At what point does combining integrated circuits yield something patentable?

Question

I am considering at which point an electronic circuit-based product becomes patentable. If I were to simply design a new electronic circuit by implementing a single integrated circuit (IC) from a manufacturer, I would think that this cannot be patentable. I would also presume that the IC manufacturer has some kind of patent on their own IC. In this case, I would suppose that there is no patentable product and therefore you can sell this product without needing to worry about patent law.

But, what about when you combine two ICs that work together to do something than one of the ICs cannot do on its own? At this point, assuming it is novel, useful, nonobvious, and not previously patented, is this patentable? Do I need to begin searching to see if someone else has patented that combination of two ICs before I begin productising an invention? Fundamentally, I would like to understand if combining ICs in a certain way leads to something that is patentable, or is something more complex required?

Answer 1

Electronic designs can be patented provided they meet the criteria for a patentable idea. But patenting a design using specific ICs is probably not a good idea as all someone has to do is redesign the same function using different ICs and they have worked around your patent.

The best approach is to patent your circuit function using "functional blocks" that are more generic so that the overall circuit's function, which is the novel idea, it patented and not the specific implementation.

Patents can be tricky to navigate. You might do well to consult a patent attorney who can advise you how best to protect your intellectual property.

Answer 2

I did a short course once from our in-house patent attorney, and he said that, under Australian law, patent was for solving a problem. So:

1. I combine two circuits for the hell of it. Not patentable.

2. I want to make automobiles go faster, which I achieve by combining the same two circuits. Patentable (subject to novelty and all that).

3. I want to test for a particular cancer, which I achieve by combining the same two circuits. Patentable (solution to a totally different problem).

4. I combine two circuits in the hope that someone will find an application, and I can extort royalties. Not patentable. (There was a case in Australia where the judge said that one party was inventing patents instead of patenting inventions).

Answer 3

The criteria for patentability are usefulness (not necessarily optimized or better than previous solutions), novel (not been done or described in a publication anywhere anytime in the past), and not obvious (to someone of ordinary skill in the art who knows about everything that has ever been done or published in the field ever in any language previously).

The number of off-the-shelf components it takes to make it is not directly relevant. If you are concerned that your new circuit might be patented by someone else you can have a freedom-to-operate search done

Answer 4

QuadmasterXLII made a comment that I think deserves to be promoted to an answer:

The heuristic "You can patent literally anything if you have money, but you can only enforce on people with less money than you" will get the practical answer 99 times out of 100

This is USA-centric, but so is most of the world these days.

The US patent office will accept any patent that uses the proper forms and is paid for.

The hard part starts when somebody tries to enforce a patent. This involves a lot of lawyers on both sides and quickly becomes very expensive for both parties. The deciding factor is often who runs out of money first.

Now everybody knows this, so if a big company threatens to sue a small company, the small company will usually fold immediately. This typically involves paying some sort of licence fee for the allegedly infringed patent.

Answer 5

It would be rare that anything as similar as combining a small number of ICs would be patentable.

If the idea is "obvious" to someone skilled in the relevant technical field, given the current state of the art, then is may not be patented, even if it is novel and useful.

Unless you were doing something that conventional wisdom inaccurately led skilled practitioner's to believe couldn't be done, it is highly unlikely that something this straightforward could meet that test.

It also doesn't help that there are multiple elite patent law firms that are monitoring anything related to IC patents very carefully and are likely to proactively contest any new effort to patent something in their field, since IC manufacturers have big $$ at stake in that corner of patent law.