Post by thread:[OT] US Patent 7138770 -- LED driving circuit (utt

> A recently issued patent (2006) for driving LEDs from AC seems utterly > obvious. Search for "US Patent 7138770" on Google Patents: > > http://www.google.com/patents > > Am I wrong, or am I missing something? Well, they do have them in two strings, not like a single string with the polarity of each successive LED reversed like how most people do it. And they did arrange it in a nice diamond pattern. I was wondering about making a bridge rectifier from LEDs a while back, but I wasn't wondering too much to actually go & try it. Looks like they beat me to it, and gained a patent. The only real difference I see is in Figure 4 the LEDs in the middle run on DC, and so don't flicker. I'm not sure what the resistor in Figure 2 is supposed to do though. I though the idea was to reduce heat, not make it. I think now I'll patent an AC LED (ie bipolar), since it's really hard and non-obvious to put 2 white chips into an LED. No flicker there (ok, less flicker). Tony

Tony Smith wrote: > I was wondering about making a bridge rectifier from LEDs a while back, > but > I wasn't wondering too much to actually go & try it. Looks like they beat > me to it, and gained a patent. What have they actually (tangibly) gained? I've since looked at other patents, some in our market niche, and some of them are really really trivial. So trivial in fact, that we've been doing the stuff in the patents long before they were filed. I wonder if they do it to secure VC. Vitaliy

> > I was wondering about making a bridge rectifier from LEDs a while back, > > but > > I wasn't wondering too much to actually go & try it. Looks like they beat {Quote hidden}> > me to it, and gained a patent. > > What have they actually (tangibly) gained? > > I've since looked at other patents, some in our market niche, and some of > them are really really trivial. So trivial in fact, that we've been doing > the stuff in the patents long before they were filed. > > I wonder if they do it to secure VC. Probably. They've basically managed to light up LEDs in a way that no one else could be bothered to. Using LEDs to make a rectifier is pretty obvious, they are diodes after all, but it's also obvious that it makes for a crap rectifier. If it was a product that required a bunch of LEDs and a small amount of DC I suppose you could use it (that's what I was wondering). The minimum I've seen is a LED & a resistor, straight off 240v. Sure it's half wave, but so is theirs, unless you consider the option with the LEDs running on the DC produced. Even then they've got additional resistors and power diodes as well, so it's less efficient. Beats me. And people wonder why I'm not impressed they're got a patent. At least they won't have to spend money defending it, it's so useless no-one would do it. Tony

> I've since looked at other patents, some in our market niche, and some of > them are really really trivial. So trivial in fact, that we've been doing > the stuff in the patents long before they were filed. I'm agreat unbeliever in patents in most cases. I filed a provisional patent recently, as it seemed to have very good potential value to do so. The contents will almost certainly never see the light of day. This gives me a 12 or 15 month (more $) interational prioority period for ANYTHING in it that I can soundly base a full patent on. As I used a rather broad but also high detail scatter gun to create the document the material which MIGHT be based on it is large. The object was to ensure that ideas in a "report" did not suffer from "parallel invention". I'm very happy for the ideas to stand on their own feet - but don't ant them standing on someone else's feet without merit :-). What does it cost for a US provisional filing? - or is such unnecessary? eg in the US just piblishing something gives you 1 year priority. In NZ it would make the material public domain! Cost of a PP here is small - main cost is in the effort involved, but it needs meet no formal standard and is never examined. Russell

On Tue, Mar 3, 2009 at 7:10 AM, Benjamin Grant <spam_OUTbenjamin.grantTakeThisOuTduke.edu>wrote: > I don't think you're wrong, I guess patents just have a low threshold for > triviality. It doesn't appear to have any content of interest in it. As an > aside - google has made viewing patents so much nicer than was previously > available. > -

yeah agreed Chris. I know also of really broad patents written by academics to secure years of research in their labs -- particularly in poorly explored fields. Such patents come out before the author really has a concrete design and are used to prevent others from encroaching on ongoing research. Really not currently an optimal system On Tue, Mar 3, 2009 at 6:12 PM, Chris McSweeny <.....cpmcsweenyKILLspam@spam@gmail.com> wrote: > On Tue, Mar 3, 2009 at 7:10 AM, Benjamin Grant <benjamin.grantKILLspamduke.edu > >wrote: > > > I don't think you're wrong, I guess patents just have a low threshold for > > triviality. It doesn't appear to have any content of interest in it. As > an > > aside - google has made viewing patents so much nicer than was previously > > available. > > --

>To be blunt, the US patent system appears to be pretty much broken. I can >think of one pretty well known example where no consideration of prior art >was taken. It seems the principle is that they'll grant a patent for pretty >much anything, and you then fight it out in the courts - which results in >the case I'm thinking of where if you can get a dodgy patent and are >prepared to defend it with a big enough legal budget, nobody will risk >fighting you. I worked for a patent attorney for several years. Basically, he said that: 1) The USPTO is in the mode of granting patents, and letting the courts decide. 2) Patent fees generate net revenue for the government. Which may influence #1. -- --- Chris Smolinski Black Cat Systems http://www.blackcatsystems.com