der_wille_zur_macht
Team RC
Haven't used one in the flesh, but I did play with TI's design tools for it back when I was working on the original NCP design I started this thread for. At that time I did not use the LM3404 because it was in an SMT package and didn't offer any benefits (based on my criteria) over the NCP. The higher input voltage sounds nice but in practical applications it's not a huge benefit to the average DIY'er (again, IMHO) because if you figure cost per watt, the typical POTRANS 24v 6.5A supply is much cheaper than any higher voltage DC supply, so the cost savings you get by having fewer drivers is negated by requiring more expensive power supplies. Though arguably an LED array with longer strings and fewer drivers will always be slightly more efficient than one with shorter strings, so you could argue that the longterm cost advantage falls in the favor of the LM.
Also, regarding efficiency, it strikes me that switchers really only have an advantage when Vin and Vout are far apart. Since we can carefully choose and control those parameters, we should be able to get efficiency with the CAT4101 that's close to or the same as what you'd get from most switchers under the same conditions.
Again though, these are all IMHO and according to my own criteria. I'd love it if someone WOULD play with that chip and report back!
Also, regarding efficiency, it strikes me that switchers really only have an advantage when Vin and Vout are far apart. Since we can carefully choose and control those parameters, we should be able to get efficiency with the CAT4101 that's close to or the same as what you'd get from most switchers under the same conditions.
Again though, these are all IMHO and according to my own criteria. I'd love it if someone WOULD play with that chip and report back!