Very good!
Very approachable yet full of deep technical info. At least to a beginner like me. I really enjoy your writing style, it is a rare skill.
May I offer a couple of comments? Keep in mind, I am a beginner so these are more editorial comments. I am only just able to understand the article itself.
The article is written for an existing practitioner, clearly. I mean, someone who probably already knows the practical model of a capacitor. So, eg, the model of a capacitor isn't so much educational (and anyway, easily and trivially found anywhere and your article isn't about such fundamentals that are easily learned) as it is intro for the rest of the discussion. So it is valuable to describe. OTOH, spelling out Surface Mount Technology before presenting the abbreviation SMT hits the wrong target audience. I mean, you don't describe what a via or micro-via is. So you don't need to expand SMT. It just slows down the flow IMHO. Same for ESR/ESL and other terms. You shouldn't expand those terms for the same reason you don't bother to expand AC, DC, PCB, etc.
You write that we must understand the construction of a capacitor, but you don't describe its construction. And indeed I don't think such understanding was necessary.
The bit about others not being able to tell you why physical proximity matters is really self-serving, even crass. One should not preface some deep explanation with a statement of how great one is for having such knowledge, and how pedestrian most others, who apparently have mere claims to greatness, are. A more simple, "it is commonly known that ..." and "let's describe why that is the case", would serve you much better.
After going to great lengths to state we must understand why placement has an effect, you do so quite well. In fact I think that is the heart of the article. But I find it ironic that you distill it down to some rules without discussing the reasons for those rules! Why is a cap ineffective further than 1/4 wavelength away? What is the basis of the 1/40th rule?
"dialectic constant"? Is that a spell check error, or is there such a thing that I am just not aware of. I'm nearly certain you meant dielectric.
I wish at the end there was some other way to introduce HyperLynx Power Integrity from Mentor Graphics. I bet HyperLynx Power Integrity from Mentor Graphics is a really great tool (sounds like it), but the way it is presented makes me think I just got burned by reading an ad for HyperLynx Power Integrity from Mentor Graphics. Perhaps if you left off the "from Mentor Graphics", and perhaps if you named another tool (since you say tools, plural).
I wonder are there any design tools for decoupling networks? As opposed to HyperLynx Power Integrity from Mentor Graphics which sounds like it is a design verification tool. Based on this article it just seems to me (as a 15+ year software veteran) this is an easy problem to completely automate away. (If you already know the power supply noise, if not you'd have to measure it but then you plug that into the tool.)
Home
Help
Search
Login
Register
