe we can give you
some other methods.

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by mixed_signal on Apr 6^th, 2012, 1:36pm

Hi,
The goal is to keep Ibias/Is constant since this ratio sets the Vbe.

Now, if Is increases then Rpinch increase. So, Ibias is derived in such a way that Ibias increases with increase in Rpinch. Consequently, Ibias/Is remains constant! (not exactly though :))

Thanks

Title: Re: Base pinch resistor in IBM 130nm cmrf8sf, temperature sensor
Post by mixed_signal on Apr 6^th, 2012, 1:40pm

Hi,

I want to find the integrated noise at the Vref of my temperature sensor. What should be my frequency range over I should integrate? My sensor will be followed by sigma delta. The sensor should give 1 reading per minute.

Thanks!