[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales High Court (Patents Court) Decisions
You are here: BAILII >> Databases >> England and Wales High Court (Patents Court) Decisions >> PCME Ltd v Goyen Controls Co UK Ltd [1999] EWHC 830 (Pat) (22 January 1999) URL: http://www.bailii.org/ew/cases/EWHC/Patents/1999/830.html Cite as: (1999) 22(5) IPD 22043, [1999] EWHC Patents 264, [1999] FSR 801, [1999] EWHC 830 (Pat)

[New search] [Printable RTF version] [Help]

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
PATENTS COURT

		22/01/99

INTRODUCTION

1. Many children at school have witnessed the effect of rubbing fur on amber to produce an electrical charge. Similarly, it is a well known party trick to take an inflated balloon and rub it on a piece of wool. The balloon acquires an electrical charge which can be strong enough to make it stick to a wall or the ceiling. These are examples of a phenomenon called triboelectric charging. That phenomenon occurs in other situations and can be put to more sober uses. For example it is known that if particles are suspended in a gas stream and made to flow along a duct or tube they can acquire an electrical charge. If the flow of charged particles is made to pass by or round an electrode, the electrode itself will acquire electrical charge. The precise mechanisms by which the particles are charged and the charge is transferred to or induced in the electrode are, even now, not known. But there are some theories which at least explain in general terms what is thought to be happening. This patent infringement action is concerned with an industrial application of triboelectricity. A brief summary of the mechanisms thought to be involved in triboelectric charging and some basic electronics will help in an understanding of some of the issues in this case.

TECHNICAL BACKGROUND

2. There are thought to be at least two mechanisms involved in the triboelectric charging of particles in a gas stream. As illustrated in Figure 1, some particles rub against each other. The result is that one or more electrons pass from one to the other. This produces two particles with equal but opposite charges. If one particle which is already charged collides with another which is not, the former may transfer its charge to the latter. The second mechanism involves particles colliding with or rubbing against the duct wall. Once again there may be electron transfer resulting in the wall having one charge and the particle an opposite charge. The net result of these collisions is a stream of charged particles.



3. There are also a number of proposed mechanisms which explain how an electrode placed in a stream of charged particles itself becomes charged. The first two of these possible explanations are illustrated in Figure 2 below. In the first, illustrated on the left of the drawing, charged particles collide with the electrode. They transfer some or all of their charge to the electrode. The second, illustrated on the right of the drawing, involves particles colliding with the electrode and generating a charge by the same basic rubbing mechanism illustrated in Figure 1 above.



4. The third mechanism is illustrated in Figure 3. As a charged particle passes the electrode it induces an opposite charge in the electrode. As the charged particle approaches the electrode the induced charge will grow. As it passes and recedes, the induced charge will wane.



5. Another mechanism of charging the electrode was suggested by Dr. Wakefield, an expert witness called on behalf of the plaintiff, during the course of his oral evidence. It was not one which had been suggested by him before, nor was it one which was known generally in the art, let alone accepted. The existence of this additional mechanism was not put to the defendant's expert, Professor Stuart. In the end this does not matter because, as I have said already, the precise mechanism by means of which triboelectric charging takes place is not known. The processes illustrated above are only likely qualitative explanations for some of what is happening. Whatever the explanation, it is a known and observed fact that particles flowing in a gas stream acquire an electric charge and that an electrode can be used to obtain an indication of its nature. Before considering the nature of the electrical signal which can be obtained from such an electrode it is necessary to bear in mind some further basic principles.
6. A voltage or current which stays at a fixed size indefinitely is called d.c. It is represented graphically as follows:



2. A voltage or current can fluctuate with time. If it does it is called a.c. It can be represented as follows:



3. Any voltage or current which varies in size with time, can be notionally split into a number of a.c. signals like that shown in Figure 5. Finally, a voltage or current can be made up of a d.c. and an a.c. component. It can be represented as follows:



7. The electrical signal which can be obtained from an electrode along or around which a stream of gas-borne particles flows is complex. It contains both d.c. and a.c. components.

THE PATENT IN SUIT

8. In this action, the plaintiff, PCME Limited, is the registered proprietor of UK Patent No. 2,266,772 which will be referred to as "Rigby" after the name of its inventor. The invention which is at the heart of this patent is directed at using the triboelectric effect to monitor the particle content of gases coming out of a chimney or stack. The specification acknowledges that in the prior art it had already been suggested that a metal probe be inserted into a duct to give an indication of the mass of the particles flowing in it. This prior art is known as DeChene. It is relied on by the defendant, Goyen Controls Co. U.K. Limited, in its attack on the validity of the Rigby patent and will be considered below. The Rigby patent identifies its contribution to the art as consisting of looking at the alternating component in the signal from the triboelectrically charged probe rather than looking at the whole signal and using that component alone to give an indication of the mass of particles in the gas flow. The reason this is said to be advantageous is as follows. The whole signal is said to be affected by a variety of what are described as "gas flow related variables" in the particle-bearing gas. These variables include some which are related to the particle flow. Examples of those are stated to include the mass flow rate of the particles and the velocity and size of the particles. Other variables are not related to particle flow. Examples of the latter are identified as humidity in the region of the gas flow, temperature of the gas flow and the thickness of the particulate layer deposited on the probe. The claimed advantage of the use of the alternating component in the signal from the probe is that it is much less affected by the latter type of variables. Therefore measuring the alternating component in the triboelectrically generated signal from the probe is likely to be a more accurate way of monitoring the variables which are related to particle flow, including the mass of the particles in the gas stream.
9. To achieve this type of monitoring, the d.c. component in the signal from the probe has to be removed. The remaining a.c. components are then used to operate a gauge, sensing circuit or an alarm. This is conveniently done by removing the d.c. component from the signal before it is amplified. Wherever the removal of the d.c. component occurs, the output amplified a.c. signal is calibrated, for example against known levels of dust in the exhaust gases. When an alarm is to be activated, it will switch into operation at the level of electrical signal which represents a particular level of dust. The method of removing a d.c. component from a mixed signal fed through an amplifier is well known. An electrical component known as a capacitor is frequency sensitive. It blocks or impedes the passage of d.c. current but will allow a.c. to pass. The range of a.c. frequencies it will allow to pass without significant impedance is determined by the characteristics of the capacitor. Capacitors can be made which allow through all but d.c. and the lowest frequency a.c. signals. To make an amplifier which amplifies only a.c. signals, a capacitor has to be inserted at a suitable location in its circuitry. Such an amplifier is said to be 'a.c. coupled' or 'capacitively coupled'. On the other hand an amplifier which can process a d.c. signal is said to be 'd.c. coupled'. An a.c. coupled amplifier is incapable of passing and amplifying the d.c. component in an input signal. For this reason the patent describes connecting the output of the probe to an a.c. coupled amplifier.
10. The electrical signal generated by the passage of the particles in the stack will change as their mass changes. The change need not be and is not asserted to be linear. The mass of the passing particles will be increased if the velocity of the gas increases or the amount of particles in the gas increases or both. The patent does not suggest any way for distinguishing between any of these.
11. The defendant is a competitor of PCME. It sells in the United Kingdom a device known as the "Goyen EMS4". It incorporates an a.c. coupled triboelectric probe and is used to monitor particle levels in stacks. In these proceedings PCME alleges that the EMS4 infringes the patent. Goyen has counterclaimed for revocation. Although a number of claims were said to be independently valid, in the end only two were the subject of significant argument. They are a method claim, Claim 1, and an apparatus claim, Claim 18:

Claim 1: "A method for detecting particles flowing in a gas flow along a stack and emitted through the stack in which a probe is positioned so that it projects into the flow of particles in the stack and is charged triboelectrically by particles in the flow and the quantities of electrical charges transferred to the probe are evaluated to provide an indication of the particle flow in the gas flow, wherein, in order to reduce the effect of variations in 'gas flow related variables' other than those relating to particle flow, an alternating component in the signal caused by the triboelectrical charging of the probe is monitored, and the magnitude of the alternating component is itself used to give an indication of the particle flow through the stack."

Claim 18: "An apparatus for detecting particles flowing in a gas flow along a stack and emitted through the stack, the apparatus including a probe for installation in the gas flow, an electric circuit for generating an electric signal from electrical charges transferred to the probe as a result of triboelectric charging thereof by particles in the gas flow, and evaluating means for providing an output signal in dependence upon the electric signal generated in the circuit, wherein the apparatus is used to reduce the effect of variations in 'gas flow related variables' other than those relating to particle flow, the electric signal that is generated by the circuit has an alternating component and the evaluating means is arranged to provide an output in dependence upon the magnitude of the alternating component of the electric signal."

12. The only other claims which were relied on as having independent validity were claims 2, 11 and 12. Claim 11 is a use claim. It is in substance the same as Claim 1. The other two claims relate to the incorporation into the circuitry of a low pass filter of unspecified frequency characteristics. A low pass filter is a device which blocks or significantly impedes the passage of alternating currents with a frequency above a pre-selected value. In many applications an electronic circuit will pick up extraneous a.c. signals. For example the mains supply in this country consists of an alternating current which has a primary frequency of 50 cycles per second (50 Hz). An electric circuit connected to the mains or located near apparatus connected to the mains will frequently pick up unwanted a.c. signals which are at 50 Hz. This is known as "hum". This can be removed by passing the output of the device through a low pass filter which blocks the passage of a.c. signals at 50 Hz and above. Of course a filter which blocks the passage of a.c. signals at, say, 40 Hz and above will also work. This is well known and frequently used. Claims 2 and 12 cover the use of such a hum-suppressing filter. By the time of the speeches, there was no serious dispute that if the other claims were invalid, these claims would fall with them.

Construction of the claims.

13. A number of features of the invention and these claims should be noted at the outset. The words "in order to reduce the effect of variations in 'gas flow related variables' other than those relating to particle flow" in Claim 1 and their equivalent in Claim 18 are surplusage. As the plaintiff accepted, they do no more than describe the alleged benefits of using the alternating component from the signal. They do not limit the scope of the claims. Secondly, in neither Claim 1 or 18 is there any bottom limit on the frequency of the a.c. signal. Even Claim 2 only limits it to frequencies of 1 cycle every 10 seconds and above. Similarly the main claims do not require the presence of a filter circuit to exclude high frequency a.c. signals. Thirdly, the claims are not limited to any particular type of stack, any particular type of probe (save that it must project at least some distance into the stack cavity), any particular velocities of stack gases or any particular loading of particles in the stack gases. They are of great width. There were two significant points of construction in relation to which the parties were in disagreement. I shall deal with them below.

Stacks v. Conveyors

14. The claims are all limited to measuring particulate flows in or along "stacks". Some of the pieces of prior art relied on relate to triboelectric devices used in conjunction with pneumatic conveyors. The distinction between the two therefore assumed some importance during the trial, notwithstanding the fact that the parties had agreed the meaning of these terms in a technical primer which was supplied to me just before the trial. That document states as follows:

"A 'pneumatic conveyor' transports powdered material within a gas stream by maintaining a sufficiently fluid nature to the mixture to allow it to be transported along a pipe. The flow rate is usually high in order to maintain the fluid character of the mixture and to transport as much material as possible.

A 'stack' is a means of conducting waste or exhaust gases from an industrial plant into the atmosphere. Particle flow in a stack is much less dense than in the case of pneumatic conveyors."

15. It is apparent that neither "stack" or "conveyor" is a term of art having a precise meaning. There was no criticism directed at the plaintiff for having used the word "stack" in its claims. In the overwhelming majority of cases a man in the art would have no difficulty in distinguishing a stack from a pneumatic conveyor. The former is generally pipe or duct shaped and would have much less particulate material in it than the latter. The difficulty is not that an imprecise term has been used for defining the monopoly. In some cases, of which this is one, it may not be possible to mark the border with precision. The difficulty is that the less precise the border, the more difficult it is to decide whether a piece of prior art (or an alleged infringement) falls within the claims. For present purposes it is sufficient to note that the parties agree that the particle flow in a stack is likely to be much less dense than in the case of pneumatic conveyors. It appears from the evidence that even when a plant malfunctions, for example where one of the bags used to filter particles out of exhaust gases before they are vented through the stack bursts, the level of particulate material in the stack will be 3 or 4 orders of magnitude less than in a conveyor operating in normal conditions.

"The magnitude of the alternating component is itself used to give an indication of the particle flow through the stack".

16. By the end of the trial, there was no dispute that the word "magnitude" had no special meaning. It refers to the bigness of the alternating component. Also it was not in dispute that the "indication" of particle flow need not be linear. In other words, a doubling of the magnitude of the alternating component need not indicate a doubling of the particle flow. But the defendant argued that although the alternating component in the signal from the probe must be used to give an indication of particle flow, this did not mean that it had to be the only component used to give such an indication. Mr. Birss, who appeared on behalf of the defendant, argued that the method claim requires one to monitor and use the alternating component but it does not require one not to monitor something else. In particular he said that if one monitors and uses the whole signal, including the alternating component, then that falls within the words of the claim.
17. The significance of this point is as follows. If a mixed a.c. and d.c. signal is passed through a d.c. coupled amplifier, one of two things will happen to it. If the amplitude of the a.c. component is smaller than the amplitude of the d.c. component, the ripples on either side of the d.c. component will be cancelled out by the circuitry. The output will relate only to the size of the d.c. component. This is illustrated in Figure 7 below.



4. On the other hand if the amplitude of the a.c. component is larger than the amplitude of the d.c. component, the output from the d.c. coupled amplifier is related to the size of both components. This is illustrated in Figure 8 below.



5. It is not necessary to understand the reasons for this difference in behaviour. It is sufficient that the parties agree that the difference exists. The result is that in this set of circumstances if the output from the d.c. coupled amplifier is used as an indication of particle flow, it will be responding to both the a.c. and d.c. components of the input signal coming from the probe. Mr. Birss says that this is within the scope of the claims. I do not agree.

18. The patent explains its underlying principle in the following way:

"We have discovered that by looking at the alternative component in the signal from the triboelectrically charged probe [in the stack] rather than looking at the absolute value of the signal, it is possible to obtain an output signal that is much less affected by gas flow related variables, other than those relating to particle flow, than in the case of, for example the apparatus described in [DeChene]". The Rigby patent, p. 3 line 5 et seq.

6. The specification goes on to say that it might appear to be disadvantageous to evaluate the much smaller amplitude alternating signal component in the signal from the probe. It asserts that the inventor has found that that component is not affected by the non-gas flow related variables which are likely to affect "the absolute value" of the probe signal The Rigby patent, p. 3 line 12 et seq. In other words the central feature of the invention is to separate the a.c. component in the probe signal from the rest of the "absolute signal" and to use it alone even though it is smaller as an indication of particle flow in the stack. This is said to be counter intuitive. If Mr. Birss' argument were correct, it would mean that the claims cover those types of arrangements which the specification expressly indicates are not desirable and form no part of the inventive contribution.

19. In my view that is not what the claims say or mean. The claims refer to the alternating component in the signal. The use of the word "component" emphasises that it is not the whole signal which is being looked at. Claim 1 then says that it is that component "itself" which gives an indication of particle size. In the context that means "by itself". Looking at the whole signal does not fall within this wording. If I was in any doubt about this, it seems to me that this is an example of an occasion where the provisions of s. 125(1) of the 1977 Act, Article 69 of the EPC and the Protocol on Interpretation are of particular significance. The specification should be used to resolve any ambiguity here and to restrict the claims to those where the alternating component alone is monitored.

VALIDITY

20. Although by the end of the trial a dispute relating to infringement continued to exist, the major issues related to validity. A number of prior publications were put in issue and one alleged prior use. It is convenient to start with the prior use.

(A) Blasto

21. A company called Blasto Industries Pty. Ltd. manufactured a system for indicating the dust content of gases in stacks. It consisted of a probe for insertion into the gas stream together with associated electronics connected to an alarm circuit. The movement of the particles in the gas stream around the probe generated a small electrical signal tribolectrically. The signal was passed to an a.c. coupled amplifier. The amplifier therefore only analysed the a.c. component in the signal. The device had to be calibrated. The device was made available with sales literature which gave some indications of how it was to be used and its performance characteristics. The defendant asserts that public sales and delivery of Blasto probes and systems took place in Australia before the priority date. On the basis of this it runs arguments of anticipation and obviousness. In the end four issue arise for consideration namely: (1) Has the defendant proved that the Blasto device and literature was used publicly prior to the priority date, (2) If it has, what does such prior use disclose to those in the art, (3) Did such disclosure amount to anticipation and (4) Does such disclosure render the Rigby patent obvious.

(1) Was there public use of Blasto probes before 30 April, 1992?

22. The defendant's case is that the Blasto probe and system were manufactured by February 1992 and sales were achieved in March and April of that year with deliveries to customers taking place before 30 April, 1992. In support of this it relies on the oral evidence of Mr. William Thorp and Mr. Geoffrey Barclay and the written evidence of Mr. Michael Jedra which was provided under a Civil Evidence Act Notice. Mr. Thorp is an employee of the company which owns the defendant in this action. In 1992 he was the Managing Director of Blasto Industries Pty. Ltd. before it was purchased by Goyen. Mr. Barclay also was employed by Blasto and is now employed by the same company as Mr. Thorp. He was the engineer who designed the electric circuitry for the Blasto devices. Mr. Jedra is employed by a company called Boral Concrete. He says that his company installed a Blasto triboelectric emission monitor at its premises in Brisbane in February or March 1992.
23. Mr. Thorp's evidence is that in January 1992 Blasto Industries received a copy of tender documents for Boral Limited. Boral required a significant numbers of triboelectric emission monitoring units. The tender documents related to approximately 270 of Boral's sites with a closing date for the tender being 30 March 1992. Mr. Thorp estimated that each site would require two of Blasto's level monitoring probes and two triboelectric emission monitoring units. At that time Mr. Thorp did not know what was a triboelectric emission monitoring unit was but after carrying out some crude experiments in Blasto's factory, decided to explore the possibility of adding triboelectric emission monitors to the company's product range and tendering to Boral. Mr. Barclay had learned that Mr. Thorp was looking for an electronics engineer and telephoned him on Wednesday afternoon, 22 January 1992. The two met on Friday, 24 January 1992. Mr. Barclay was offered a contract position to start on Monday, 27 January 1992. In fact Mr. Barclay worked on the project over the weekend and on the Monday already had a hand drawn schematic design for the filter amplifier circuit for the monitor. According to both Mr. Thorp and Mr. Barclay, within two weeks of starting the project Mr. Barclay had an operating prototype ready. Mr. Thorp's evidence is that Blasto's first emission monitoring units were offered for sale to the public in early March, 1992 designated generically as B01779. The low price set by Blasto apparently caused potential customers to be reluctant to buy. Blasto therefore offered to let customers take delivery of one or more units with which they would be quite welcome to conduct any tests of their devising. Unconvinced potential customers were offered the supply of a Blasto triboelectric probe for a period of one to three months for evaluation and testing before they were asked for a definite purchase commitment. Mr. Thorp says that many of the B01779 devices initially supplied this way were invoiced to the customers on the basis that if returned the device would be credited and Blasto intended to sell returned devices so long as they remained in a salable condition. No customer who was provided a device on approval or who initially bought any device outright was put under any obligation or restriction of any kind on how or whether that customer tested the device or to whom he or she disclosed any aspects of the Blasto devices. Mr. Thorp says that his company made good sales of the new product. Mr. Barclay's evidence supports that of Mr. Thorp. It explains the speed with which he was able to design the necessary circuitry for the Blasto monitor.
24. In support of Mr. Thorp's and Mr. Barclay's version of events, a number of documents and prototype printed circuit boards were disclosed and are relied on. One document is an invoice dated 10 April 1992 relating to a sale to Air Engineering Pty Ltd. Mr. Thorp has reason to believe that the unit sold to Air Engineering Pty Ltd was immediately on-sold to Controlled Environment Pty Ltd of Heidelberg, Victoria, Australia but delivered to a Sydney firm of patent attorneys, Spruson & Ferguson who immediately delivered it to their client Auburn International in the USA, the owners of the DeChene patent. The defendant also produced one of the probes which it says it had supplied to Boral. Unfortunately it was misplaced before the trial and efforts to find it have failed.
25. Although the plaintiff has not formally accepted the authenticity of any of the documents relied on by the defendant, no suggestion has been put to any witness that any of them are not authentic. On the other hand the plaintiff says that evidence of prior use is wholly within the hands of the defendant. It is therefore under an obligation to put its best case forward. In particular it is said that the defendant should have produced one or more customers from Australia to support the allegation of prior use. A witness statement from Mr. Jedra is not enough. It says that if there are any doubts as to the accuracy of the defendant's documents it should be resolved in the plaintiff's favour. I accept that the court should pay particular attention to whether the defendant could have produced better or more complete evidence at the trial. That is a factor I have taken into account.
26. Mr. Meade subjected Mr. Thorp and Mr. Barclay to detailed and searching cross-examination. He analysed with Mr. Barclay the details of the circuit diagrams for the Blasto probe. He pointed out that one of the entries on one of the documents relied on by Mr. Thorp included a reference to a sale which was recorded on that document as having taken place on a date which in all the circumstances must have been incorrect. On the basis of this admitted error in Blasto's records he suggests that none of the dates given for other recorded sales in that document can be relied on. He also says that the first advertisement for the Blasto probe showed a picture of a similar device, but not the probe itself. The cross examination on these and similar points covers many pages of transcript. Notwithstanding the precision of this cross-examination, I have no hesitation in holding that the defendant has convincingly demonstrated that the Blasto probe and emission monitor and its associated literature was made public before the priority date. Such limited inconsistencies as Mr. Meade was able to point to were readily and credibly explained as minor clerical errors and the like. Although it would be possible to go through all of the evidence in detail, one example of the material put forward by the defendant will suffice.
27. Mr. Thorp says that once the Blasto emission monitor had been put on the market it had come to the attention of Auburn. As mentioned above, it appears that Auburn obtained a sample of the Blasto product by placing an order through Air Engineering Pty Ltd. which forwarded it to Controlled Environment Pty. Ltd. Mr. Thorp says that a complaint relating to this product was made to him by or through Controlled Environment on 27 March 1992. In response to that he said that he sent out a comfort letter to customers. That letter is dated 2 April 1992 Bundle E Tab 11.. It refers to the telephone contact by Controlled Environment, asserts that the management decision to develop and market a dust monitoring device based on the triboelectric operating principle was made on 27 January 1992 and goes on to state:

"The product was tested and then released for marketing during the first week of March 1992 at a price of $2000.00 per unit."

28. Mr. Thorp's evidence on this issue was not shaken. There was nothing to suggest that the letter of 2 April 1992 was not accurately dated. The reference to the marketing of the Blasto triboelectric product nearly 2 full months before the priority date is consistent with the various invoices which have been exhibited by the defendant and the evidence of Mr. Barclay. The whole of the evidence hangs together. I find the prior use to have been proved.

(2) The extent of the disclosure

29. The prior use and publication of the Blasto device and its associated literature does not prove that the circuitry in the device was published. The literature does not describe in any detail how the circuitry in the device operates. Furthermore the device itself was a 'black box'. In 1992 Mr. Thorp was concerned to hinder any competitor who might want to copy his company's new products. Therefore the devices supplied to customers had their electric circuitry embedded in resin. A customer would not be able to dismantle the device physically. A crucial issue on this part of the case is what, if anything, the Blasto device and its literature would have disclosed to members of the public. On this subject my attention was drawn to Lux v. Pike [1994] RPC 107, Milliken Denmark v. Walk-Of Mats [1996] FSR 293 and Availability to the Public, EPO Enlarged Board [1993] EPOR 241. In the latter, the Enlarged Board said:

"Where it is possible for the skilled person to discover the ... internal structure of the product and to reproduce it without undue burden, then both the product and its ... internal structure become state of the art. ...

It is the fact that direct and unambiguous access to some particular information is possible, which makes the latter available, whether or not there is any reason for looking at it."

30. Basing himself on this, Mr. Meade argues that it is only information directly and unambiguously obtainable from an examination of the black box which is to be treated as published. Furthermore, relying on lines in Jacob J's judgment in Milliken, he says that it was only "easy and ordinary techniques" and "simple test(s) ... of the ordinary sort" which could be used to examine the prior art. Therefore, tests which, as a practical matter, a skilled man would not employ are irrelevant. These submissions are directed particularly at the evidence of Professor Jan Stuart who was called on behalf of the defendant. The Professor explained how he would have analysed the Blasto device non-destructively to draw firm conclusions as to how it operated. In particular he said that it would be easy to discern by testing that the device was a.c. coupled. This meant that it must have been measuring or responding to the a.c. component in the signal produced by the probe. However the plaintiff argues that the ordinary skilled but uninventive worker would not have carried out the tests suggested by Professor Stuart. According to Mr. Meade what counted was what the skilled man could and would have done at the priority date. Mr. Meade says that the skilled worker would not have had the abilities of a skilled forensic engineer, such as Professor Stuart, so he would not have carried out the testing which Professor Stuart said was simple and obvious to do.
31. The disclosure in a document or a prior use is the information that document or use makes available to the skilled addressee. The skilled addressee may be a team. As Aldous J., as he then was, said in Lux, a machine, like a book, can be examined and the information gleaned can be written down. At its outer reaches this may appear an artificial exercise. The contents of a document written in the most obscure language is treated as if it was written in English. This is so even if it is most unlikely that any worker in the field would be able to read the language of the document. In effect the skilled team is treated as if it contains an ordinary skilled interpreter. It seems to me the same applies just as much to interpreting a piece of hardware. Mr. Meade relies on cross-examination of Professor Stuart in which the Professor accepted that for a man skilled in the relevant art, reverse engineering of the sort he was suggesting "would be out of the ordinary, but not totally unusual" Transcript page 744, line 26. and that "it would be something new to have a shot at ..." Transcript page 745, line 13.. In my view this does not undermine the value of the Professor's evidence. Even if by these answers the Professor meant that the tests he put forward would only have been simple to a forensic engineer rather than an ordinary skilled addressee in the art, it seems to me that the notional skilled team in the art must be treated as if it included such a skilled but uninventive forensic engineer. Were it otherwise, in some cases it would be possible to re-patent information which any competent forensic engineer would be able to find out by obvious analytical techniques. Similarly where an ordinary skilled man in the art of making particular types of chemical products has no experience of reverse engineering, it would be possible in some cases to patent information about a commercially available product which was deducible by a competent but uninventive forensic chemist. I do not think that is the law. In any event, having read Professor Stuart's evidence and listened to his oral testimony, I do not believe that he was agreeing that the tests he was proposing were outwith the normal skill and expertise of a man in the art. It seems to me that all he was saying was that most men in the art are rarely faced with the task of reverse engineering so it is not something that they do frequently in the ordinary course of their work. I accept his evidence that they are the types of tests which would be obvious to any ordinary skilled worker in the art who wanted to find out without undue effort what he could about the operation of the Blasto probe. The fact, if it be one, that a man in the art would not frequently need to make such inquiries is irrelevant.
32. It seems to me that three questions need to be answered in any case where one is trying to decide what information about the composition or structure of a product is made public by reason of the public availability of the product namely (1) what tests or analyses would it be obvious to carry out on the product, (2) what would be the result of those tests and (3) how would those results be interpreted by the man skilled in the art. The third of these questions is of particular importance in the light of the way Mr. Meade puts his case. If, as he says, it is only information unambiguously obtainable from an examination of the black box which is to be treated as published, it is easy to conclude that unless the interpretation of the results are unambiguous, they must be ignored. The fallacy of this can be illustrated by way of an example. A series of obvious tests are carried out on a product. There is no dispute as to the results which are obtained. Furthermore there is no dispute that any man skilled in the art would conclude "it is highly probable that this product contains ingredient A but there is a small chance it contains ingredient B instead". Because the interpretation is ambiguous, does that mean that it must be ignored? In my view the answer must be no. What are to be treated as published are the results of the obvious test procedures and their normal interpretation. The fact that the latter may be inconclusive does not mean that it is not made available to the public. If the statement "it is probably A but may be B" is found in a document it is no less published than the statement "it is definitely A". The same applies to information derivable from inspection of a product. If the information deducible is imprecise, that will affect whether it is sufficient to support anticipation. It does not mean that it is to be treated as not published. It may well render a patent claim obvious.

(3) Anticipation by Blasto

33. The Blasto product was made available with limited documentation. Mr. Meade accepted that they would be analysed together by at least some customers. The product summary leaflets See Bundle D Tab 3. make it clear that the devices work by triboelectricity. The invoices make this clear also. The leaflets also state that this effect is harnessed to monitor factory emissions, i.e. from stacks, and that the sensing element or probe is given "a minute residual voltage" by the triboelectric effect "which is then amplified by the internal circuitry and then processed for further output control." So it is clear that the circuitry contains an amplifier which amplifies the very small electrical signal from the probe and that it is the amplified signal which then is used for further analysis or control. It also says that the output can be used to feed to external industry standard monitoring equipment such as programmable logic controllers or chart recorders. The literature also says that the device has adjustable sensitivity Bundle D Tab 3 page 9. and it gives a brief explanation of how to calibrate the device. What the literature does not disclose is the nature of the amplifier used and, in particular, whether the whole or only a part of the signal from the probe is amplified.
34. The analysis which Professor Stuart says would be carried out is set out in paragraph 10 of his first report. In summary it is as follows. The first step is to subject the probe to a small d.c. voltage and to read the output from the device. It will show an initial rise in output but within 2 seconds the output would fall to zero. In other words it was failing to amplify a d.c. signal. The probe would then be subject to a variety of a.c. signals of differing frequency. Looking at the output from the device would show that some frequencies are more amplified than others. The frequency response of the amplifier would therefore be elucidated. This would show that the circuit in the device responded to frequencies between about 0.5 Hz and 25 Hz (or 0.3 and 17 Hz, depending on which particular circuit was used in the device). None of this is complicated. The plaintiff's expert, Dr. Wakefield, was cross-examined on this in some detail Transcript Day 3 pp 323 - 343. . Although it was apparent that he was reluctant to concede points, in the end he did not dispute that the analysis proposed by Professor Stuart was a sensible and obvious one to follow. He said that it might prove tedious and testing the device would have to be done with care, but the overall approach was not one that he challenged with any enthusiasm.
35. Professor Stuart's conclusion was that any skilled man in the art who carried out these tests would realise at a very early stage that the Blasto device was a.c. coupled and that its circuit had a frequency response of between about 0.5 Hz and 25 Hz. It would be apparent therefore that the device was responding to an a.c. signal derived from the probe. Dr. Wakefield said that this conclusion was not inevitable. A flavour of his evidence is given by the following extract:

"Q. What I want to suggest is if you knew what the shape of your pulse was, whether it was a [sine] wave or a square wave pulse, and you saw signals like that coming out, you would know exactly what was going on. You would say, "There is a capacitor in parallel near the output producing this kind of figure."

7. A. You do not know at what stage in your circuit at all this is being produced. It could be produced right at the beginning. You have no means of knowing that. The second point was that you would be at a loss to know why it was doing that.

8. Q. Never mind worrying about why it was doing it. You would know how it was doing it. You would know that somewhere in the circuit, there was a parallel capacitor producing that kind of wave form.

9. A. Yes, but there are some, I admit, pretty unlikely other circumstances which relate to instability. You cannot dismiss them altogether, but if the thing was unstable at certain frequencies, if devices limit and there is a capacitor somewhere in the system, you can get output of that kind. I admit that is not a likely scenario. It is a possibility, but not a very likely one." Transcript day 3 p. 339.

36. At other passages he accepted that it was "likely" there was a capacitor in parallel in the Blasto circuit. However there are two passages in the cross-examination on which Mr. Birss relies in particular. The first is as follows:

"Q. The fact that it has gone down again -- it will go up, but it will go down again -- with a longish time constant tells you that the device is AC coupled, does it not?

A. That is a likely inference.

Q. Can you think of another likely inference other than that one?

A. The only other one is that, for some reason, you can saturate the device.

Q. You would be able to go back and test it and make sure you have not blown it.

A. You would. Unfortunately, if you applied a volt to it, it would be inactive for some time because you would have pushed the value on the capacitor well beyond the maximum value required for a full-scale output. This will mean that it will take some time to recover to add to your confusion, but eventually you should, if you persevere long enough, arrive at a conclusion." Transcript Day 3 p. 332.

37. Here Dr. Wakefield's objection was that it was not desirable to start with one volt as the testing voltage. But subsequent cross-examination demonstrated that he would have expected a worker to have started with a smaller voltage and work up, so the problem would never have arisen in any event. But even ignoring this, it is apparent that the only ground for holding any doubts that there was a.c. coupling was one which would disappear with perseverance. The second passage is:

"Q. That would tell you that the most likely explanation is that the device is AC coupled.

A. I think you would probably decide that." Transcript Day 3 p. 335.

38. I have come to the conclusion that Professor Stuart is right and that obvious and simple testing techniques would disclose that the Blasto device was a.c. coupled and would also disclose its frequency range. As Dr. Wakefield accepted, it would be easy to make a device which imitated these features of the Blasto device. This means that the Blasto device would disclose to those skilled in the art that it incorporates circuitry which responds only to an a.c. signal derived from the probe, or the a.c. component of the signal if it contains a d.c. component as well. Although the detailed structure of the Blasto circuitry could not be discovered by reverse engineering, these broad features would be disclosed and the claims are directed to just such broad features. For these reasons I have come to the conclusion that the Blasto prior use anticipates both Claim 1 and Claim 18. The instructions to use it anticipate Claim 11.

(4) Obviousness over Blasto.

39. However, even if I accepted Dr. Wakefield's hesitation as representing the views of the average skilled man, it would not alter my view on validity. Although at times Dr. Wakefield says that a worker could not be certain that the Blasto device was a.c. coupled, he accepted that that such a worker would think it was the most likely explanation of its behaviour. I have no doubt that a skilled worker in this art presented with the Blasto probe and told, as he was, that it could be used for monitoring dust levels in a stack and interested in making a competing product would have tried to copy it or even improve on it. In relation to the type of amplification circuit to be used, although his options would be open, I think it would have been obvious to try to use the type of circuitry which he would think was most likely to have been used by Blasto itself, namely an a.c. coupled amplifier. Such a device would fall within the claims of the Rigby patent.
40. Notwithstanding this finding, I must deal also with the other attacks on validity.

(B) DeChene (PCT Application WO86/02454)

41. The patent application of DeChene is acknowledged in the Rigby patent as prior art. It is directed to a flow measuring device which uses triboelectric charging to measure the flow velocity and mass flow rates of gas flows containing suspended particles. To do this it says that a metal probe is inserted in the stream of flowing solids (i.e. the particles in the gas stream) and that the charge is turned into a current which is fed through an electronic circuit which monitors the rate of charge transfer. It says that this rate of charge transfer (i.e. current) is approximately proportional to the flow rate of the solids hitting the probe Page 2, line 23 to page 2 line 5.. The patent is concerned solely with the electronic circuitry used to monitor the charge on the probe. The following circuit diagram is used to illustrate the DeChene invention:



42. In fact this diagram is a composite. It incorporates two possible circuits. The first starts with the probe (1) which is represented by the rectangular box immediately below the letters "FP" on the left of the diagram. The latter letters refer to the flow pipe within which the particle bearing gas is flowing. The probe, as shown, projects into the flow pipe. The charge transferred to or induced in the probe is then led through the circuitry from the input relay (5), through the amplifier (10), the absolute value circuit (12), adjustable filter (14, 15) and then through the amplifier voltage follower (16). The output is then led to four separate destinations namely (a) a voltage to current circuit (18) This is recorded on the diagram as "4 to 20 mC OUT". This should read "4 to 20 mA OUT". In other words it is a reference to a 4 to 20 milliamp output. There was no dispute between the parties that this is what the wording on the diagram should have said and that it is how a man in the art would have read it., (b) a milliammeter (20), (c) a 0 to 8 volt output and (d) an alarm circuit (22, 24, 23 etc.). It is not necessary to understand the precise operation of this circuit. There is no dispute between the parties that it is d.c. coupled and is designed to give an output which indicates particle flow. The second circuit starts with what is described as a "capacitively coupled probe" (1P). When this is used, most of the circuit in the diagram is bypassed by the operation of the switches 32 and 17. So, the circuit illustrates and the specification refers to both a d.c. coupled and an a.c. coupled probe. Mr. Birss said that both versions of DeChene anticipated at least the apparatus claim of the Rigby patent.
43. The argument in relation to the d.c. coupled version of DeChene was as follows. Although this was intended to monitor only the d.c. component in the probe signal output, when the a.c. component in that signal was larger than the d.c. component, the circuitry would in fact respond to an output which was related to both components. This is explained above in relation to Figures 7 and 8. Therefore the device would respond to the a.c. component and the disclosure falls within the claims.
44. There are at least two reasons why this argument fails. First it depends upon Mr. Birss' contention that the claims cover devices which monitor the whole signal, including the alternating component. For the reasons set out at paragraph 16 et seq. above, this contention fails. The claims of the Rigby patent are restricted to devices which monitor the a.c. component of the probe signal alone. There is no dispute that the d.c. coupled version of DeChene does not do that. Secondly, the argument of anticipation starts from the premise that sometimes the a.c. component of the probe signal will have a larger amplitude than the d.c. component. The precise mechanisms which generate each of these components is, even now, not known. There is no way of predicting what the relative sizes of the a.c. and d.c. components will be. There is no material before me which demonstrates that the a.c. component can ever have a larger amplitude than the d.c. component. It follows that it has not been proved that there will ever be circumstances where the output will be of the type illustrated in Figure 8 above, rather than that illustrated in Figure 7.
45. I therefore turn to consider the a.c. coupled version of DeChene. Once the switches have been operated to put it in this mode, the circuit will look something like this:



10. The description in DeChene of what this circuit is supposed to do is notable for its brevity and obscurity.

"A capacitively coupled probe 1P with OVP, which produces an AC signal, can be used with the aforementioned circuitry. With this prove (sic) 1P and switch, with section 29, 32, and 17, places the auto zero function in continuous operation and directs the signal from the current to voltage converter op amp output 8 directly to the voltage to current converter 18."

46. There are numerous peculiarities in relation to this part of DeChene. If what is being illustrated is a particle monitoring device it is surprising that the outputs most likely to be used for that purpose (i.e. those items referred to as (b) to (d) in paragraph 42 above) have been excluded from the circuit. No one seemed to understand how the circuit operated and DeChene itself does not state what it is supposed to be doing. Professor Stuart volunteered the view that the circuit was full of errors, was an "oddity" and was confusing. When he was asked to venture an opinion as to why certain parts of the d.c. coupled circuits had been omitted when they would have been useful even in the a.c. coupled circuit he readily accepted that he could not be sure. In one case he said:

"A. I found it difficult to be sure, but assuming that it was not yet another error, the only explanation I could think of was ..." Transcript p. 719.

11. This gives a flavour of his evidence in relation to the a.c. coupled version.

47. Mr. Birss says that none of this matters. The apparatus claim in the Rigby patent, Claim 18, is directed to an apparatus "for" detecting particles flowing in a gas flow. Mr. Birss says, and Mr. Meade agrees, that this means "suitable for". If a device is suitable for carrying out the analysis required in the rest of the claim it falls within it. This is so even if the designer of the device is not aware that it has the relevant characteristics and has not described it in that way. Since the claim is directed to an a.c. coupled triboelectric device suitable for producing an output which is dependent on the magnitude of the alternating component of the probe signal, all that the defendant needs to prove is that the a.c. coupled version of DeChene has those characteristics. Since it is described as a.c. coupled, it will inevitably select out the alternating component from the signal and the output will inevitably be dependent on the size of that component. The device must fall within the scope of Claim 18. Mr. Birss also says that the function and purpose of the a.c. coupled version of DeChene is not really as obscure as some of the oral evidence suggested. In particular he relied on the fact that the title and introductory passages of DeChene tell the reader that the patent is directed to triboelectric sensing devices for detecting mass flow rates of particles suspended in gas and that the output monitoring device which was left attached to the right hand end of the a.c. coupled circuit was the 4 to 20 mA voltage to current circuit. Professor Stuart said that this was a standard process control signal range. It follows, says Mr. Birss, that this is being used to gauge the amount of particulate material in the gas.
48. As attractive as these arguments are, I do not accept them. Common sense has to be applied to the law of anticipation. As we now know, a prior disclosure must be enabling before it can render a claim bad for anticipation. Here we have a device which is described in an incomplete, confusing and erroneous way. A man in the art who tried to build a device in accordance with the information provided in relation to the a.c. coupled version of DeChene could not be certain that what he built was what DeChene was describing. The description is not enabling and the attack of anticipation fails.
49. There is no material upon which an attack of obviousness based on DeChene could succeed and, as I understand Mr. Birss' submissions, in the end no such attack is pursued.

(C) Svarovsky (U.S. Patent No. 4,179,934)

50. This document is relied on both for anticipation and obviousness. Its title is "method and apparatus for monitoring particle sizes". It is particularly concerned with monitoring the particle size distribution of particles being carried in pneumatic conveyors. This is achieved with apparatus of the following general layout:



12. The pneumatic conveyor (3) carries particles and gas. A sampling probe (5) diverts a stream of the particles into a pipe (7) and into a chamber (9) in which it is mixed with an incoming jet of compressed air. The output goes along line (11). Because there has been mixing with compressed air, the particle/gas ratio in the line (11) is lower than in the conveyor. The diluted mixture is supplied to a device (13) which splits the incoming stream into two fractions. One is a relatively large air stream containing a low concentration of solids and the other a relatively small air stream with a high concentration of solids. The former will have mainly small particles in, the latter will have mainly large particles. These two streams are then subjected to triboelectric monitoring by use of the following apparatus:



13. One stream of particle-bearing air passes down tube 19, the other passes down tube 27. In each there is a probe (35, 37) which is charged triboelectrically by the passage of the particles. The electrical charge from the probes is fed to electric circuitry which compares the output from the two streams thereby giving an indication of changes in the quality of the product in the pneumatic conveyor.

51. The patent also describes how the signals from the probes are monitored. Each is passed first to an a.c. coupled amplifier. The reason for this is described as follows:

"Solid particles, when being conveyed in a turbulent gas stream, acquire electrostatic charge, mainly through triboelectric contact charging. A suitable probe inserted into the gas stream carrying the particles may be used to pick up both DC and AC current components. The former results from transfer of charge from particle-probe collisions (and also by induction) but the resultant DC current is subject to electrical leakage between the probe and the pipework carrying the gas stream and also due to the surface condition of the probe. However, the AC component, which is the electrostatic noise, is, like the DC component, strongly dependant on the solids concentration in the flow, but it is little affected by leaks or probe surface condition."

52. There is little doubt what this is saying. It is because the d.c. component is susceptible to errors due to leakage and the surface condition of the probe (e.g. it is covered by dust particles) that this signal is less reliable than the a.c. component (which is called in this patent "noise"). The a.c. component does not suffer from this problem. The use of the a.c. coupled amplifiers ensures that it is only the a.c. component which is used to give a measure of particle flow. It will be recalled that one of the examples of "gas flow related variables other than those relating to particle flow" referred to in the Rigby patent is the thickness of the particulate layer deposited on the probe. The reason given for the use of a.c. coupling in the Rigby patent is that it avoids the errors which the d.c. component associated with such gas flow variables. Svarovsky is choosing to use the a.c. component for the same reason as the Rigby patent.
53. The value of the signals obtained from the two fractions is also explained by Svarovsky. The noise or signal from each fraction is described as being "related to the mass flow rate of the solids in that fraction" Column 2 line 5.. The separator (13) effectively splits the contents of the conveyor into two streams one only containing particles above a particular size (the so-called "cut size") and the other only containing particles below the cut size. So by use of the triboelectric probes:

"... particle size measurement is reduced to the determination of the mass flow rate or concentration of the solids with the cut size calibrated by a standard laboratory method." Column 2 line 14.

14. Again the meaning of this is clear. Each probe measures the mass flow rate or concentration of solids flowing in its associated tube (19 and 27) by triboelectricity. This information is then used to build up a picture of the particle size distribution of the contents of the conveyor.

(1) Anticipation by Svarovsky

54. Mr. Birss says that the probes and associated circuitry described in Svarovsky would be "suitable for" use in a stack and therefore anticipate the apparatus Claim 18. Mr. Meade's argument against this is that the twin barrelled device illustrated in Figure 12 above is not suitable for insertion into a stack. The only thing which goes in the stack is the sampling probe (5) illustrated in Figure 11 above. Central to the Mr. Meade's submission is the implicit assertion that the whole of the disclosure in Svarovsky must be looked at together. It must not be divided into parts. In particular his suggestion is that the whole Svarovsky device is applied to the conveyor (3) as illustrated in Figure 11 above. So, one must consider the whole of that device applied to a stack. If so, then it is the sampling probe (5) on that figure which must be inserted into the stack, not the triboelectric probes (35 and 37).
55. I do not accept that this is the correct approach. One way of looking at this issue is to consider the question of infringement. Subject to the issue of enablement where, in accordance with recent jurisprudence, it appears anticipation and infringement differ, something which falls within and infringes a claim if it post-dates the patent, will anticipate the claim if published or used before the priority date. Assume, therefore, that the Svarovsky device was built after the priority date. Would it infringe? In my view it would. It incorporates a probe (35 or 37) which is "suitable for" inserting into a stack. The associated circuitry monitors the a.c. component of the probe signal. It meets all the features of claim 18. In the same way, Svarovsky describes a probe and circuitry which is suitable for use in a stack which has all the features required of claim 18. It anticipates. However it does not anticipate claim 1 because that requires the probe to be positioned in the stack, something which is not disclosed by Svarovsky. Similarly, claim 11, which is directed to the actual use of the probe in a stack is not anticipated.

(2) Obviousness over Svarovsky.

56. Independent of the attack based on anticipation, Mr. Birss argues that all the claims of the Rigby patent fail for obviousness. To do this I bear in mind Windsurfing International Inc. v. Tabur Marine (Great Britain) Ltd [1985] RPC 59. That case sets out a structured approach to the question of obviousness and sometimes simplifies analysis. In this case the inventive concept which is put forward as underpinning the Rigby patent is the application of triboelectric monitoring to the analysis of the flow characteristics of particles in a stack, that monitoring being achieved by analysis and use of the a.c. component only of the triboelectrically generated signal.
57. Mr. Birss' argument is that a man or team in the stack pollution control field looking at Svarovsky would realise that it was using triboelectric monitoring to test for changes in the mass flow characteristics of particle bearing air in a duct, that is to say in the ducts (19 and 27) in Figure 12. He would also appreciate, because it is a major part of the teaching of Svarovsky, that there are advantages associated with monitoring only the a.c. component of the signal from the probe. Svarovsky's teaching is general in the sense that it is not dependent upon any particular type of particles being monitored. It would be obvious to take the Svarovsky probe and the Svarovsky suggestion of a.c. monitoring and to use it to monitor the particle bearing gases in a different type of duct, i.e. a stack. Of course, Professor Stuart supported this conclusion.
58. Mr. Meade answers this by pointing out that a major feature of the Svarovsky patent is the teaching of a ratiometric type of analysis of two streams of particle-bearing air. It was not necessary for Svarovsky to assess the characteristics of the particles in one stream by itself, as would be involved in a stack. For example if the quantity of particles in each stream in Svarovsky went up by 50% it might be that the a.c. component of their respective probe signals would also both go up by 50%. Ratiometrically, nothing would have changed. This is satisfactory from Svarovsky's point of view since he is only concerned with seeing if there is a significant change in the balance of particles in the conveyor. If both large and small particles increase by the same amount, this does not suggest a fluctuation in the quality of the material flowing along the conveyor. As Mr. Meade points out, and Professor Stuart agreed, the use of the probe and a.c. monitoring in a stack would jettison this feature of Svarovsky's patent. Secondly Mr. Meade uses an argument which he also deploys as a major part of his defence against the attack based on the Beck prior art (which is considered at the end of this judgment). In essence it is that problems with electrical leakage between the probe and the pipework carrying the gas stream were by the priority date of the Rigby patent easily and better solved by use of a d.c. coupled amplifier and there were reasons why a man in the art would not wish to take the a.c. component of the signal alone.
59. I have come to the conclusion that the first of these points is of little weight. It is true that Svarovsky is adapting triboelectric techniques to a quite sophisticated application. He is not looking as mass flows for their own sake. But there is no doubt that he is determining mass flow rates (see paragraph 53 above) and is then using it for the purposes of comparison. In my view an ordinary skilled worker in the field looking for a method of monitoring the mass flow rates of particles in a stack would realise that he did not need the further sophistications of Svarovsky. Subject to Mr. Meade's other argument, which I will consider below, it would be obvious to such a worker that the Svarovsky equipment and approach monitored the mass flow rates of particles in a gas and that this is just what he needed for pollution monitoring purposes.
60. The second branch of Mr. Meade's argument is somewhat more complex. It can be divided into three parts: (1) by the priority date there were good technical reasons for preferring d.c. coupled amplifiers over a.c. coupled amplifiers, (2) there were good reasons for preferring to take the larger d.c. component than the smaller a.c. component of the signal and (3) there are reasons why a worker would not assume that the triboelectric effect would produce the same signal in the Svarovsky testing chambers as it does in a stack. I will consider each of these in turn.

(a) Preferences for a.c. or d.c. coupled amplifiers.

61. Professor Stuart's evidence was that there were technical reasons for preferring a.c. coupled amplifiers to d.c. coupled ones. He said that the latter were difficult to make. If he had the choice, he would use a.c. coupled amplifiers. Dr. Wakefield on the other hand said that there were technical problems with d.c. coupled amplifiers and that the criticisms levelled at them by Professor Stuart may have been true in the 1970's but were no longer true by the 1990's. By that date they were readily available off the shelf. Professor Stuart agreed that such amplifiers were available off the shelf by the priority date but he was unshaken in his preference for a.c. coupled amplifiers. Mr. Meade put to him in a most delicate manner that his preference for a.c. coupled amplifiers might well not be shared by younger electronics engineers. The evidence and cross-examination of the pros and cons of these two families of amplifiers took up a considerable amount of paper and time. In my view it is unnecessary here to rehearse the points made. That each type of amplifier has advantages and disadvantages is clear. However I have no doubt that by the priority date both types were readily available or could readily be made. I also accept that some older electrical engineers at that date may have been prejudiced against using d.c. coupled amplifiers. Such prejudice owed more to early experiences than the then current capabilities of these devices. I also accept that younger electrical engineers would not have been deterred from using d.c. coupled amplifiers and, in some cases, may have preferred them. The details of the merits and demerits of each are irrelevant. I have no doubt that by the priority date a worker in the field would decide what he wanted his circuit to do first and select the components to do it second. If the circuit demanded use of an a.c. coupled amplifier, he would have no difficulty in buying or building one. Likewise if the circuit demanded a d.c. coupled amplifier. If it made no significant difference which type of amplifier was used, his own personal preferences or what happened to be available from readily available stockists would determine his selection. The result is that I do not accept that by the priority date there was a general prejudice against d.c. coupled amplifiers as Professor Stuart's evidence suggested nor do I accept that there was at that time a general prejudice in favour of those devices as Dr. Wakefield suggested. Both types of amplifier were readily available tools. Their availability and the way in which they could be used would have been part of the common general knowledge of any worker in the field at the time.

(b) Preferences for taking the d.c. component in the probe signal.

62. It is central to the plaintiff's case that at the priority date there were good reasons why an ordinary skilled worker would not wish to jettison the d.c. component in the probe signal and look only at the a.c. signal particularly when considering the application of triboelectric monitoring to stacks. This is because the a.c. component in the signal would be much smaller than the d.c. component. Selecting the former and throwing away the latter would mean that most of the message-carrying information was being lost. If the total electric signal was small, throwing away most of it would be counter-intuitive. Furthermore, there were at the time good reasons to believe that this would be a particular problem where stacks are concerned. As set out in the primer and as mentioned above, the particle density in a conveyor would be expected to be much larger than that in a stack. On the basis of this it would be expected that the size of the total triboelectric signal obtained from a probe in a conveyor would be much larger than that obtained from a stack. The reluctance to throw away the large d.c. component would be particularly strongly felt.
63. The suggestion that the a.c. component of the total probe signal would be much smaller than the d.c. component is supported by an assertion to that effect in the Rigby patent:

"... it might appear to be disadvantageous to evaluate the much smaller amplitude alternating signal component involved ..." Page 3 line 13.

15. Furthermore Mr. Meade relies on one of the figures in the patent which shows a plot for the raw signal obtained from the probe:



16. Mr. Meade says that this depicts a very small a.c. signal (i.e. the wavy fluctuation) imposed on a much larger d.c. signal.

64. I do not think this material takes the plaintiff very far. The fact that Rigby asserts that the a.c. component is much smaller does not prove that it was much smaller nor does it prove that that would have been the expectation of anyone in the art. When assessing such a statement it must be borne in mind that Rigby was putting particular emphasis on factors which pointed away from his invention being obvious. The same point can be made in relation to the figure portraying the raw signal from the probe. No one suggested that it was meant to be accurate. It is a freehand illustration of the point Mr. Rigby was trying to make in the text. In fact, if the contents of the Rigby patent are to be looked at to see what the relative size of the components is, then it is unhelpful to Mr. Meade's argument. Figure 6 of the patent purports to show the a.c. and d.c. current outputs from a probe monitoring an air flow to which coal particles are added at a constant rate:



17. The lower plot (A) is the a.c. component. The upper plot (b) is the d.c. component. Although the former is smaller than the latter, it is only in a ratio of about 1:3. The Rigby patent explains that these are the plotted results of an experiment he had carried out.

65. Furthermore, even ignoring this material, no reasons were put forward why anyone in the art would have assumed that the a.c. signal would be significantly smaller than the d.c. component in the signal. As I have mentioned already, the precise mechanism by means of which these two components are generated was and still is unknown. There was, and still is, no known way of predicting their sizes either individually or relatively. It was not suggested that it was common general knowledge that the a.c. component was always much smaller than the d.c. component. Indeed even now there is no material to suggest that that is true. In my view a man in the art at the priority date would have no preconceived ideas as to the relative size of the a.c. and d.c. components of a triboelectric probe signal. If he had wanted to carry out simple experiments he may have done the same sort of test that Rigby had conducted with coal dust. He may have found that the a.c. component was smaller than the d.c. component. I was not persuaded that any skilled worker would be put off using the a.c. component because of worries about its relative magnitude. This would not have been a deterrent.

(c) Expectations that the signal from a conveyor would be much larger than from a stack.

66. Once again, what is important is the view or expectation of a man in the art at the priority date. Mr. Meade argues that he would have expected the signal from a probe in a stack to be so much lower than from a conveyor that he would set his face against discarding any of it. He would not want to amplify just a part of the signal. I accept that a man in the art would have thought it quite likely that the signal from a stack probe would be smaller than that from a conveyor containing a much higher density of particles. However he would have no way of knowing how much smaller the signal would be nor would he have any way of knowing whether the a.c. component in it was as much diminished as the d.c. component because again there is no precise or quantitative basis on which such a conclusion could be based. I do not accept that he would have assumed that the signal would be so small that it was not worth trying to look at the a.c. component in it alone. If he had technical reasons for looking at that component, he would have tried to do so. Consideration of the overall size of the signal might have made him think that the a.c. component might be too small to measure but I am not persuaded that it would have put him off trying it.
67. Furthermore, this argument has no application to Svarovsky in which the triboelectric probes are not inserted into the conveyor at all but into ducts containing much lower density of particles. As noted above, no limits are placed on the particle density to be monitored by apparatus made in accordance with the Rigby patent. But some idea of what is involved in pollution monitoring can be obtained from one of the documents introduced into the case by the plaintiffs. It is an a publication called the Handbook of Air Pollution. It is dated 1984. The page relied on has a table which sets out typical particle densities of exhaust gases from various types of industrial plants. To show the effectiveness of the filters, the density of the particles in the air before filtering is also set out. In other words the table shows what the particle density in the plant exhaust would be if there is a total failure of the filtering system. Examples given include the following:

Industry Application	Particle Loading before filters (g/m3)
Copper Ore Plant Crusher	5.42
Feldspar Pebble Mill	13.8
Fuller's earth Fluid energy mill	2.38
Fuller's earth Raymond mill	11.99
Industrial boiler	1.24
Iron ore plant fine crusher conveyor	2.99
Kaolin Raymond mill	10.36
Kaolin Roller mill	4.03
Limestone primary crusher scalping screen and hammermill	2.84
Municipal refuse boiler	1.14
Utility boiler pulverized fired	5.7
Utility boiler stoker fired	4.6

68. The claims in the Rigby patent would cover a monitoring device put in the stack of a feldspar pebble mill. It could be set to trigger an alarm at, say, 5.00 g/m3. This should be compared with what is being detected in Svarovsky. The latter states that the separator (13 - See Figure 11 above) operates with a solids concentration of less than 5 g/m3. That then produces two outflows, one of which is relatively large containing a relatively low concentration of solids and the other of which is relatively small but with a high concentration of solids. Each is monitored by the triboelectric probe. It is apparent that the Svarovsky apparatus is designed to operate at particle density levels within those covered by the claims of the Rigby patent.

(d) The signal from the Svarovsky testing chamber would be expected to be larger than that derivable from a stack.

69. This point is based on the fact that in Svarovsky compressed air at 70 p.s.i. is shown as being injected into the sample of particle-bearing gas from the conveyor (see Figure 11 above). The diluted particle flow is then described as turbulent. The plaintiff suggested that this turbulence would be significantly different to the flow conditions in a stack so that a worker would not assume that the same level of signal would be obtained from the probe if Svarovsky was applied to a stack. In my view there is nothing in this point. I accept that a worker in the field might expect the turbulence in the Svarovsky test chambers to be different to that in a stack. There is nothing to suggest how different it would be and, more importantly, there is no material which shows that a skilled worker would assume that the differences were so great that, whereas a useful and useable signal could be obtained in the Svarovsky chambers, a useful and useable signal could not be obtained from a stack. Once again, the mechanism of triboelectric charging is not known and I do not accept that a skilled worker would assume that he would obtain insufficient signal from a stack. On the contrary, in my view he would have thought it likely that a useful signal would be obtained.
70. I have come to the conclusion that a skilled man thinking of developing a pollution monitor for use in a stack and presented with Svarovsky would read and appreciate that he was using triboelectricity generated in a probe to measure the mass flow of particles in a gas, that there were valid reasons for using a.c. coupling and that the levels of particle density detectable in the system were similar to those with which he might be interested. It would be natural and non-inventive of him to try to use the same concepts in a pollution monitoring device for use in a stack. There were no "lions in the path" which would have deterred him from trying this route to the solution of his problem. All the claims in the Rigby patent are invalid for obviousness over this piece of prior art.

(D) Beck (UK Patent No. 1,335963)

71. This is concerned with monitoring the detection of electrostatic charges in flowing materials. It is particularly concerned with the build up of such charges in pneumatic conveyors and the risk of fire or explosion which may be caused by them. Beck's solution is to use a device of the following general arrangement:



72. Particle bearing gas flows down the conveyor (1). Into the tubular wall of the conveyor is built a curved electrode (2) which is insulated from the rest of the wall by an insulating material (3). Charge built up on the electrode by triboelectricity is fed to a sensing circuit (4). This incorporates an a.c. coupled amplifier (8). The output is fed to an alarm or control circuit. Beck discloses that the signal from the electrode consists of an a.c. and a d.c. component. The magnitude of the former is about one third that of the latter. Beck explains that d.c. sensing is not employed because of signal errors which will be caused by atmospheric humidity and the state of cleanliness of the insulation of the measuring electrode. It also states:

"This alternating component has a character similar to band-limited white noise, and it can be shown that its magnitude is dependent on the mean value of the electrostatic charge concentration and the mean velocity of the flow. Where the latter is constant, as will commonly be the case in situations in which such an apparatus is used, the magnitude of the alternating component will give a direct measure of the mean value of the electrostatic charge concentration." Page 1 line 71 et seq.

73. Because Beck does not employ a probe inserted into the duct, there is no question of this being used for an anticipation attack. It is only advanced in support of an argument of obviousness. However for many reasons it appears to me to be further away from the Rigby patent than Svarovsky. Unlike the Svarovsky patent, Beck is not measuring particle flows of the same sort of density as may be of significance in a stack. In fact it is not measuring particle flows at all. Although it records that the a.c. component is dependent on the mean velocity of flow, it does not actually suggest that it can be used to measure it. If it is not obvious to make a monitoring apparatus within the claims of the Rigby patent from Svarovsky, it cannot be obvious to do so from Beck. On the other hand if it is obvious, Beck adds nothing. In these circumstances no purpose would be served by considering this as an independent starting point for attacking validity.
74. For the reasons set out above, I have come to the conclusion that the Rigby patent is invalid. Before ending this judgment there are two matters I wish to add. First I would like to express my gratitude to both counsel for the way in which they conducted this case. Their advocacy was of the highest order. Secondly I would like to express my thanks to Mr. Summerhayes who was made available by the Patent Office and sat with and assisted me during the early part of the trial.