Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
06/29/09 01:24
Modified:
  06/29/09 01:33

Read: times


 
#166558 - Scheme to provide fast turn-on after full discharge
Responding to: ???'s previous message
As we have already discussed, some micros don't like slow power-ups. What they need is a full power-up, from 0V to Vcc in about 1msec. Unfortunately, in many cases it's not enough to only sharply turn-on Vcc, but also to provide a full discharge before the power-up. So, a good circuit should allow the micro's supply voltage to have fully discharged before power-up.

Also, in many schemes the threshold between turn-on and turn-of voltage is way to little, causing the ciruit periodically turning-on and -off, when the supply voltage has reached the turn-off level, resulting in an annoying "pumping". In the scheme I propose here, the turn-on level is 12.0V and the turn-off level is 8.2V. This seems to be huge, but remember, that the turn-on level sees the supply without any load.

Have a look the following scheme, maybe you find it useful. (The scheme hasn't been tested!)



The both zeners are BZX85 types (500mW). The circuits needs to be connected to the storage cap of power supply, behind the rectifier. The supply current drawn by the 7805 is assumed to be less than 0.1A.

The lower part of this circuit works properly when the supply voltage at the terminal reaches a level of about 3.5V. Then the supply voltage at LMV393 is at about 2V. To prevent the MOSFET IRF9540 from unintentionally turning-on at lower supply voltages, the MOSFET driver BC546 is blocked by the 5V6 zener, until the supply voltage reaches about 6V. Additionally, the supply voltage for the zener is filtered by the 100R and 100µF parts, allowing the lower part to stabilze within a few milliseconds.

The first LMV393 is a comparator providing a hysteresis: At 12.0V supply voltage the output goes low and at 8.2V the output goes high again. When the output is low, the 1µ cap becomes charged, making the output of second LMV393 going low after exactly 100msec. Within this time the output voltage of 7805 must be able to decrease to 0V. Add a resistor (here 100R) to guarantee this.

After the 100msec delay, the upper BS108 turns-off, allowing the BC546 to turn-on and pulling down the gate of IRF9540 to 0V. Then, the 7805 section is supplied via 4R7 resistor, crerating a time constant in combination with the associated caps. Adjust the 4R7 resistor and/or the caps in order to allow the output of 7805 to rise to full level within about 1msec.

The BC546 uses rather high impedances (4k7 resistors) to turn-on and -off the MOSFET, resulting in moderate switching times. Adjust these impedances if you need faster switching times, but remember that you will increase EMI by this at the same time.

Kai

List of 58 messages in thread
TopicAuthorDate
soft-start powersupply and controllers            01/01/70 00:00      
   supervisor chip or processor with good brownout detection            01/01/70 00:00      
      Per gives a good solution if you can tolerate restarts            01/01/70 00:00      
   Long rise times of Vcc are critical!            01/01/70 00:00      
      replace the 7805            01/01/70 00:00      
         Lot of information            01/01/70 00:00      
            look at PowerTrends (TI)            01/01/70 00:00      
            Some answers...            01/01/70 00:00      
               Bigget reset capacitor and resistor            01/01/70 00:00      
                  it is NOT the capacitor value            01/01/70 00:00      
                     Thanks Erik            01/01/70 00:00      
                        I partially disagree            01/01/70 00:00      
                           disagree with what?            01/01/70 00:00      
                              With thinking only in a reset chip.            01/01/70 00:00      
                                 Use the reset chip to reset the powersupply            01/01/70 00:00      
                        How about $0.15            01/01/70 00:00      
                     A small clarification            01/01/70 00:00      
                        3 pin regulators with uvlo            01/01/70 00:00      
                        Eh ?            01/01/70 00:00      
                           Only a clarification            01/01/70 00:00      
                           Steve, I meant turning on MCU supply            01/01/70 00:00      
                        and power down            01/01/70 00:00      
                           Erik, brown-out is handled by Silabs MCU!!            01/01/70 00:00      
                              Did I miss that?            01/01/70 00:00      
                                 But Erik...            01/01/70 00:00      
                                    In application notes, direct pull-ups have been used...            01/01/70 00:00      
                                       some idiot, not there any more ...            01/01/70 00:00      
                                          You seem to be wrong about the capacitor, Erik            01/01/70 00:00      
                                             Not necessarily...            01/01/70 00:00      
                                             the risetime has nothing to do with reset ... unless            01/01/70 00:00      
                                             which SiLabs part?            01/01/70 00:00      
                                                Andy..            01/01/70 00:00      
                                                   SILabs urban legends            01/01/70 00:00      
                                                      Interesting...            01/01/70 00:00      
                                                         you may            01/01/70 00:00      
                                                            Thank you, I really appreciate it!            01/01/70 00:00      
                                                      Clarifying, Erik, Kai            01/01/70 00:00      
                                                         The reset from the supply monitor will be there            01/01/70 00:00      
                                                            Good advice...            01/01/70 00:00      
                                                               I am guilty of not recalling all differences            01/01/70 00:00      
                                                         Only open-drain but not push-pull!            01/01/70 00:00      
                                                            Good one, Kai            01/01/70 00:00      
                                                            why the 'love' of the MAX1232?            01/01/70 00:00      
                                                               True, Erik            01/01/70 00:00      
                                                                  Have you seen this post?            01/01/70 00:00      
                                                                     Have you seen this post?            01/01/70 00:00      
                                                                        Turn on your brain, Erik, and read what I wrote!            01/01/70 00:00      
                                                                           a picture is worth 1000 words            01/01/70 00:00      
                                                                              So you haven't read the words??            01/01/70 00:00      
                                                               I was suggesting him the MAX1232...            01/01/70 00:00      
      Well, even more demanding...            01/01/70 00:00      
   Bigger, not bigget            01/01/70 00:00      
   Only external electronics can help you            01/01/70 00:00      
   Scheme to provide fast turn-on after full discharge            01/01/70 00:00      
      Regulator with output enable            01/01/70 00:00      
         You never exactly know what you get...            01/01/70 00:00      
            Bypass caps            01/01/70 00:00      
               Please don't stress this scheme too much...            01/01/70 00:00      

Back to Subject List