HACKED!: ATX Power Supply with Variable Output Voltage?

In a previous video I showed you how to convert a common ATX power supply into a crude lab bench power supply That features a 12 volt 5 volt and 3.3 volt output Such a supply is definitely a decent start when it comes to testing your electronic circuits But it comes nowhere close to the convenience of a proper lab bench power supply with a variable voltage output Now of course you could add a buck boost converter to the outputs of the atx power supply and and thus create a variable voltage output as well Only problem is that the converter takes up nonexistent additional space inside the supply housing So there must be another solution to this problem That is why in this video we will have a closer look at the circuit of such a switched-mode power supply and find out whether it makes sense to modify it in order to achieve a variable output voltage as well Let's get started First off I removed all the case screws from my atx power supply test subject and removed the top cover of the housing at first glance the circuit looks very big and confusing But when we consider the functional principle of a switch mode power supply it becomes much easier to understand on the left side we got our 230 volts Ac mains voltage which firstly gets rectified by full bridge rectifier And then smoothed out by capacitors in order to create a DC Voltage of around 325 volts This DC voltage is then switched on and off rapidly by some kind of transistor and Thus a PWM signal is created with a high frequency of around 100 Kilo Hertz Next this high-voltage high-frequency PWM voltage is connected to the primary side of a transformer which therefore induces a low voltage high frequency PWM voltage into the secondary side Last but not least this voltage gets once again smoothed out by capacitors in order to create the initially targeted 12 volt DC Now we can check this output voltage by connecting the green wire of our supply with a black wire Plugging in the power and measuring the voltage between a yellow and black wire As you can see here the voltage is around eleven point three five volts and if we connect different loads to this 12 volt output we can see that the voltage stays pretty much constant No matter how much current we draw from it That means that if we go back to the switching transistor part of the circuit There has to be some kind of feedback from the 12-volt output Which tells the transistor to increase the duty cycle of the PWM signal if we draw more current Because we need more energy on the output side to sustain the 12 volts So let's remove the circuit board from the housing and have a closer look at the ICs to find out which one was responsible for the PWM signal The two ones in the top left corner were LM358 and thus operational amplifiers While the third one in the same corner was a TPS3510 and thus a PC power supply protection IC The last IC was a UC3842 Which according to its datasheet seems to be the IC we've been looking for Sadly though there was no direct feedback connection between 12 volt output and the pins of the IC Instead the 12 volt output was connected to an optocoupler which on the other side connected to pin 2 of the IC AKA a the voltage feedback pin and while measuring the voltage of this pin while the supply was powered on I asserted a voltage of 2.5 volts In conclusion that means that there exists a kind of voltage divider between the 12 volt output and the voltage feedback pin Whose resistor values are chosen in a way that the 12 volts create a 2.5 voltage drop at the voltage feedback pin That means that by varying the resistor values the feedback pin will regulate the duty cycle so that the output voltage will always create the 2.5 volts at the feedback pin And thus we can achieve a variable output voltage So I had a closer look at the passive components around the IC and noticed a suspicious 10 kilo ohm SMD resistor after removing it I soldered the two wires to a 50 kilo ohm potentiometer and soldered it to the pads of the SMD resistor Then I set the potentiometer value to 10 kilo ohm turned on the supply and checked whether the 12 volt output still functioned correctly Which it did and by slowly varying the potentiometer value the output voltage did change accordingly But only upwards and only to a certain degree Since the over voltage protection of the supply was still active Luckily though another optocoupler close to the PWM IC was directly connected to pin 3 of the PC Power supply protection IC which was the full protection output pin whose job was to trigger the over voltage protection To get rid of this protection feature I simply shorted two pins of the optocoupler which resulted in a successful voltage adjustment up to 16 volts Which was the maximum Since the output capacitors were only rated for up to this voltage The only remaining problem was that I was not able to lower the voltage beneath 12 volts So I unsoldered a second suspicious resistor and replaced it once again with a potentiometer This time a 100 kilo ohm one And due to my sheer "luck" the atx power supply was done for after this modification Which almost led me to give up on this modification attempt But thankfully I had a backup supply laying around whose overall circuits looked a bit simpler and best of all its PWM IC the KA7500 was directly connected to the 12 volt outputs through its pin 1 That means all I had to do was to remove this one pin from the circuits prepare a new potentiometer with three wires Solder its right wire to the 12 volt output the left wire to the ground output and the middle wire to pin 1 of the IC Now by powering the supply I was capable of varying the output voltage easily between 3 volts and 12 volts but it seems like an over-voltage protection was still active But nevertheless after adding two more resistors to adjust the output voltage more fluently and Adding a proper load to the output I put a stop to this modification As you can see it does work but while tinkering on such a mains voltage circuit that has the potential to kill a person and constantly trying to disable protection features I realized that this is not the way to go if you want a variable power supply So I do not recommend to replicate this modification Instead buy yourself a beginner's lab bench power supply They are not that expensive Anyway, I hope you enjoyed watching this modification attempt. Don't forget to like share and subscribe Stay creative and I will see you next time

Channel: GreatScott!

Views: 828,697

Rating: 4.8087115 out of 5

Keywords: atx, ATX, power, supply, hack, hacked, variable, output, voltage, lab, bench, tutorial, how to, diy, project, make, switch, switched, mode, smps, SMPS, feedback, resistor, potentiometer, functional, principle, explain, explanation, mod, modification, electronics, greatscott, greatscott!, pwm, PWM, duty, cycle, modify

Id: oeNAhP-GIjo

Length: 7min 44sec (464 seconds)

Published: Sun Jul 09 2017