I was looking at this power switch from pololu:
I'm curious if anyone knows where I can find the schematic for this? I'm interested in the overall design aspect behind this switch.
Thanks in advance!
The Pololu Pushbutton Power Switch is a patented design
If it is really patented it should be in the patent databases ...
search search search => http://bit.ly/keiNng
Look to rohm. I'm sure someone makes the "latching circuit" for soft-power applications.
Isn't the "latching circuit" a microcontroller (running some firmware to read the button etc)?
florinc:
Isn't the "latching circuit" a microcontroller (running some firmware to read the button etc)?
Yes, the MC program has to be in the 'logic' loop if it's desirable to have a programmable auto-power off function, which is the whole point of having a solid-state power switch, otherwise a simple press on / press off manual power switch would be all that's required.
Here is the same concept that uses a latching relay rather then a FET 'power switch':
On pressing the manual momentary power switch, the MC starts and first outputs a digital high to 'set' the latching relay that 'seals' the power switch contacts. When the MC program wishes to power down it just outputs a low and the relay releases it's contact and the whole system powers off.
The advantage to using a latching relay with capacitance coupling (rather then a standard relay) is that once the relay is 'set' and the cap fully charged there is no continuous power being consumed by the relay coil, so this is as 'energy efficient' as the solid-state switch, about the same size and the relay can be had for $1 on e-bay.
Lefty
Or, you could simply replace that microcontroller with a NC pushbutton "OFF" switch. That is how we did it back in the dark ages before whizzy micro-circuits.
I guess the idea is to avoid having 10A going through the mechanical switch.
OK, I understand now what you meant.
Hello.
KE7GKP:
There is no effective way of "patenting" an electronic circuit. As soon as you substitute one component or change a resistor value, it is a different circuit.
I don't think this is true. For example, you can change the resistor value in a low-pass filter and the circuit is still functionally a low-pass filter. If you want to make a different circuit, I think you need to significantly change the way the circuit functions and/or accomplish something the original design does not. I'm not an expert on patents, though, so I could be wrong. All I know is I'm skeptical of your claim that patenting circuits is entirely ineffective.
If you look closely at the photos, you will likely see the part numbers of the ICs, and the circuit is probably "borrowed" from one of the reference designs or application notes for the ICs.
robtillaart posted a link to the circuit schematic right above your post, which I expect will be more insightful than looking closely at product photos. The circuit is not from a reference design or application note (why is that your first assumption?), and it does offer certain advantages over alternate approaches to push-on/push-off power circuits.
- Ben
Quote from: KE7GKP on June 21, 2011, 06:55:31 PM
There is no effective way of "patenting" an electronic circuit. As soon as you substitute one component or change a resistor value, it is a different circuit.I don't think this is true. For example, you can change the resistor value in a low-pass filter and the circuit is still functionally a low-pass filter. If you want to make a different circuit, I think you need to significantly change the way the circuit functions and/or accomplish something the original design does not. I'm not an expert on patents, though, so I could be wrong. All I know is I'm skeptical of your claim that patenting circuits is entirely ineffective.
I own several patents myself so although not a legal expert, no noob either but all disclaimers apply 
First legal rules wrt patents differ per country and are even subject to change, and are not respected in all countries either ...
- Electric circuits can be patented, but they should be really new AND solve a real problem (there are definitely some degrees of freedom wrt what a problem is)
- Using a known electric circuit (low pass filter) as a new solution for an exisiting problem can be patented if it is not a trivial solution (there has to be some inventiveness in it)
Bens is right that a non significant change in a schematic still falls under the patents protection. I think all patents I've seen claim that trivial changes to the proposed solution are part of the claims. Sometimes the claims go so far that they claim the solution to be new and that the implementation can be either software and/or hardware implemented. (a lowpass filter can be made in many ways!)
Wrt Patents:
Patents normally consists of 4 parts: (1) a problem description, (2) the state of art solutions/prior technology, (3) a proposed new solution and (4) claims. For the legal people the claims are the farmost important as that is what is claimed, second the state of art is important as that decides if the proposed solution is really new or not. If you failed to find/describe all earlier solutions it might be that your solution is not new at all (this happens very often!)
Getting a patent is not easy but after "the bright spark" and the necessary work (literature research/editing/rewriting etc) it can be done. The real difficult part comes then, to enforce your patent. If someone else uses your patent you get in a legal fight and you have to proof someone is using your patent. If the other party can show it make things that way before your patent was filed => say bye to your patent as the work of the other party is prior technology. These legal fights are very costly and often difficult to win. Some would say only the lawyers win ...
Proofing someone else uses your design in electronics is quite difficult as you often need to be able to reverse engineer chips these days. That makes it difficult to enforce your rights, so often it is easier (and economical more interesting) to keep the design secret and use your money to be the first in market and fill up all the sales channels. And if you patent something, don't patent a specific solution, try to patent a generic (implementation independant) one
That said, for real patent advice contact a patent attorney.
my 2 cents,
Rob