2
\$\begingroup\$

I came across term "half bandgap voltage reference". I expect it is somehow related to bandgap circuit. Do you know what it is, know its schematic or principle of operation?

\$\endgroup\$
  • 2
    \$\begingroup\$ Maybe you should mention where you "came across" that term. \$\endgroup\$ – AndreKR Feb 8 '11 at 1:11
  • \$\begingroup\$ @AndreKR That was informal conversation with engineer from other company. He work on this type of circuits (design them). From conversation I got impression that it is standard industry name for such device but I did not get a chance to ask for details as we had to do other things. \$\endgroup\$ – mazurnification Feb 8 '11 at 5:50
  • 1
    \$\begingroup\$ Maybe he just didn't finish his work on a bandgap circuit, so it was only half. ;) \$\endgroup\$ – AndreKR Feb 8 '11 at 14:39
  • \$\begingroup\$ I got info what it is - it is synonym of sub-bandgap voltage reference. This is easily "googlable" term and several circuit solutions are posted on the net. \$\endgroup\$ – mazurnification Feb 8 '11 at 15:40
3
\$\begingroup\$

A bandgap voltage reference is a circuit that is relatively easily made on an integrated circuit. The basic premise is that it uses two different sized NPN transistors driven at the same current, and the sum of these two voltages can be an accurate reference voltage near the bandgap voltage of silicon at 0K, which is 1.25eV. There are many descriptions of the circuit, which began as the Brokaw cell, then the Widlar bandgap, and then the many variations you can see today. The Wikipedia page is a decent brief read.

It has several benefits:

  • It gives roughly the same result regardless of power supply (PSRR)
  • It can be temperature-compensated to give a very flat response over temperature
  • It derives its accuracy from the ratio of transistors instead of the absolute value of components, which is a major benefit for integrated circuits.

The traditional method for generating the reference voltage is to add the voltages at the two paths, which will give the typical bandgap voltage of 1.25V, or two forward diode drops. As supply voltages have dropped, there has been a push for lower output voltages. They typically use more current mirroring to derive an output voltage that is closer to a single diode drop.

While I have never seen the term "half-bandgap" while doing research on low-voltage reference circuits, there are plenty of bandgap reference based circuits with an output voltage of 0.5 - 0.7V.

| improve this answer | |
\$\endgroup\$
  • 3
    \$\begingroup\$ Thanks for the answer but what you are describing is "bandgap circuit", my question was regarding "half bandgap circuit". (Note: I am linking to the same Wikipedia article in question) \$\endgroup\$ – mazurnification Feb 5 '11 at 9:49
1
\$\begingroup\$

Sometimes we use a phrase like this to represent a portion of a bandgap circuit inside an IC. Often there isn't the feedback loop (which regulates the transistor currents to be equal), or it is used as a comparator with a built-in reference.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Thanks, that was good hint, but it turned out this was synonym for sub-bandgap voltage reference. \$\endgroup\$ – mazurnification Feb 8 '11 at 15:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.