More voltage please....

Beau Schwabe

I thought this might be useful to some....

I'm working on an app where one of the IC's require a little bit of a voltage at a relatively low current. I actually
needed 9V at about 4mA. The object below will provide at least 10mA at 9V (15mA at 5V) from a 3.3V 50mA
source. Overall this converter is about 50% efficient

3.3V * 50mA = 165mW

165mW / 5V = 33mA <--measured 15mA before regulator started dropping off (15mA/33mA = 45%)
165mW / 9V = 18mA <--measured 10mA before regulator started dropping off (10mA/18mA = 55%)

Revision History:
                  Version 1.0   -    original file created
                  Version 1.1   -    added schematic version for -V DC-DC converter
                                     and back EMF diode across transistor(s)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 7/24/2006 3:26:12 PM GMT

David B

What would it take to make negative 9 to 12 volts, at also maybe 5 to 10 mA? Would something similar work?

Beau Schwabe

David B,

I updated the schematics to include a "negative" DC-DC converter version. ....also added a back-EMF diode across transistor(s)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Phil Pilgrim (PhiPi)

Beau,

That's really slick. I was going to ask if the efficiency would increase if you used Schottkys, but I see the 1N419s are germanium! 'Didn't know you could even get those anymore.

For regulated output, it would be interesting to see if you could use the Prop's ADC hookup to close the loop and control the pulses...

-Phil

David B

Thanks! That'd be really cool, having one cog that generates custom voltages for a project.

Beau Schwabe

Phil,

Thanks... I have about a handful of 1N419's but I think for this circuit, just about anything would work.

"For regulated output, it would be interesting to see if you could use the Prop's ADC hookup to close the loop and control the pulses"

....Phil, are you standing over my shoulder again?!!..knock it off! get away! -just kidding


Actually I think varying the pulse width would aid in more efficiency, and also cater to regulating the output. Initially I tried the square
wave turning it (the 13.5kHz) completely ON or OFF depending on if the voltage was above or below a specified preset. Let's just say
it was not ideal. Instead of ON or OFF, I think varying the duty cycle would be the best approach here. I haven't had a chance to try
this though.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Brian Riley

Beau,

I was looking over the schematics in the spin file. With regards to the comments in the section on the negative voltage generator; shouldn't that read 7905 and 7909 rather than 7805 and 7809? The "79" series is the designation for negative regulators.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cheers ... brian riley, n1bq, underhill center, vermont
See the K107 Serial LCD Controller at
www.wulfden.org/k107/

Beau Schwabe

Brian,

Your absolutely right... I have corrected the spin file.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Brian Riley

Beau,

I recognized the error because I tend to do the same thing for similar structures, copy/paste, then edit the pasted, and I too often have to go back and get things easily overlooked. <sigh>

Now, in view of the magnitude (or lack thereof) of current involved, and board real estate being what it is ... why not suggest 78L05/78L09 and 79L05/79L09 ???

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
cheers ... brian riley, n1bq, underhill center, vermont
See the K107 Serial LCD Controller at
www.wulfden.org/k107/

Beau Schwabe

Here is another method using a charge pump that can be used to power a light duty 5V component from a 3.3V supply ... not quite as efficient, but the parts are easier to find.
This is designed for very light loads 5V@10mA max.

[b]CON[/b]
  [b]_CLKMODE[/b] = [b]XTAL[/b]1 + [b]PLL[/b]8X
  [b]_XINFREQ[/b] = 5_000_000

  PinA  = 0
  PinB  = 1

  Frequency = 400_000                                    'DC to 128MHz

[b]PUB[/b] CTR_Demo
    Synth(PinA,PinB, Frequency)                         'Synth(PinA, PinB, Freq)
    [b]repeat[/b]                                              'loop forever to keep cog alive


[b]PUB[/b] Synth(pA,pB, Freq) | s, d, ctr, frq
  Freq := Freq #> 0 <# 128_000_000     'limit frequency range
  
  [b]if[/b] Freq < 500_000                    'if 0 to 499_999 Hz,
    ctr := constant(%00101 << 26)      '..set differential NCO mode
    s := 1                             '..shift = 1
  [b]else[/b]                                 'if 500_000 to 128_000_000 Hz,
    ctr := constant(%00011 << 26)      '..set differential PLL mode
    d := >|((Freq - 1) / 1_000_000)    'determine PLLDIV
    s := 4 - d                         'determine shift
    ctr |= d << 23                     'set PLLDIV
    
  frq := fraction(Freq, CLKFREQ, s)    'Compute FRQA/FRQB value
  ctr |= pA                            'set PINA to complete CTRA value
  ctr |= pB<<9                         'set PINB to complete CTRA value

  [b]CTRA[/b] := ctr                          'set CTRA
  [b]FRQA[/b] := frq                          'set FRQA                   
  [b]DIRA[/b][noparse][[/noparse]pA&#093;~~                           'make pin output
  [b]DIRA[/b][noparse][[/noparse]pB&#093;~~                           'make pin output
     
[b]PRI[/b] fraction(a, b, shift) : f

  [b]if[/b] shift > 0                         'if shift, pre-shift a or b left
    a <<= shift                        'to maintain significant bits while 
  [b]if[/b] shift < 0                         'insuring proper result
    b <<= -shift
 
  [b]repeat[/b] 32                            'perform long division of a/b
    f <<= 1
    [b]if[/b] a => b
      a -= b
      f++           
    a <<= 1    

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Lawson

cool stuff, I think you could get closed loop regulation for this DC-DC converter with a schmidt triggered comparator.· Hm... I've only seen this described for a "buck" step-down DC-DC converter, not shure if this'd work for a "boost" or "inverter" style DC-DC converter.·

Basically the Inductor is given power untill the output reaches a value slightly above the target voltage.· At that point the inductor input is grounded untill the output drops slightly below the target voltage.· Now the Inductor input is powered up again and the cycle repeats.· A resistor bridge to two input pins should suffice for the input threasholds.· (i.e. the resistor bridge is setup to give·one pin a tiny bit more voltage than the other pin)·This system should provide a tightly regulated output, but the switching frequency is NOT fixed with this regulation scheme so any noise from the regulator won't come at predictable frequencies.·

Hope this helps,
Marty

P.S. it's amazing the stuff a kid learns reading the engineering trade magizines his dad keeps scattered arround his office at home :P