The best bandgap TC you have done

[smilelangjun]

I have been doing bandgap for some time, and have tried some kinds of bandgap in standard cmos process using subtrate PNP bjt. The best temperatute coefficient is this,3σ of ΔV is ±6mv from -40 ° to 80°, the bandgap voltage is 1.2. ΔV =±|Vmax-Vmin|, if the temp of Vmax is smaller than Vmin, the ΔV is negative, vice versa.
The data came form test, not simulation, and they are statistic results, not the best one in thousands of chips.
Have you done a bandgap ? then what about your temperature coefficent ?

 

[wpchan05]

I have BG in my chips, which returns to me recently. I have not done analysis on the temp sweep. May I ask you how do you set up a -40C envirnoment ? I even dont know how to test my bandgap voltage other than at room temperature. Can you tell me what your setup is?

However, my bandgap voltages are staying consistent on different day measurements, and the variation of my bandgap voltage is around 50mV variations out of 10 chips.

 

[smilelangjun]

mount the chip on the pcbs, put them in a big machine which can warm and cool the pcbs. The gnd and power wires can thread out from a aperture.

 

[wpchan05]

Hi,

any recommendation for the big machine you suggested? I am interested in doing this test, but I don't have much money in buying a decent thermal bath. Thanks.

 

[smilelangjun]

It is a big machine which can adust the temperature and humidity. Its shape is much like a refrigerator, and its chinese name is 恒温恒湿箱. Mybe you can understand. I think it worth a lot of money.

 

[snafflekid]

He is talking about a temperature forcing system, Thermostream. They are expensive but accurate. If you cannot afford this, you can point a hair dryer at your chip or use some freeze spray and monitor the Vbg change. It will not be an accurate measurement!

 

[winsonpku]

I am curious about your architecture of your BG
I think the performance of your BG is very good.
Could you give more information about you BG, such as process, power consumption? Thanks!

It is a big machine which can adust the temperature and humidity. Its shape is much like a refrigerator, and its chinese name is 恒温恒湿箱. Mybe you can understand. I think it worth a lot of money.

 

[wpchan05]

He is talking about a temperature forcing system, Thermostream. They are expensive but accurate. If you cannot afford this, you can point a hair dryer at your chip or use some freeze spray and monitor the Vbg change. It will not be an accurate measurement!

Thanks for the keyword.

 

[smilelangjun]

I am curious about your architecture of your BG
I think the performance of your BG is very good.
Could you give more information about you BG, such as process, power consumption? Thanks!

It is a big machine which can adust the temperature and humidity. Its shape is much like a refrigerator, and its chinese name is 恒温恒湿箱. Mybe you can understand. I think it worth a lot of money.

The most popular architechtre with op, dont have correction。That is not very good, I have seen a bandgap with the same architechtre have only (ΔV)max=±3mv drift from -40 to 80 in one of ADI's SOC chip. But I dont know how they did it, maybe the layout is the most important reason.

 

[sreevyas]

simple way of measuring at 0 and 100 C
cover the PCB and chip in aluminum foil and put ice cubes on the box have zero deg mesurement
and similar way keep the foil near stem of boiling water and mesure for 90 c

most inexpensive way of doing it

__sree

 

[sreevyas]

simple way of measuring at 0 and 100 C
cover the PCB and chip in aluminum foil and put ice cubes on the box have zero deg mesurement
and similar way keep the foil near stem of boiling water and mesure for 90 c and extrapolate

most inexpensive way of doing it

__sree