3
\$\begingroup\$

I'm seeing the schematic of a Notebook HP Pavilion dv9000 (download). On page 17 in the corner left you can see this part:

images

Two questions:

  1. Why they didn't use a MOSFET instead of MMBT3904?
  2. What's the usage of capacitor C851? haven't seen to use a capacitor for a BJT at its base, so far.
\$\endgroup\$
  • 3
    \$\begingroup\$ My $0.02 : about $0.02. (However, 1.2V isn't very much for switching a MOSFET while it's plenty for a BJT) \$\endgroup\$ – Brian Drummond Aug 23 '17 at 13:47
  • \$\begingroup\$ Both alternatives are fine, it really is about cost (although MOSFET transistors typically sink less current than bjt). The capacitor is to filter noise that could be at the base of transistor, so it really acts as a DC switch. \$\endgroup\$ – lucas92 Aug 23 '17 at 13:53
3
\$\begingroup\$

A mosfet requires usually a fair bit more gate voltage than the transistor ~0.7V so that's an immediate advantage. For a crude N-fet switch / emitter follower it also means that you've got that same ~0.7V voltage drop over the transistor instead of the mosfet >2V.

BJT is a current-driven device so the base will drain the line charge by itself even if there is no pull down/up resistor. Transistor base is also not bothered at all by a high voltage drive signal as long as you have a suitable series resitor in base. A mosfet will fizz if you connect e.g. 48V signal to the gate. Beware of the reverse voltage rating thought.

The transistor base current on the other hand is a source of many headaches if you are using a transistor as an inverter. That 60µA current is perfectly fine for lighting up a signal LED for example..

\$\endgroup\$
  • 1
    \$\begingroup\$ You should add a short note about the capacitor, for a complete answer. Not sure what they are worried about without seeing more of the schematic but maybe you can take an educated guess.. \$\endgroup\$ – pipe Aug 23 '17 at 14:01
  • \$\begingroup\$ It may be a simple delay circuit, 10k + 100nF => around 1ms to turn the transistor on. Conversely it'll ignore any glitches. I'd go with the delay circuit, what it seems to do is to pull HT_VLD down until 1.2V_HT is on properly. The author is using rather confusing way of creating connections, there's no indication for example that the "HT_VLD" is actually a connection to elsewhere on the schematic.. But HT_VLD_RC does not, it's just a name. And the way the PDF file is created, the proper PDF search barfs on you. There's probably a circle in robot hell reserved for people like that. \$\endgroup\$ – Barleyman Aug 23 '17 at 14:45
2
\$\begingroup\$

This looks like a power sequencing circuit to enforce some control between the 1.2V_HT and 5VPCU lines. The 10k and 0.1uF capacitor delay the 1.2V_HT signal a few milliseconds before the BJT starts conducting.

As @barleyman said MOSFETS need tend to need more voltage at the gate than a BJT, especially if it needs to operate with less than 1.2v. Common,cheap MOSFETS such as the 2N7002 (also used in this circuit) have threshold voltages that can be as high as 2.5V.

\$\endgroup\$
  • \$\begingroup\$ Mosfet gate threshold is also usually not that well defined, for example the 2N7002P mosfet has 1.1 to 2.4 Vth at room temperature at 250µA. This grows to 0.6 to 3.2V for the 150C to -55C temperature range. The ubiquitous BC846 has 0.3 to 0.95 Vth from 100µA @ 150C to 100mA @ -55C range. \$\endgroup\$ – Barleyman Aug 24 '17 at 11:14

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.