Lower output voltage from 12V to 8V possible ?

[Gabriel]

Got this 12V 1.5A psu and I wonder if it is possible to lower the output voltage to 8-9 V ?

I have no idea if it is possible without redesigning the whole psu or changing the transformer.

The TO220 on the left is a SB10100FCT and the TO220 on the right is a KSH13007A if it matters.

[Th3_uN1Qu3]

Re: Lower output voltage from 12V to 8V possible ?

Possible, just that it'll need a higher minimum load to work properly.

You can adjust output voltage by adjusting the resistors on the TL431, which is the TO-92 package on the secondary side of the optocoupler. Just to the left of it in picture 2.

The TL431 will have a voltage divider from output to ground, with its middle point connected to pin 3, by replacing the lower resistor of the divider (the one which goes from pin 3 to ground), with a higher value it is possible to decrease the output voltage.

[Gabriel]

Re: Lower output voltage from 12V to 8V possible ?

The TL431 will have a voltage divider from output to ground, with its middle point connected to pin 3, by replacing the lower resistor of the divider (the one which goes from pin 3 to ground), with a higher value it is possible to decrease the output voltage.

Holy crap, how would you know that ? sometimes it hurts when I think about what bus to get on just to go home.

Will have a look at this tomorrow.

[ben7]

Re: Lower output voltage from 12V to 8V possible ?

Looks like R11, the through-hole resistor next to the TL431, is the lower resistor (from adjust (pin 3) to ground).

Looks half decent, with a fuse, and input filtering. I'd just replace those caps too

[Gabriel]

Re: Lower output voltage from 12V to 8V possible ?

Just got it working.

1.3Kohm 12.11V
1.5Kohm 11.04V
2Kohm 8.89V
2.2Kohm 8.33V

Ended up using a 2.2Kohm resistor for R11, I see no voltage drop under load.

I'm using this for a brother p-touch label writer which draws 6W - no problem so far.

What are the drawbacks changing the output voltage this way ? more rippel ? less efficiency ? possibility for psu to oscillate ?

Why does this psu have no lytic startup cap as many other have ?


Th3_uN1Qu3 and ben7, thanks a lot for your help - Both of you were spot on.

[ben7]

Re: Lower output voltage from 12V to 8V possible ?

Just got it working.

1.3Kohm 12.11V
1.5Kohm 11.04V
2Kohm 8.89V
2.2Kohm 8.33V

Ended up using a 2.2Kohm resistor for R11, I see no voltage drop under load.

I'm using this for a brother p-touch label writer which draws 6W - no problem so far.

What are the drawbacks changing the output voltage this way ? more rippel ? less efficiency ? possibility for psu to oscillate ?

Why does this psu have no lytic startup cap as many other have ?


Th3_uN1Qu3 and ben7, thanks a lot for your help - Both of you were spot on.

This PSU appears to be a self-resonant type. So yes, it might oscillate. But that is much more likely if the capacitors on the secondary go bad, since there is no 'critical' capacitor on the primary side.

This 'critical' capacitor would, in certain computer power supplies, such as the infamous Bestec ATX-250 's, cause the normal oscillation to go wacky. The output voltage can go from the normal 5v, to 20+ volts! This easily kills motherboards.

As long as the secondary rectifier diode could handle the current, this could output its original 18w spec (12v, 1.5A) at 8v - which would be 2.25A. So It will provide plenty of power for the label writer machine.

Moreover, Th3_uN1Qu3 knows a crap-ton about SMPS design , so I am sure he will have more to say (or possibly correct me )

[tom66]

Re: Lower output voltage from 12V to 8V possible ?

You also have to consider capacitor ripple current and transformer ratings.

[ben7]

Re: Lower output voltage from 12V to 8V possible ?

You also have to consider capacitor ripple current and transformer ratings.

I thought so, but wasn't sure - better to keep my spewing mouth shut

[Th3_uN1Qu3]

Re: Lower output voltage from 12V to 8V possible ?

Holy crap, how would you know that ? sometimes it hurts when I think about what bus to get on just to go home.

It comes from experience. Seen and fixed many of them, have a prototype right now on my bench, that i built myself from the ground up. Anytime i see an opto with a TO-92 next to it, i immediately think TL431.

What are the drawbacks changing the output voltage this way ? more rippel ? less efficiency ? possibility for psu to oscillate ?

Mostly stability issues. Two things happen at low load when a SMPS of this kind is modified for lower output voltage. First, the gain of the circuit increases, which means the crossover frequency increases as well. This can lead to oscillation at low load. Secondly, even though this is a self-oscillating design, it still has an auxiliary bias winding that provides the voltage for driving the power device (as a matter of fact, this is what starts up the oscillation in the first place). When you lower the output voltage, the voltage on this winding decreases as well, so it takes a higher minimum load at the lower voltage, for the auxiliary winding to come up to its designed operating parameters.

If the device you are powering draws a fairly constant amount of power, above the minimum load limit of your SMPS, this will not be a problem. You will recognize a too low load by a characteristic ticking sound coming from the transformer - this is commonly known as "hiccuping". It is a low frequency oscillation that occurs when the voltage on the aux winding drops too low to continuously power the circuit on the primary, so oscillation stops for a little while, then after building up enough voltage it resumes, and so on. A power supply operating in this mode will interfere with the attached device.

Ripple will not be an issue unless you pull more current from it than it was originally rated for.

[Gabriel]

Re: Lower output voltage from 12V to 8V possible ?

How would you know what the minimum load limit would be ? is it possible to calculate it based on the components ?

Is it a better idea to fit a lower ohm resistor to a 5V supply to get it up to 8V compared to fit a higher ohm resistor to a 12V supply to lower it to 8V for stability reasons ?

[Th3_uN1Qu3]

Re: Lower output voltage from 12V to 8V possible ?

How would you know what the minimum load limit would be ? is it possible to calculate it based on the components ?

It's dependent upon the topology, control method used, number of turns on the auxiliary winding... It can't be calculated without a full schematic, including inductances of the transformer.

Is it a better idea to fit a lower ohm resistor to a 5V supply to get it up to 8V compared to fit a higher ohm resistor to a 12V supply to lower it to 8V for stability reasons ?

Largely, yes.

[tom66]

Re: Lower output voltage from 12V to 8V possible ?

I've played around a lot with the self-oscillating two transistor design. I kind of finally understand it -- ish. The auxillary winding is important for start up but it forms a much more important part of the operation of the supply. It essentially drives the MOSFET gate effectively allowing it to operate at high frequency and reasonable efficiency. The gate drive capacitor couples the drive pulses so the supply can start up (a large capacitor slows start up but increases drive current.)

This design will also start up with just the load from charging the output capacitors up, then hiccup to keep the output in regulation, and still be fine with heavy load after doing so. See below. It doesn't require a TL431, can just use a zener for feedback.

Only real application I can see a supply like this useful for is driving LEDs. It's very efficient and cheap. For standby supplies it's too expensive to implement for more than about 2A. Though I have seen a three-transistor variant in a 32" LCD TV... for the backlight inverter power! (About 150W.)