Modifying power brick output voltage question
Good day folks. With the joint I work in starving for parts, I use everything at my disposal and stock everything as much as I can. Take this situation: not a single 12v brick around ! Instead, I scrounged up two HP printer bricks. They're rated at 36v 2A, which is useless for anything other than powering the printers they're designed for. What do we do ? Hack them, of course. I did this in the past successfully on a Vaio adapter, with the help of some of our wonderful chaps here, but this one gave me a bit of trouble and I was wondering what I'm doing wrong. I'm aiming for a large drop here as well: 12v out of a 36v adapter, unlike the previous one where I only dropped it from 16v to 14v, so I'm not even entirely sure this is possible. It's not so much about whether I need it or not (that surely doesn't help 
), but mostly about learning how such a circuit works.
It was the nicest and most wonderful thing to take apart: 3 screws holding the two halves together ! Pure bliss - if only all adapters were put together like this
I had a look around the board and identified the AS431 reference IC, responsible for driving the single optocoupler. I drew a schematic of the feedback loop on the primary, though please note that the values you see are the actual numbers written on the SMD resistors - I know I should've marked them as they are on the board, but it's just a personal choice when drawing schematics involving SMDs, because my brain automatically processes the first thing it sees, rather than the part number on the silk screen.
Based on the schematic, which we assume to be correct at this point, the resistor I chose to modify was R49 (originally a 01D resistor, meaning 100kOhms). Since it goes between the output V+ and the REF pin of the AS431, lowering this value should also lower the output of the brick, since the transistor inside the AS431 conducts more, turning on the LED in the optocoupler nice and bright, which prompts the primary to decrease the frequency.
To check this, I started off with a 20k trimmer pot turned all the way "up" in place of the resistor: 20k instead of 100k did indeed lower my output, but now it was TOO low - 8v. I eventually found a 31kOhm resistor in place of the trimmer, which brought my output closer to what I was expecting: 13.5v.....sort of. See the trouble I'm having is that it's not stable for one thing (it fluctuates between 13v and 13.8v, so regulation is off) and it also lacks current: the slightest load immediately kills it and it goes all over the place. The green power LED I marked as P.LED in my doodle also pulses regularly, like twice every second and I think the actually issue is here: I gave that LED no second thought, but then had a closer look and noticed it's what actually powers the optocoupler, if I'm not mistaken. My theory is that with the circuit tuned for 36v, with the output now at 12v, the 129 resistor in series with the LED (in reality R51 on the board) is too low for the LED in the optocoupler to turn on, causing the primary to boost the output, but as this happens, the optocoupler suddenly turns on and the primary cuts back again, then the cycle repeats, hence why I'm getting this "semi-regulated" output - it doesn't overshoot, but it's not stable either. The LED ITSELF could be playing a part in this: with V+ at 12v, the drop across it is now too low
Solution: remove and jump the LED OR replace R51 with a lower value.....which should also help keep the LED on during operation, which would make the end result more professional. Any thoughts based on my doodles there ? Thank you.
), but mostly about learning how such a circuit works.It was the nicest and most wonderful thing to take apart: 3 screws holding the two halves together ! Pure bliss - if only all adapters were put together like this
I had a look around the board and identified the AS431 reference IC, responsible for driving the single optocoupler. I drew a schematic of the feedback loop on the primary, though please note that the values you see are the actual numbers written on the SMD resistors - I know I should've marked them as they are on the board, but it's just a personal choice when drawing schematics involving SMDs, because my brain automatically processes the first thing it sees, rather than the part number on the silk screen.
Based on the schematic, which we assume to be correct at this point, the resistor I chose to modify was R49 (originally a 01D resistor, meaning 100kOhms). Since it goes between the output V+ and the REF pin of the AS431, lowering this value should also lower the output of the brick, since the transistor inside the AS431 conducts more, turning on the LED in the optocoupler nice and bright, which prompts the primary to decrease the frequency.
To check this, I started off with a 20k trimmer pot turned all the way "up" in place of the resistor: 20k instead of 100k did indeed lower my output, but now it was TOO low - 8v. I eventually found a 31kOhm resistor in place of the trimmer, which brought my output closer to what I was expecting: 13.5v.....sort of. See the trouble I'm having is that it's not stable for one thing (it fluctuates between 13v and 13.8v, so regulation is off) and it also lacks current: the slightest load immediately kills it and it goes all over the place. The green power LED I marked as P.LED in my doodle also pulses regularly, like twice every second and I think the actually issue is here: I gave that LED no second thought, but then had a closer look and noticed it's what actually powers the optocoupler, if I'm not mistaken. My theory is that with the circuit tuned for 36v, with the output now at 12v, the 129 resistor in series with the LED (in reality R51 on the board) is too low for the LED in the optocoupler to turn on, causing the primary to boost the output, but as this happens, the optocoupler suddenly turns on and the primary cuts back again, then the cycle repeats, hence why I'm getting this "semi-regulated" output - it doesn't overshoot, but it's not stable either. The LED ITSELF could be playing a part in this: with V+ at 12v, the drop across it is now too low
Solution: remove and jump the LED OR replace R51 with a lower value.....which should also help keep the LED on during operation, which would make the end result more professional. Any thoughts based on my doodles there ? Thank you.
Attached Files
)
Comment