Beginner needs Schematic expert review


#1

Hi!

Firslt, I am very new to KiCAD but Being passioanate about circuit design, I am willing to take the dive. This is the schematic of the primary side of a power supply. My instinct says it has so many flaws from design point of view especially it leading to the pcb. I want to get a review on it so I can improve any suggestions are more than welcome and Thanks for taking the time to take a look.


#2

Ahhh. That transformer (L_core_ferrite_coupled) is sideways. If you were to plug this into the wall as is, you would almost certainly injure yourself.

You want the transformer to physically isolate the hot+neutral side from the low voltage side. What you’ve drawn would pass everything through, blowing up your downstream components and possibly electrocuting you if you touch part of the circuit that was supposed to be at a lower voltage.

I’m struggling here because I really don’t want to sound too negative and turn you away from electronics – I definitely think it’s good that you’re curious and I want to encourage you to continue learning and building things. But given the specific risks of making a mistake with mains electricity, I think it would be a good idea to hold off on a mains power supply design like this until you’ve read a little bit more about it and are confident that you understand how to do it safely. Making mistakes is part of learning, and it’s much much better to make mistakes at 5 or 10 volts than at 100 or 200 volts. The high voltage stuff isn’t difficult, but the cost of a mistake can be very high. Make your mistakes at a lower voltage, and then move on to higher voltages once you’ve gotten past the smaller, safer explosions.

In general, it’s a good idea to make notes in your schematics so that it’s easy to understand what you’re trying to do. Text notes like “AC input section” “Rectified output” “Output filter” “Regulator” can be super useful for others, and for yourself when you come back to a schematic after a few months and don’t remember what you were trying to do. It’s usually a good idea to group things together by function rather than as groups of similar components. Pulling out the power mosfet like you did makes it really difficult to understand what the circuit is doing. It’s really an integral part of the switching converter and should be drawn directly connected to the IC output.

It’s also a good idea to include additional parameters like voltage ratings for capacitors, especially if different parts of the circuit are operating at different voltages. You absolutely need to include turns ratios for transformers – this is the difference between passing 30 volts or passing 300 volts.

You probably also want to clearly indicate which parts of the circuit are inputs and which are outputs. This is particularly confusing in your schematic. Are you going to have some sort of terminal block / pin header / connector for this board? There are a lot of labels that don’t seem to go anywhere, which makes the schematic harder to read.

Finally, there are a lot of components in your schematic that… either don’t seem to be doing anything, or that probably shouldn’t be in there. The schematic is difficult to follow, so it’s hard for me to be more specific, but I suspect that a lot of the diodes and transistors don’t really belong in this design. My suggestion is usually to try to design the simplest circuit you can to do a particular task. If you don’t understand how some complicated part of a reference design works, leave it out until you understand what it’s doing and why it’s there.


#3

It is unclear, from the schematic what exactly your circuit is supposed to do and what the intended inputs and outputs are. Can you clarfiy what you are trying to accomplish?

It seems you are trying to implement the UC3842 Reference Schematic for a offline SMPS based on a current mode PWM controller, the datasheet for this part is 60 pages and has many details that are impossible to review without more thorough documentation on your side. At the bare minimum many of the capacitor and resistor values can be chosen only when certain details like output voltage, input voltage, and expected load, is known.

Additionally, as @ohazi pointed out, all of your components are lacking ratings and specs. Most, if not all of the difficult engineering (“real engineering”) will be in specifying the exact rating and specification of all the components, like transformers, chokes, and diodes. So please elaborate, otherwise the schematic is quite incomplete at this juncture. An incomplete copy of the reference schematic only serves to diminish your learning of electronics (in my opinion)

Switching Supplies have many caveats, and the reference implementation is just a reference, it is incomplete for most purposes and needs to be adapted.

AC/DC power supplies are extremely inexpensive and use highly optimized, safe, and verified designs. Academic curiosity has it’s place, but from a purely engineering perspective you are reinventing the wheel, somewhat, but you are starting not from a normal wheel, but a formula-1 wheel and trying to reverse engineer.

I would highly suggest you refocus on implementing a DC/DC SMPS supply using a different current mode controller. It shares many of the same design challenges, but you do not have to worry about Mains isolation and is much safer. If you have a risky side, and simply want to work with AC voltages, I would highly suggest instead a simpler, traditional transformer and rectifier based ac supply (like the one @ohazi presents)

If you do choose to implement this circuit. Please use an AC power supply and do not plug into the wall.

This component is a common mode choke, and not a transformer, it is connected correctly for its purpose (to filter high frequency transients from the supply line). The rectifier is connected per the reference design, which is not traditional configuration because this is an offline (online? whatever, riding on mains) supply arrangement.

https://www.murata.com/en-us/products/emiconfun/emc/2014/10/28/en-20141028-p1

image


#4

Ah, you’re right. When I see “AC” --> “other stuff” --> “Bridge rectifier,” my usual heuristic is to expect a transformer somewhere in the other stuff.

The TI schematic is a little bit clearer (note the clearly indicated inputs and outputs, and the obvious symbol differences between mains-coupled ground and output ground), but it still isn’t super easy to follow. Partly because it’s a fairly complicated design.

Some semiconductor manufacturers have a tendency to publish some pretty ridiculous rube-goldberg-like reference designs. TI seems to do this a lot. I guess they’re trying to showcase the flexibility of the chip they’re selling, but this is often a trap for beginners. Most of the time you don’t want anything close to what the kitchen-sink reference design is doing. Start with the simplest design and go from there.

When the design is big and complicated and you’re not really sure what all the surrounding components are doing, it’s easy to get overloaded and just start dropping parts in simply because “the reference design said so.” This is where it becomes easy to make a mistake, because you don’t understand how that part of the circuit works, so anything that might be slightly off won’t jump out at you.

If you start with a minimal design, and then only add complexities that you understand, and only when your particular application needs them, it’ll be easier for you to see at a glance that something is off, because you have a clear understanding of how everything is supposed to work, and what everything is supposed to be doing.


#5

For me your schematic is completelly unreadable. It is hard to see where each wire is going to just to imagine what signal is where.
You can go farther with that idea and use no wires at all - just separate elements with net names :slight_smile:


#6

Thanks alot for all the pro tips. From both yourself and @ohazi I have learnt a bunch of new things. I apologize for not being very descriptive before but I’ll try to explain myself better. Below is the reference circuit that Im using. It is 100-230VAC input with 12VDC 10A output smps.

A friend of mine came to me with it and told me that he needs a custom pcb design for this power supply (dunno why he assumed being a fresh engineering graduate I would know all this :confused:) It is intended to drive a 4amp load while charging a battery. He said he needs a specific shape with reduced size which, I assume, could probably be achieved by replacing a lot of THT with SMT components (the one he has uses all THT).

The reference fig has specs mentioned and I should do so too on my schematic just like you said. I have only drawn upto the primary side of the transformer until now. Its gonna have connectors for input and outputs. I tried to specify some of the circitory and this is how it looks now. I didn’t change the references because I dont know how it would effect the annotation later. here goes

The choke and transformer I have to figure out still. I searched through the internet for the exact calculations involved in designing the smps but could not find them I would highly appreciate if you can provide them as they may help me with this project. Thanks again


#7

May be this will be usefull:
http://schmidt-walter-schaltnetzteile.de/smps_e/smps_e.html

I am working with low voltages and currents (12V, 300mA) so such project I would clasify as difficult.
I know only that in fly-back trafo you should be care to have each winding fillling your carcass fully (from its edge to edge) to minimise the losses and unwonted emissions. I know that windings are sometimes made of 2 or 3 wires in pararell not because of current but to reach this goal.
And 2n2 at your schematic is very important for EMC.