Patents 101 for DIY Audio Enthusiasts

Main Index

Articles Index

Greetings.  My name is Ian Millar.  I'm a semi-retired 'Registered Patent Attorney' working from home (occasionally) in the northern Sydney suburb of Cheltenham - just down the road from where our friend Mr Elliott resides.  My CV is now offline, it used to be linked from here, but not as a blatant attempt at self-promotion - I really don't want any new work - it's tedious.  I also have a DIY audio blog, where a few of the ESP projects are quietly hiding.

I have been a long-time reader of Rod's audio pages and have built a number of the ESP projects over the years for my own use and this has provided enormous personal satisfaction.  I have learned a great deal from Rod and am eternally grateful for his generous assistance when I've occasionally bitten off a little more that I could chew so to speak (like trying to put too many projects that weren't designed to go together in the same box with multiple power supplies!) .

Rod has asked me to put down a few words about patents and how they might (or might not) be relevant to DIY audio enthusiasts and in particular readers considering or having built any of the excellent projects published on the ESP site.  There are many other places to read detailed analyses of the patent system generally and up-to-date case law is available by subscription to (quite boring) specialist journals.  Instead, this will be pretty basic and I will try to keep it as relevant to DIY audio enthusiasts as I can.  I'll break it up into topics of relevance, some of which may seem a little generic, but necessary for a basic understanding.  I'm probably pretty good at the limited scope of what I do, but not very good at presenting 'courses' so don't expect to pass any exams by reading this.

What I hope to achieve here then is a concise overview as seen from my own perspective.  Hopefully after reading here for just a few minutes a lay person can go away with a basic understanding of something that they never really considered before.

By the nature and context of this article it really must be incomplete.  If questions arise or glaring omissions are pointed out, I can always update the text later.  So I'll type this 'off the cuff' without resorting to anything but my acquired knowledge with a view to keeping it as non-technical as possible.

The disclaimer bit:&nsp; OK - I accept no responsibility for omissions or errors.  Please make no decisions on the basis of what you read here, but make your own enquiries.  I'm sure Rod will add his own disclaimer too.

ESP has published this material solely as basic information to describe patents, and the processes involved.  No part of this material is to be taken as legal advice.  Patent law where you live may be slightly different from that described here.  Information is provided in good faith, although there may be errors, omissions, local variations or other circumstances that may affect the accuracy of the material for you.  No decision should be made without consulting a Patent Attorney.

Well it's not a Trade Mark (which is usually an indication of a brand) and it's not Copyright.  It's not a Registered Design (covering the appearance of an article) either.

A patent is a limited-term monopoly awarded by the government as an incentive to innovation and the advancement of technology.  At the end of the term of a patent the invention becomes a gift to the public.  Like most things in a commercial world the patent system is driven by basic human greed and self-interest.  Some disagree with the concept of monopolies while others (at least try to) profit from it.

A patent is a form of Intellectual Property.  It is an asset which can be sold like any other asset or licensed to others.  A patent can depreciate like a rusty car as it approaches the end of its usefulness.

A patent in Australia is a statutory monopoly for any useful and inventive 'manner of new manufacture' which might range from electrical and mechanical devices and methods of manufacture and apparatus use, to pharmaceuticals and DNA sequences.

Patents are generally not available for mere working directions, business schemes, rules for playing games, or instructions for operating known machinery.  In my opinion software is really nothing but a list of instructions for operating known computers, but in recent years applications have been passed for software-related inventions with a 'leave it for the Courts to decide' approach.

A patent application for (say) a method of under-biasing a power amplifier output valve by turning an existing trimmer pot slightly to the right then slightly more to the left ought to be refused.  That may sound stupid (it was supposed to), but many years ago Rolls Royce was refused a patent application for a method of flying a jet aeroplane more quietly over built-up areas simply by throttling down a bit!  Brilliant eh?

The layout of copper tracks on a PCB or silicon wafer is generally not patentable subject matter either.  There could be exceptions in unusual circumstances where factors beyond the mere track layout were involved.  Say if the board had to be non-planar or something special, but patent protection in such cases would be unlikely to be limited to a specific track layout.  In Australia there is separate (non-patent) legislation protecting circuit layouts.

Human beings are not patentable, although I have seen an Australian patent for a radioactive glowing 'pig'. 

Patents are unlike Copyright in which it must be proved that actual copying has taken place for infringement to be found.  Technically you cannot infringe copyright by independently creating a work that happens to be similar to one in which copyright happens to subsist.

A scope of monopoly afforded by a patent is defined by a list of numbered paragraphs called 'Claims' which define an alleged invention.  I say 'alleged' because claims can fall over in Court for being shown up as defining non-inventions.  Claims are long, oddly punctuated and formatted sentences at the end of a specification which includes drawings and a description of a 'preferred embodiment' of the invention.  Actually a good claim (for the patentee) is short, but they are difficult to achieve.  I once drafted a two-line independent claim and saw it through to grant.  I don't know whether that ever took a tumble later on though.  Most are more like the example that follows.  Some 'smarty pants' Patent Attorneys invent words and apply the word 'said' instead of 'the' excessively in an attempt to impress the easily impressed.  One particular disgrace is the addition of 'ingly' to words such as 'seal' or 'press'.  For example "… wherein said plug sealingly engages with said aperture" instead of simply "… wherein the plug seals the aperture".

An 'independent' patent claim for (say) an isobaric loudspeaker (when that was a new thing) might have read something like this:

Actually the last two limitations are probably unnecessary, but I included them anyway.  Usually there are further claims which depend from a claim like the above one which serve a back-up role.  A 'dependant' claim for the isobaric speaker might have read something like this:

2. The loudspeaker of Claim 1, wherein the first and second transducers are substantially coaxial and both face in the same direction.

Dependent claims have within their scope every limitation of the claim(s) from which they depend and sit quietly just in case their precedent claim(s) is/are found to be invalid by a Court.

At the time of writing (May 2010) there were several recent US in-force patents and patent applications for isobaric speaker systems lodged with the USPTO (US Patents & Trademarks Office), the most recent being an application filed in 2006.  Based on a (very) quick read of these, it seems highly unlikely that any would stand up if taken to court.  The vast majority of all claims are based on things that loudspeaker box manufacturers would seem to do regularly in the normal course of building an enclosure.

There are many patents in the audio industry that appear to be intended to do nothing more than make it appear to the uninitiated that the manufacturer is 'clever', and can do things that others cannot.  Many of these patents might have managed to get past the examiners for whatever reason, but they would fall over in an instant if challenged.

The claims are directed at a hypothetical person skilled in the art of loudspeaker construction and need not spell out everything.  For example Claim 1 above omits any mention of the apertures across which the transducers would typically be mounted and whether the transducers are connected in series or parallel, but these details are not crucial.  The claim is intended to define the invention and not the minutia of construction.  The claims include no explanation of how the loudspeaker really works or what its benefits might be.  The independent claim(s) just define of what is (hopefully) 'covered'.

Assuming claim validity, unauthorised exploitation of a loudspeaker as defined by the any one of the independent claims during the term of the patent will infringe.  There is a principle called 'exhaustion of rights' by which it is OK to sell legitimate patented goods which are second-hand, because the patentee made his profit from those already.  This is subject to certain cross-border provisions which come into play in circumstances where a patent for the same invention is owned by or licensed to another party in another country.

An independent claim can be broken down into 'essential integers' or more simply 'features' and it is the combination of these features in which the monopoly resides.  For example an independent claim could be broken down as  ...

1. An invention that has:
      Feature A;
      Feature B;
      Feature C; and
      Feature D;
      wherein Features A, B, C and/or D can or do mutually interact in a new and inventive manner.

... and I'll include a dependant claim for reference later ...

2. The invention of Claim 1, further having Feature E which is attached to Feature A in some special way.

Features A to D might all be off-the-shelf items, but if the combination was new and inventive at the date of filing the patent application, the owner is entitled to monopolise it.

There is no world-wide patent for anything.  Whilst there is an International Patent Application process (called a Patent Cooperation Treaty application) , it is only a means for filing a bundle of national applications at one place and time.  The end result is that national patents are granted in some or all of the places designated when the PCT Application was filed.

Each patent is enforceable only under the law of the country in which it is ultimately granted.  An Australian patent cannot be infringed by the sale in the USA of a US or Chinese-manufactured knock-off.  The manufacture of a knock-off in China is not of itself an infringement of the Australian patent either.  The Australian patent is infringed when such goods arrive in Australia (there are temporarily visiting vessel exclusions to this).  The importer infringes.  To stop the Chinese factory a PRC patent application ought to have been filed.

20 years is now the maximum term of a standard patent in most if not all countries.  Some countries allow an extension of term for pharmaceutical patents because it takes some years for government agencies to approve new medicines before they can be sold - rendering the first years of the monopoly useless.

Renewal fees must be paid at regular intervals throughout the term of a patent.  If these are not paid the patent lapses.  After expiration everything that was disclosed in the document becomes free for the public to exploit (subject to possible restoration of patents that lapsed because a renewal fee was unpaid accidentally).

Patent Offices (that's where you file patent applications) are not interested in your potential to infringe earlier patents.  Their primary job is to filter out public burdens in the form of applications for known inventions and allow only technically enforceable patents for new inventions.

Standard patent applications are examined for 'novelty' (newness) and 'inventive step' (non-obviousness) in most jurisdictions.  There are lesser patents called 'Innovation Patents', 'Utility Models' or 'Short-term Patents' and the like in some countries in which a tick-a-box approval process fast-tracks a 'grant'.  Post-grant examination is necessary in some countries before rights in these can be enforced.

For standard patent applications searches are performed by examiners (public servants at the Patent Office) who are usually graduate engineers, scientists and people studying to become patent attorneys (for some reason).  Examiners scrutinise applications for possible rejection.  Some are heavy-handed, but most are reasonable.  Applicants can't select a particular Examiner.

Applicants usually start by claiming broadly - say by claiming just the combination of Features A, B and C in an independent claim, knowing that Feature D is really the crux of the invention.  They temporarily place Feature D in a dependent claim (say Claim 2) with high hopes of it slipping past an examiner.  The Examiner searches for and finds older documents (usually patent documents) and will reject the broad independent claim if he considers it to be 'anticipated' by any one of them.  The cited document of relevance could be a hundred years old or quite recent.  It could even be owned by the same applicant.  The applicant is then allowed to amend the text of the independent claim to include Feature D (and delete Claim 2 as you can't have two claims of the same scope).  Feature D cannot be plucked from thin air mind you.  It must be a feature as disclosed in the specification as filed.  New subject matter cannot be added after filing unless a further application is filed.

Where an Examiner cannot find a direct knock-out, but a document that anticipates A+B+C, he might reject the claim anyway on the basis that the addition of Feature D seems obvious.  Usually such rejections can be argued successfully either directly with the Examiner or via a process of appeal.

So lets say that by convincing argument and the benefit of any doubt being decided in the applicant's favour the case is accepted for grant with feature D included in Claim 1 (as shown in the above example).  Feature D is now a limitation of the scope of monopoly afforded by the patent.  The patent is therefore of lesser commercial value than the applicant had initially hoped.

Quick Summary:

• Apply for an old combination of A+B+C and fail examination.
• Add Feature D to the independent claim and the combination of A+B+C+D becomes your limited monopoly.
• A third party can take Features A+B+C, or A+C+D or B+C+D and avoid infringement.
• A third party must take all of Features A+B+C+D and their defined mutual interactions to infringe.
• Dependent Claim 2 is just a back-up.  Taking Feature E and none or only some but not all of Features A to D will not infringe.

It should perhaps be emphasised that not all granted patents are valid.  Examiners are not expected to (and don't) have a high level of personal knowledge of the current “state of the art” in every field of invention likely to come across their desks.  Reliance must be made heavily on what prior art they can find in a search when attempting to reject an application in an unfamiliar field.  It stands to reason then that applications are regularly accepted “wrongly” for dubious inventions.  Some countries have “opposition” provisions whereby interested parties can oppose grant on the basis of what they can demonstrate to be old, but still many applications slip through the system resulting in invalid patents being granted.

Patent applications are generally published 18 months after their 'priority date'.  This is either the filing date of the application, or that of an earlier application to which the application is linked.  This means that if you were to search the public records today by subject matter for applications filed in the last 18 months, you generally wouldn’t find them.

A concept that applicants and novice patent owners have trouble understanding is that they can be granted a patent for their own invention and infringe an existing patent at the same time.  This arises where there is an earlier in-force patent (less than 20 years old) cited by the Examiner or otherwise overlooked.

Lets say the earlier patent (I'll call it 'the reference') was just 10 years old and remains in-force.  Remember if the Patent Office cited the reference during examination, it didn't care about its renewal status or what it claimed.  They merely considered its disclosure as though it were a published journal article to ensure that our claim included something extra.

Lets also say that the reference has a Claim to the combination of Features A+B+C.  Our patentee was entitled to a patent to A+B+C+D because feature D was not disclosed by the reference but he cannot exploit his own invention in Australia for the next 10 years without answering to the owner of the reference.  Adding feature D does not get our applicant off the hook.  He must take A+B+C to exploit his invention.

This common situation can be resolved by cross-licensing.  I.e. our patentee pays a royalty to the owner or licensee of the first patent and takes it from there.  The next inventor will come along with Feature F sooner or later too, so filing a patent application is like stepping on a ladder and slowly climbing up as the topmost patentee gets off (when his patent expires) and someone else steps onto the bottom rung with the latest addition to the original invention.  A truly brilliant invention is a rare thing indeed.  These tend to avoid the ladder game.  In 22 years in the profession I have seen only a handful of these.

Australia has a patent novelty standard known as 'Absolute Novelty'.  Subject to certain 12-month filing grace provisions, this means that the validity of a patent claim can be contested on the basis of 'prior art' in the form of printed public disclosure or use by anyone (including the patentee) anywhere in the world before the 'priority date' of the claim.  The priority date is either the Australian filing date or a foreign filing date for the same invention filed by the applicant up to 12 months earlier and from which 'Convention priority' is claimed under an international convention.  Other jurisdictions have a standard known as 'Relative Novelty' where 'use' outside the region is ignored and very few if any countries still apply a 'Local Novelty' test.

Prior secret use of an invention by parties other than the patentee is irrelevant to patent validity.  But where the defendant used (secretly or otherwise) an invention immediately prior to the patent filing date, this can be a valid ground of non-infringement.  "Why should we stop doing what we've been doing all along just because this guy came along with a patent?" would seem a valid rhetorical question.

Patentable Inventions are supposed to involve some (at least small) level of non-obviousness and patents can fall over in Court for want of it.  Novelty is one thing, but if Feature D is an obvious addition, then the patent claim ought to fall.  Some Patent Offices (USA and the European Patent Office in particular) are becoming very tough with this requirement.  The Australian Examination threshold is somewhat lower (but under review) and the current practice is not to uphold obviousness rejections where there is any element of doubt.  What might be obvious to a genius with the benefit of perfect hindsight might not be obvious to the hypothetical skilled, albeit unimaginative, addressee of the patent.

This hypothetical person is taken to be an ordinary skilled artisan in the field of the invention (say a loudspeaker manufacturer in the above example) and an obviousness assessment is to be taken by him in the light of common general knowledge of his peers before the patent application was filed.  Patent Examiners are not in a position to establish what is commonly known in the field of every application that they examine and there are insufficient public resources at their disposal to gather the necessary number of expert witness declarations to establish it.  I underline 'unimaginative' as it's a good knock-out for expert witness credibility.

An 'expert witness' is often called upon in patent matters to support a case of obviousness against a patentee, but the expert must have no imagination by which he might be capable of invention himself.  If you are capable of invention, then your opinion as to obviousness must be tainted by it.  A quick search of patent records can sometimes reveal that the expert witness giving evidence against you was once nominated as an inventor on a patent application.  "Hey that guy was an inventor for Patent 6480100 in 1979.  He must be imaginative".  Goodbye expert evidence!  .

I've lost count of the number of times over the years that I've set people straight on a strange misconception that permeates society for some reason.  It's like the many myths that abound within the audiophile fraternity - like the one that says that left and right interconnect cables must be the same length or the perceived sound will be delayed at one side, or that cryogenically treated RCA sockets sound better.  It's the "Don't they just have to change it by 10 percent?" myth.  Sometimes it's 20 percent or some other arbitrary percentage.

There is no provision in the legislation or common law for altering an invention by any 'percentage' to avoid patent infringement.  It's ludicrous.  Percentage of what anyway?

If you take every essential feature that is claimed in any independent claim of the patent (Features A+B+C+D in the above example) with the defined interrelationship (and E, F and G if you like), you generally infringe the claim.  There are provisos relating to non-essential features and 'mechanical equivalents' of these, but I won't go into them here because in my opinion non-essential features have no place in an independent claim anyway.

A patent in Australia affords the patentee the exclusive right to exploit the invention and to authorise others to do so during the patent term.  Unauthorised supply of the whole patented product is of course an infringement.  There are also 'contributory infringement' provisions.  For example the unauthorised supply of a key component (say Feature B) of a patented product will infringe if that component has only one reasonable use (being an infringing use).  Even the supply of a non-key component can infringe if it can be proved that the supplier had 'reason to believe' that the component would be put to an infringing use.  There are also inducement provisions whereby any instructions by the supplier to use a product in an infringing way will constitute infringement by the supplier.

Infringement can be by end-use and by supply.

Each country has its own provisions which are somewhat similar in most regards to these.

The defendant in an infringement action usually files a counterclaim that one or more of the Claims is invalid say on the basis of knock-out prior art that has come to light.  The whole patent doesn't fall just because one claim is found to be invalidated however.  It could be that only 2 of 10 claims are infringed, and these two claims are shown to be invalid.

In Australia, apart from injunctions to cease trading in the goods and the right of seizure, a successful patentee/ litigant has the choice of an award of 'Damages' or an 'Account of Profits'.  In the USA there are triple damages available where blatant infringement is shown to be wilful.

These awards are generally calculated on the basis of infringing activity which occurred after the date of publication of the patent application and awards are calculated retrospectively to that date.  For this reason patent applicants aware of a potentially infringing activity in the market place can request immediate publication and expedited handling of their application.  In countries with opposition provisions they must then sit patiently and refrain from making any threats against the offender until a patent is granted.  Infringement proceedings cannot be instituted on the basis of a pending application and any threats would be nothing more than an invitation to oppose grant.

The public is generally deemed to be aware of every patent.  There are 'Innocent Infringement' provisions that relieve the defendant of these remedies if they can show that they were not aware of the patent in suit, but these are quashed if the patented products were marked to indicate a patent number for example and were sold in quantity in Australia before the infringement took place.  That's why it is important for patentees to mark their goods with a patent number.

So What about the ESP PCBs and the detailed directions that Rod so generously provides?

For the sake of illustration, there is an in-force Australian patent for the NTM™ (Neville Thiele Method) crossover, a schematic for which is published on the ESP website.  If ESP were to publish detailed construction information and sell or offer to sell PCBs in Australia for the project, this would make ESP liable for patent infringement under the contributory infringement provisions.

This is a perfect example of Ian's '10% myth' described above.  It doesn't matter if the NTM™ circuit published on the ESP site is the same as that normally used, or is a completely different topology.  The patent describes the method of achieving the result, and is not concerned with the specific mechanism or component values.  It's immaterial if the circuit I describe is different from or identical to that used by licensees, if it achieves the same end result then it infringes the patent.  For example, changing a few resistor values (say 10% of them) has no effect whatsoever - the patent is infringed no matter how many resistor values are changed.

In Australia we are all deemed to know about the patent if the patentees or their licensees sold a number of their own crossovers with patent numbers marked on their packaging.

Indeed, Rod advised me that a challenge from the patent licensees was issued when the NTM schematic was published on the ESP site, but they were appeased when promised that the article was the end of the published information, that no further information would be given (tuning formulae etc.) and specifically that PCBs would never be made available.

With few exceptions for the published ESP projects for which PCBs are supplied, the PCBs are the sole physical item included in the sale.  If there were patents in force with independent claims covering the finished projects, the sale of the PCBs could constitute contributory infringement and/or infringement by inducement.

But What About Me?  I'm just a DIY End-user.  They won't sue me will they?

We all go out and buy the resistors, capacitors and opamps etc. ourselves.  The suppliers of those legitimate non-key components don't know what we plan doing with them so they are off the hook.  The suppliers of any fake components have their own answering to do.  But when we finish soldering it all together (in Australia) we could infringe patents under the 'use' provisions whether we purchased a PCB from ESP or not.

Technically the patentees could sue us, but they'd have to be pretty crazy.  It would be thrown out of Court as a complete waste of time and I do not know of any reported case in which an individual one-off end user has been sued for patent infringement.  Litigation expenses are enormous.  Hundreds of thousands (and even millions) of dollars are spent in patent litigation matters.  'Costs' are awarded to the successful litigant, but these don't scratch at the actual legal expenses.  So the likelihood of being subject to litigation over the purchase of a $25 PCB plus components to populate it (say another $80 ) is unlikely in the extreme.

Unless you are out there selling your completed wares or using your finished project in a public manner, a patentee will not even know about you.  Even if they did, their awards as either damages or an account of profits would be meagre.  There would be no sale profits as you didn't sell anything.  Even if you used your project back-stage in a concert or in the mixing of a musical recording or something, it is seriously doubtful that any profits from ticket sales or CD sales could be attributed to your having used the project somewhere in the background.  So that leaves damages which would be based on the item that you 'ought to' have purchased from the patentee if you hadn't decided to be a DIY type, and they'd have a tough time proving that a DIY type was ever going to spend more than 100 bucks on a commercial crossover anyway.

Well, although not originally designed so, the patent system has evolved into an 'elite sport' to be played out in the Courts only by those with very deep pockets and enormous commercial interests.  Back yard inventors would be blind to file patent applications with a view to self-funded enforcement of their potential rights.  What most applicants hope for is to attract the interest of the large players with a view to selling or licensing their patent rights to them.

Neither Rod nor myself are aware of any patents which might cover any of the projects for which PCBs are currently supplied by ESP.  If any reader knows of a commercial product similar to an ESP project that is marked with a patent number, please contact Rod with the information.

Well that's Patents 101 (more or less).  I hope it was easy reading, a little informative and not too boring.

All the best, Ian

Firstly, my thanks to Ian for putting this together.  The patent system is mysterious to most people, the language is arcane, and the audacity of some people to patent products that are clearly a minor variation of common knowledge is mind-boggling.  One of my articles on the use of current drive for loudspeakers was cited as prior art in a patent (which was granted), and the patented product does nothing markedly differently from the current drive system I described.  Some digital signal processing was added, but it appeared to be no different from analogue filtering that was done in the past by others.

There is no doubt that the 'elite sport' that Ian referred to is in full play.  If in any doubt, read the claims and counterclaims of major microprocessor manufacturers and personal computer operating system vendors (I think you can guess to whom I refer).  Multi-million dollar court battles have become a spectator sport ... at least for the IT observers.  The only thing we can be sure of is that we, as customers, will be the ones who pay the price of these battles, with inflated processor and/or operating system prices.

Useful patent references are ...

Google Patent Search (US only at the time of writing).
Esp@cenet - Worldwide Patent Search, European based.
IP Australia
British Library Patents Information
UK Patent Office
Canadian Patent Office
European Patent Office
United States Patent and Trademark Office

Main Index

Articles Index