Protection of Reversed Polarity on DC plug-in

From Qi-Hardware
Jump to: navigation, search

Contents

[edit] Schematics

Sch. A:

advantage -

  • simplicity, low cost

disadvantage -

  • without protections on current and voltage limit suppressor, reversed polarity
  • needs a dedicated tolerance of DC adapter 5V +/- 5%

Sch. B, it's the feedback from Joerg[2].

advantage -

  • protections on current limited suppressor and reversed polarity;
  • using a general fuse that will be broken both positive and negative polarity input.
  • a traditional diode is enough.

disadvantage -

  • without protection on voltage limit suppressor;
  • one-time fuse needs to be replaced a new one when it broke on end user. But it's not necessary when using PTC[3] type fuse.
  • needs a dedicated tolerance of DC adapter 5V +/- 5%

Sch. B-1:

advantage -

  • protections on current limited suppressor and reversed polarity;
  • Zener diode can be an over-voltage protection for input.

disadvantage -

  • one-time fuse needs to be replaced a new one when it broke on end user. But it's not necessary when using PTC[4] type fuse.
  • need a resistor (or PTC fuse) to be as current limited when Zener breakdowns.

Sch. B-2:

advantage -

  • protections on current limited suppressor and reversed polarity;

disadvantage -

  • one-time fuse needs to be replaced a new one when it broke on end user. But it's not necessary when using PTC[5] type fuse.
  • better to use a Schottky Barrier diode which drops low forward voltage .

Sch. C:

advantage -

  • protections on current limited suppressor and reversed polarity;
  • using a general fuse that will be broken when positive input and over-current.
  • without risks on fuse broken when inputing a negative source.
  • no needs to replace fuse if using PTC type.
  • with few nano-amperes when MOSFET is OFF. This is mostly designed on consumer products when reversed battery-in.

disadvantage -

  • high cost when using PTC fuse and N-MOSFET[6]. Still without voltage protection.
  • there's resistance drain source Rds(on) existed internally while MOSFET is ON. An input voltage will be shared by that small internal milli-ohm resistance. A well-considerable design is necessary. A VSD source to drain drops forward voltage on body diode.
  • needs a dedicated tolerance of DC adapter 5V +/- 5%

Notes:

  • Since the adapter's rating current for Milkymist One is 2A, a trip current lower than or equal to 2A is safe. For example using a resettable fuse holding current 1.1A and it's trip current on 2.2A. Since this is sample that overs a trip current 2A. So using a trip current of 1.5A is preferred due to there's no common PTC with trip 2A at 20 ° C. So performance from datasheet[7]'s Fig. 9 Time-to-trip curve at 20 ° C when a expected fault 2A trip occasion happened:
if using microSMD110F, holding current 1.1A, trip current 2.2A, needs 10 seconds to trip.
if using microSMD075F, holding current 0.75A, trip current 1.5A, needs 6 seconds to trip. 
if using microSMD050F, holding current 0.50A, trip current 1.0A at 13.2VDC max., needs 0.6 seconds to trip.
if using microSMD035F, holding current 0.35A, trip current 0.75A, needs 0.2 seconds to trip.
  • Since 5V net/symbol used every where in RC2 design, thus any variance directly influences the specification met by:
all step-down regulators, two usb host connector power, a USB compliant supply will provide between
4.75 and 5.25V and at least 0.5A of continuous power. 4V3 supply for audio codec

[edit] Experimental Conditions

  • Equipments:
    • Agilent 34401A Multimeter
    • ABM DC Power Supply 8303D
    • Tektronix TDS1012
    • TES-1326S IR Thermometer
  • M1 RC2 0x1a/0x2c Board without reset ic & diodes to fix no configuration bug, only usb mouse and VGA monitor connected.
  • Software:
    • FLickernoise 0.1(built on Nov 19 2010), OS: RTMES 4.10.99.0, Platerform: Milkymist SoC1.0RC1, CPU: LatticeMico32, Board: M1(PCB rev.)
  • Data measured under circumstances: All average voltages and currents are measured 5 minutes after power-up and entered control panel. Bare M1 board without case. Room temperature is around 20°C degrees.

[edit] Sch. B

  • F2 - MICROSMD075F-2, 6VDC/0.40Ω; D14 - MBRA240T3G;
  • Since this schematic is simply understood while using non-reversed polarity source, so no more experiments for it.
  • According to F2 datasheet, it needs roughly 6 seconds to trip at 2A, 2.5s @ 2.5A, 1.3s @ 3A. D14 has forward max. 3A capability. Let's try to provide a 2.5A 5V non-reversed polarity source power:

[edit] Sch. C

  • F2 - MICROSMD075F-2, 6VDC/0.40Ω; Q3 - FDD6530A;
  • 100 times on testing reversed polarity input with - 5.00V and the results on If (A) is stayed at net 0.412mA.
  • According to LM4550B, the Analog Supply Range is Min.4.2V. A descended tune a DC JACK power input at 4.516V which voltage of (TP26) is 3.829V. A very loud noise starts up from LINE OUT. This extreme test conflicts with the recommended operating ratings.
  • From data below: when the input is 5.00V, regardless of the voltage VGS or TP26, they have been lower than the specification required. This is because Q3 its resistance Rds launch of formation of the results of the voltage divider effect. For example, the VGS is equalent to the required voltage of USB host, but from these tests they are all unqualified. This is not schematic C with wrong design. It's because we don't change a new DC adapater with dedicated output voltage range. If the parameters of good design can be best among the central values, but sometimes backfired.

[edit] Sch. D

  • room temperature: 33.3°C
DC Jack pluggin polarity VJ11, Voltage (V) VZ, Zener Voltage If stable(A)/Transient(A) Temperature Fuse(°C) Temperature Zener(°C) Voltage of symbol 5V marked on M1 RC2 TP1(V) - 3V3 TP2(V) - 2V5 TP3(V) - 1V8 TP4(V) - 1V2 TP26(V) - 4V3 Notes
non-reversed, 0x1e 5.25 5.209 0.76 / - 44.7 47.0 5.199 3.250 2.491 1.782 1.158 4.293 booted, L11(3.4mV), L10(3.5mV), Vfuse(38.4mV)
non-reversed, 0x1e 5 4.961 0.76 / - 44.9 47.4 4.951 3.250 2.491 1.782 1.157 4.294 booted, L11(3.3mV), L10(3.5mV), Vfuse(38mV)
non-reversed, 0x1e 4.75 4.710 0.76 / - 44.7 46.2 4.699 3.252 2.491 1.782 1.158 4.292 booted, L11(3.3mV), L10(3.5mV), Vfuse(37.8mV)

[edit] Reference

  1. Original Milkymist One RC2 schematic
  2. Joerg's feedback
  3. PTC Resettable fuses
  4. PTC Resettable fuses
  5. PTC Resettable fuses
  6. MOSFET
  7. complete datasheet

[edit] Link

  • MICROSMD110F-2, complete datasheet, IH(room temperature)A = 1.1, IT(room temperature)A = 2.2, Vmax(operating)V = 6, Time to Trip (sec. @ A) = 0.2 @ 8A values specified were determined using PCB’s with 0.030”X1.5 ounce copper traces.; OHMS WATTS AT 20°C, 6V, MIN = 0.07;
  • ON Semiconductor MBRA340T3G, Schottky (Diodes & Rectifiers) 3A 40V
  • Fairchild FDD6530A, MOSFET Power 20V N-Ch PowerTrench
Personal tools
Namespaces
Variants
Actions
Navigation
interactive
Toolbox
Print/export