Voltage follower capable of up to 150mA delivery - what do I need?

Discussion in 'General Electronics Chat' started by peskywinnets, Apr 1, 2012.

  1. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0
    I've used opamp voltage followers in the past to great affect, but they can't supply a decent amount of output current into heavyish loads.

    I therefore seek a device that will output exactly the same voltage that it receives, but also upping the current drive ability to at least 15mA at about 3.5V ....what type of device or circuit should I be looking for?

    Many thanks.

    pesky.
     
    Last edited: Apr 1, 2012
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,753
    760
    If it is an OPAMP then u need TDA2030 or something similar

    Or do u need it to operate at 3.5V
     
  3. #12

    Expert

    Nov 30, 2010
    16,283
    6,793
    You can add as many higher current transistors to the output of an op-amp as you want, as long as you connect their output to the negative feedback loop.
     
  4. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0
    I actually need it to operate 3.3V (I said 3.5V, but that was incorrect)

    The problem with putting a tranny in the feedback loop (which I'd not thought of - so thanks!), is that presumably the VCC would need to be at least 3.3V *plus* the voltage lost across the tranny, which means the VCC would then need to be about 4V (assuming 0.6V-0.7V lost across the tranny)...I was hoping to get a solution here that would follow the input voltage 0->3.3V but with a VCC of 3.3V.

    I'm surprised they've not come up with a high current voltage follower in SOT23-5 or similar.
     
  5. #12

    Expert

    Nov 30, 2010
    16,283
    6,793
    So connect an npn common emitter to the op-amp with its collector pulling down on a pnp transistor connected to Vcc. That is the tightest saturation you can get with bipolar transistors and you can still connect the output of the pnp to the feedback loop of the op-amp.

    I'd say the next step is a MOSFET but they don't work at 3.3 volts.
     
  6. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    That will probably require some frequency compensation to keep it from oscillating.
    Peskywinnets, what kind of load are you wanting to drive?
     
  7. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
  8. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0
    I'm driving 18 x 3mm blue LEDS...I realise that PWM is normally the way to go with LEDS, but what I'm experiencing is that even with a PWM value of just 1 (10 bit PWM , therefore 1024 levels), that the LED is still a little too visible (a PWM level of zero has the LEDs fully off, therefore the brightness step between PWM level 1 & PWM level 0 is too great).....I'm therefore thinking that driving the the LED with a filtered PWM DC voltage between the 'just visibly off' DC voltage, right up to the correct LED FWD voltage (so for example for a blue LED, somewhere between say 2.5V at the lowest setting right to 3.3V at the highest setting), but to do this I need a voltage follower that can pump out enough current at the 3.3V end of things.
     
    Last edited: Apr 1, 2012
  9. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Last edited: Apr 1, 2012
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Post a schematic, including the LEDs, of how you want to drive them. I think you need current limiting, which you don't seem to have.
     
  11. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0

    I haven't a schematic (the present version using PWM and a MOSFET is all on breadboard, & the version I want to move to ...a high current voltage follower is in my head & what propmted me to post!), but will try & knock up something.

    My take (after looking at the LED voltage vs current chart) was that maybe current limiting wouldn't be needed *if* supplying the LED with less than it's rated FWD voltage (what I did was test a selection from the batch of leds I received, established the current I required by setting different voltages & then ran with that voltage. So for example if the current at 3.4V across the led is 30mA, I set the max voltage to be 3.3V across the LED yielding something like 20mA ...well under the rating - it all works well, except like I say, even at a PWM setting of 1 (for 1024 levels) it's a touch too visible (not bright, but still too high for a setting of 1/1024th!)
     
    Last edited: Apr 1, 2012
  12. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    All LEDs of the same type will not have identical forward voltage at a given current level, or, conversely, all LEDs of the same type will not have identical current at a given forward voltage level. If you put them all in parallel, some will draw more current than others, so some will be brighter than others. If they are all run at the same current, they will all be very close to the same brightness.

    Why can't you set the PWM to 0/1024?
     
  13. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0
    Point taken, but in practise, I'm not seeing major visual differences in brightness between all the LEDs

    I can...and the LEDS are fully off then...what's irking me is the difference in brightness between PWM 1/1024 & PWM 0/1024 ...it's too great...in my opionion a setting of 1/1024 should mean you've really got to get up real close to the LED see if they're even on...but presently it's quite clear they are on (from a distance), so therefore I wanted to convert the switched PWM 3.3V I'm usinging to drive them to a smoothed DC voltage....I'm pretty certain that a smoothed PWM voltage of 1 will have them pretty much off ...which will allow for a much smoother fade from full brightness! (which is ultimately my end goal)
     
    Last edited: Apr 1, 2012
  14. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    What is your PWM frequency? What does the driver circuit look like? Switching circuits can distort duty cycle, which will be especially noticeable at low (or high) duty cycles, such as 1/1024.
    Post a schematic of your driver.
     
  15. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    My Ultra-Bright Chaser projects use PWM for a brightness pot. With extremely short-duration pulses the LEDs were also too bright.
    I added a 1nF capacitor from base to ground on the cathode-driver transistor to reduce the level of the very narrow pulses and now it dims perfectly.
     
  16. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0
    my PWM frequency is 15.625Khz...which might seem crazy high for an LED, but I need it to be inaudible as this is in part of an audio circuit(well, at least inaudible for an old duffer like me!). Like I say, no schem, but the signal path is simple...... PIC PWM pin, driving a MOSFET gate, mosfet source is to ground, moseft drain is to LED cathode, LED anode connects to 3.3V


    thanks but that won't help here...I'm using a MOSFET, which obviously is just a switch, so all the cap will do is alter the on/off threshold.
     
  17. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    The 1/1024 pulse width should be 62.5ns. I doubt your driver /MOSFET can faithfully reproduce this.
    What MOSFET are you using?
    What are you driving it with?
     
  18. peskywinnets

    Thread Starter Active Member

    Jan 19, 2009
    47
    0

    The Mosfet I'm using is this one...

    http://www.nxp.com/documents/data_sheet/BSH105.pdf (it seems pretty fast at switching)
     
    Last edited: Apr 1, 2012
  19. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
  20. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Like all Mosfets it has a high capacitance.
    It switches fast when it is driven with a driver that is only 6 ohms or less.
    A high current from the driver quickly charges and discharges its high capacitance.

    Please post your schematic.
     
Loading...