Schottky rectifier, linear supply, excess ripple?

[AmmoJammo]

I've recently acquired an old GoodWill bench PSU.

Along with the two adjustable outputs, it also features two fixed 5volt outputs. One rated for 1amp, the other 3amp.

Feeling I didn't need two current limited fixed 5volt supplies, I modified one to do 12volts instead. (the 1amp channel)

However, under a ~900ma load, the input to the regulator dropped too low for regulation and I ended up with 11volts on the output.

I've replaced the original rectifier with one made of MBR745 Schottky (to hopefully increase the voltage to the regulator) and now, on the multimeter, it will do a solid 12volts, with a 900ma dummy load.... HOWEVER!

It has 600mv of peak to peak ripple...

[mohare]

Did you upgrade the filter caps (on the rectified side)?  You've more than doubled your output voltage so you'll have to upgrade the filter circuits too.

[AmmoJammo]

Originally had a single 4700uF from memory, now it has two 3300uF in parallel.

The ripple across the rectified, filtered supply is a 1.2volt peak to peak, triangle wave.... 

Are Schottky's just not suitable at all for this application?

[madires]

What's the voltage of the transformer's winding?

[AmmoJammo]

What's the voltage of the transformer's winding?

I don't know, because its a premade supply, but with the ~900ma load on the DC output, the AC measures 12.7volts, and the rectified DC measures 15.7volts.

[madires]

That doesn't look too bad. Have you checked the ripple when the load is attached to the filter caps (in front of the voltage regulation)? It seems a little bit strange to use a 12V secondary winding for a 5V regulated output since this would be a waste of power. Do you have a schematic for the 5V regulation? Or is it simply a 7805?

[AmmoJammo]

That doesn't look too bad. Have you checked the ripple when the load is attached to the filter caps (in front of the voltage regulation)? It seems a little bit strange to use a 12V secondary winding for a 5V regulated output since this would be a waste of power. Do you have a schematic for the 5V regulation? Or is it simply a 7805?

The ripple across the rectified, filtered supply is a 1.2volt peak to peak, triangle wave.... 

It uses an LM723, which has a 1ohm resistor in series with the output for measuring current, resulting in the input needing to be 1volt higher than one may expect.

[dacman]

The LM723 has a 3 V dropout (the input needs to be at least 3 V greater than the output).  With 12 Vdc on the output through a 1-Ohm current sense resistor at 1 Adc, the output of the LM723 would be 13 Vdc, which means that the input needs to be at least 16 Vdc, even at the low peak of the triangular ramp.  It might work with, for example, a 2.5 V differential, but you would be out of the design criteria.

The Schottky diode is faster (usually) than a standard rectifier but is usually larger than a similarly rated (PIV and A) rectifier diode.

[AmmoJammo]

The rectified, filtered, unregulated voltage looks like this...

1st pic is no load, but occasionally, it changes to the second pic, for seemingly no reason. 

Third pic is with the ~900ma load after the regulation.

[madires]

With the LM723 based regulation you might get a stable 10V output. For 12V you would have to use something different with a lower dropout.

[macboy]

Try something lower dropout, but not a "true" low dropout regulator. That means, avoid anything with a PNP output device. They are not inherently stable and are therefore not suitable to a lab/bench supply where various types of loads may be applied. Almost anything with NPN output devices are inherently stable and should work well. Look at Linear Tech LT1086-12 which is a simple, 3-terminal regulator similar to a 7812 but with around 1 V dropout. Although it doesn't offer true current regulation, it will limit short circuit current to around 2 A or less (see datasheet).