Bench power supply

The engineer wanted to know what should be the minimum rating of the bench power supply for a Buck application with 5V at the input and 3.3V of output, delivering 1A of load current. I think his question initially arose because he was trying to make a small converter powered off a USB port for driving a peripheral device. He was worried how the current limit of the USB port was going to affect the load current he could draw. 

Note that assuming 90% efficiency, we would get an input power of 3.67W. In that case, Iw would be equal to 3.67/5 = 0.73A. So the actual number is somewhere around 0.7A as I told him. In reality, besides having to account for the less-than-perfect efficiency, we may need some additional margin just to provide the inrush current at startup. So it is likely we may end up using a 1A or 2A bench power supply. But that is not relevant here. The problem here is, we still don t know where the additional current is coming from in steady state Intuitively, not via math ... 

Therefore all the DC being demanded by the inductor must come solely from the bench power supply, that is, 0.67A (Block 2). 

But we must also confirm that the bench power supply is not the problem, if possible by swapping it with a completely different one (brand/rating) lying around the lab. 

Question 9 Is there some interaction with nearby circuitry Yes, you could be picking up fields from nearby circuits, but that shouldn t affect a typical switcher, simply because it produces enough noise and fields of its own. However, it is a good idea to do the reverse-peel here. If I find the converter is on a larger system board, I immediately and carefully first cut off all the traces leading from its output and divert them to my predictable electronic load. I also cut the input traces and divert them to my bench power supply. If the problem is gone, it is an interaction problem. 

Another test that I have found always brings out the inherent weaknesses of the part is the hard dV/dt test. Basically, I simply slam the red banana plug into the already-powered-up DC bench power supply and look for overshoots (voltage or current) in the switcher. There is a fair amount of natural input bounce created by this rather unofficial test, but that can really help aggravate/expose any startup logic issues with the IC. Of course we may later decide to specify a smooth (non-jittery) input AVI At for the IC and just move on. My colleague used to use a mercury switch for the same purpose. That gives almost the same hard input AV At, but without all the bounce. 

Power converter circuits are often the most overlooked aspect of a system. During the engineering phase, power is not a concern. There are plenty of bench power supplies scattered around the laboratory for use in breadboarding. Even in SPICE, the trusty voltage source element provides infinite voltage and infinite current for new circuit designs. 

Your bench power supply is set fairly close to the maximum input voltage rating of your IC. You have just changed the input ceramic capacitor of the converter from 22 0,F to 10 iF, and the supply line still looks very clean (under steady conditions). But the part gets damaged almost every time you connect the red lead coming from your bench power supply directly to your board. Why ... 

---