SG3525 (AKA KA3525)

I have a switch mode power supply I was trying to repair that got put to the side. Getting back to it now. It had many burned parts including SG3525 SMPS controller IC. Basically put’s out a PWM to drive the main power MOSFETS.

I am slowly tracing out all the circuits in the board. There are missing parts etc. There are also 3 pots that are involved. It is not triggering the mosfets (also replaced).

I decided to put one of these IC’s on a breadboard and play with it until I had a real clear understanding of it. I am studying the test circuit on page 4 of the white papers which I have included here. There are a few things am slightly unclear about. Please refer to the drawing on page 4 for the numbers I am referencing.

Near pin 10 there is a switch to shut down the chip. It looks like it is a normally open switch that is being pushed down LOL. Is this really supposed to be just a normally open momentary switch?

Near pin 7 same question. Is that supposed to be a normally closed momentary switch?

To the right of that the text says RT and there is a symbol. I believe they are just representing a test point here? RT is what that pin is called etc.

Some of the capacitors in this circuit have 3 diagonal lines connecting the plates. What does that mean please?

For pin 10 can any .7 v general purpose diodes be used?

Over near pins 11 and 14 on the right side there is a ganged switch. Looks like it can choose between VCC or ground for the outputs. I see it has to do with setting up collector and emitter voltages inside the chip but I don;t understand what it’s for.

What is deadtime please? I understand discharge is relating to the cap CT that sets up the frequency etc.

I have plans to use this chip for my own designs in the future if I can get my head around the finer details. When I have read enough I want to try to design a tube amp and maybe use a switch mode instead of a transformer if it turns out to be a good idea.

Thank You for any help you can give.

KA3525A.pdf (84.1 KB)

It's a switch - could be momentary (probably would be for testing), but likely for practical use, it would be controlled differently.

I think so on the second switch too - though if my impression of the purpose of the pins is right, that's only useful for testing.

Those two diodes are internal - notice that they're shown on the internal block diagram earlier in the document (this is horrible documentation, btw - most manufacturers will put like a dotted line around the parts that are internal when they show it like that)

That's an old chip - there are probably better parts available for new designs.

Can you tell if they are both normally open or closed or which? I agree....Horrible documentation :frowning:

Thank You

You can see that for yourself, just need to know what to look for.
Switches are supposed to be always drawn in rest (i.e. not actuated).
If the little bar you see, touches the contacts, it's a n.c.
If it doesn't it's a n.o.

In this case the button on the switch is also shown.
The one near pin 10 is n.o.
The one near pin 7 is n.c.

Furthermore, you should take a look at the pin descriptions.
The description near pin 10 is SHUTDOWN with a line over it.
The line over it tells it's an active LOW pin, so you have to pull it LOW to get that function.
Other ways to note this, is by preceding the text with a forward slash, or an exclamation mark.
Once you know this, you can also work out yourself which switches to use.

Is what I figured. The one near pin 10 should be drawn so that it looks like it has room to move when the button is pushed IMO. If you zoom in, neither button is drawn touching. Just wanted to make sure.

Thank You.


What is deadtime please?

In the case of your IC, dead time is the time between one output going from high to low and the other going from low to high.
It means that if you are driving an H-Bridge the outputs will be open for a time (dead) between each leg being alternately switched on.

Tom.... :slight_smile: