0
\$\begingroup\$

Im designing battery powered application and need reverse polarity protection.

My system specs: 4.5v(3x AA batteries) in and system running on 3.3v. Im using MCP1711T-33I/OT LDO to convert 4.5v to 3.3v. My system will take about 50 milliamps maximum. Therefore i can't obviously use any diodes since voltage drop and power waste will be too big. I read some threads and people tend to suggest P-channel MOSFET. I was looking for low rds(on) MOSFETS and all their packages seem so huge. Like this one:

http://www.farnell.com/datasheets/2049687.pdf

Size isn't too big problem though(sot23 package would be perfect though). Can you suggest me anything else than P-channel MOSFET? If not, can you suggest me some decent P-channel MOSFET for my needs? Thanks in advance.

\$\endgroup\$
  • \$\begingroup\$ Are you seriously saying 3 mohm is huge? What type of wires do you use to connect these 3 AA batteries? Superconducting ones? \$\endgroup\$ – Huisman Jul 20 at 16:53
  • \$\begingroup\$ @Huisman He means package size \$\endgroup\$ – DKNguyen Jul 20 at 19:44
  • \$\begingroup\$ @DKNguyen I wonder... they refers to the rds(on) in the sentence. Moreover: "Size isn't too big problem though" \$\endgroup\$ – Huisman Jul 20 at 19:47
  • \$\begingroup\$ @Huisman I think he means he can deal with physical size, he just doesn't want to because it seems inappropriate for 50mA. \$\endgroup\$ – DKNguyen Jul 20 at 19:47
  • \$\begingroup\$ Yes, DKNguyen is right. Sorry for the confusion. I made up that sentence pretty badly. By saying huge i meant indeed package size and yes package size isn't too big problem, but it's huge for 50 mA in my opinion. It's power ratings are way over what i need. \$\endgroup\$ – Mini Tamm Jul 20 at 20:00
0
\$\begingroup\$

Im using MCP1711T-33I/OT LDO to convert 4.5v to 3.3v.
...
Therefore i can't obviously use any diodes since voltage drop and power waste will be too big.

If you are really concerned about voltage drop and power waste, I think you'd better focus on the MCP1711.
It is dissipating the difference between input voltage and output voltage so, 1.2 V * 50 mA = 60 mW at most.
Any RDS(ON) up to 1 Ω waists at most (50 mA)2 * 1 Ω = 2.5 mW.

I'd suggest to replace the MCP1711 by a high efficient step-down converter (next to using the P-mosfets suggested by others for polarity protection).

\$\endgroup\$
  • \$\begingroup\$ Yes, i just realized that myself too. Just in case i will ask is DMP2100U-7 MOSFET decent? I have burned down some MOSFETS in the past since i couldn't choose the correct one(farnell.com/datasheets/1669947.pdf). \$\endgroup\$ – Mini Tamm Jul 20 at 20:48
  • \$\begingroup\$ Also just to be sure, will MCP1603/B/L be better choice? farnell.com/datasheets/1669947.pdf \$\endgroup\$ – Mini Tamm Jul 20 at 21:02
  • \$\begingroup\$ @MiniTamm Yes, DMP2100U-7 is a good choice if it is for personal use (for product development it's not recommended for new design, use DMP2045U) \$\endgroup\$ – Huisman Jul 20 at 21:21
  • \$\begingroup\$ And the MCP1603/B/L looks a nice candidate too (I briefly looked at it) \$\endgroup\$ – Huisman Jul 20 at 21:36
0
\$\begingroup\$

With RDS of say 1 ohm you would have 50mV drop. This drop reduces the amount of energy you can drain from the battery a little, before the voltage drops too low, but otherwise it doesn't effect your system. Power loss in the RDS doesn't matter because you use a linear regulator so total power loss doesn't change. So I'd say you can easily find good enough MOSFET even in SOT23 package.

\$\endgroup\$
  • \$\begingroup\$ Im sorry, but what you mean by: Power loss in the RDS doesn't matter because you use a linear regulator so total power loss doesn't change.? \$\endgroup\$ – Mini Tamm Jul 20 at 20:14
  • \$\begingroup\$ A linear regulator works so that it adjusts its resitance so that it "burns" the excess voltage to achieve the correct output voltage. If you have some series resitance with a linear regulator it does part of the "burning". As a result the total power dissipated in the regulator and the series resistance is the same that the linear regulator would dissipate alone. \$\endgroup\$ – TemeV Jul 20 at 22:40
0
\$\begingroup\$

Even ~500mOhm RDSon is plenty low for this.You are searching for too low an RDSon for your application. I use the IRLML6402 (65mOhm) or IRLML6302 (600mOhm) because that's what I like. It makes little difference at 50mA. You'll notice a 30mV drop as much as you will notice a 3mV drop.

\$\endgroup\$
0
\$\begingroup\$

I have used Si2305 which works great. Check it out. It dropped <0.2v @150mA

\$\endgroup\$

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.