Optimum Senior Designer Scott Nance presents a 45 minute seminar on PCB design for switching power supplies. Originally presented at PCB West 2013 optimumdesign.com
Thank you for this. I also picked up your article from your website, which shows many of the pictures in your presentation. The first part of the video is just background information on switching supplies, so if one is already familiar with the topic, they may wish to skip ahead to the practical advice starting at about time 18:30.
Really good speaker and an informative video. I'd love to have a beer with this dude and discuss Electronic Design, seems like her really knows his stuff..
for the layout example at 28:10 i would not have extended the vout copper pour under the inductor to minimize capacitive coupling from the switch node to the output, one could possibly even have a ground sliver under the inductor to further reduce the coupling, also be aware of inductor polarity for multilayer inductors, you would want the start of the winding (inside) towards the switch node
Great video! I have 1 question on the slide (at time 31:30) the "thermal pad" for the fet is connected to SW. How do you provide cooling for that fet and keep the SW node small?
@@sain8827 Hi, I'm still alive ! :) The thermal pad in 99.9% of IC is connected to GND (or the lowest potential of the IC) and usually they specify that it should not be left floating. In this example with the IC in question is a fet, the thermal pad is also source connection and on Q1 you can see that it is connected to a big power/ground plane with plenty of thermal vias to spread the heat. I think my questions still stands, and Q2 has only a small area to cool, via its source connection.
Can you suggest a software to me for doing mix signal analysis. I have to design and POE++ device with Power Source side and PD controller side to power up high power LEDs like 70W. Client requires detailed simulation prior to making the hardware. I have seen LTSpice but it doesn't support micro-controllers. Candace tools are super expensive. ...what are my low cost options.
At 31:09 gate drive layout is shown. Having two paths from driver common pin to transistor source, will it form a ground loop? Moreover, ground plane is much wider, so current should follow this path, not a dedicated Kelvin trace. Or, maybe, I get something wrong?
9:46 - "Linear regulators are typically 60% efficiency". Ok, that sentence did not really help, as it would be hugely different depending on the ratio of input and output voltages.
Good Presentation but make sure Presentation always in bigger window and presenter in small window. Otherwise we have to pause many times to zoom and check. People will be interseted in Presentation not presenter.
for the people who edit these things: Just display the slides full screen. The info there is relevant, seeing the talking head is NOT. Especially when there are schematics (or anything besides text) you I'd prefer just seeing the slide.
So many traps in this video. Explain how you determined linear regulators are only 60% efficient for any circumstance (Hint, you can't). Also, perhaps research your topic a bit more, as you find that many switching regulators are just linear regulators with an additional controller. Also does not talk about disadvantages of Switch Mode Power Supplies (what SMPS stands for that you happened to leave out). For example, at low load currents, SMPS is nominally worse than linear. Likewise, when input and output voltage are close, linear regulator performs better in effiency, introduces no AC noise, less complex (less supplemental components), and is probably cheaper. Horrible talk on SMPS and linear regulators. Not even sure if your PCB design is good, as im not going to watch it seeing as though you did no research for the beginning of your presentation.
He specifically mentions SMPS. If you bothered to watch the video before criticizing you would see that. The 60% efficient figure is "typical", aka a common example. Not set in stone.
@@mikek5206 read my comment once more. Never said he DIDNT talk about SMPS, he never spoke on the disadvantages. As for the efficiency argument, a linear regulators efficiency scales LINEARLY with application. The fact your sticking up for that ridiculous comment shows you too do not know what you're talking about, no offense. What I mean by that is if I regulate 6v to 3v, that is 50% efficient, half the energy is roughly wasted as heat. This assumes the bare minimum topology, no power save features. Likewise, 12v to 3v is less than 25% efficient. Big difference in efficiency for the same regulator based on application alone. That is why that comment makes 0 sense.
