Technical evaluation of a Hubbard electrometer, the device employed for Scientology auditing.
By © Leonardo Serni, Febbraio 2009.
Scientology considers the E-meter "an electropsychometer, an instrument measuring emotional responses through the small electrical impulses produced by thought. The E-meter tells what the preclear's mind is doing when he's made to think of something. The E-meter gives the reading before the preclear is aware of the data. It is therefore a pre-conscience electrometer. It runs a small current through the preclear's body, and this current is influenced by mental masses (memories - N.d.R.)" (L. Ron Hubbard, "Dianetics and Scientology Technical Dictionary"). A Preclear is anyone undergoing the process devised by L. Ron Hubbard.
According to Scientology, an E-meter "is a device latching on to the electronic structure of the reactive mind (the body of mental aberrations - N.d.R.). The reactive mind may be considered an electrical energy field, and when a current is run in the vicinity, it alters the flow of this current. [...] For the E-meter to be read, the small flow of electrical energy through the preclear must be kept constant. When this small flow is reduced by an increased resistance, the needle will move on the quadrant. This happens because the preclear is gathering mass. This is effective mind mass (condensed energy), mass that opposes resistance to the E-meter current flow." (On the side of this explanation, there is a picture in which a preclear recalls a motorcycle accident, where the "mass" causes an increase in resistance). "We see then that the tone arm (the potentiometer on the E-meter, NdR) measures the density of the mass present in the preclear's mind. This is real mass, not imagined mass, and may be weighed [...] It has been proved that the mental energy, such as that contained in a mind image, and the electrical energy of the Earth or the power company, only differ in wavelength. Therefore whenever the preclear thinks to something, relives an episode of his past, or moves around some part of his reactive mind, an electrical alteration will take place". (L. Ron Hubbard, "Understanding the Electrometer")
The external controls are two continuous rotors, one rotary selector marked "32 - 64 - 128" and a "ON - OFF - TEST" switch. There are also two small 7-segment displays with three control buttons not unlike those of the first digital clocks.
On the back there is a voltage selector type USA/Europe (230/115 VAC), whose purpose is unclear given that the device is powered by a DC jack labeled "21 VDC" (weirdly, the power unit, which they tell me is the original one and looks like it, is labeled "12-18V".
Practically all the controls are connected to the lower board (in yellow), which will hence be called METER board, and only a trio of wires goes from an additional, electrically insulated module of the biggest rotary control to the upper board (in green), which we will call the DISPLAY board and has an "output" in the two digital displays, while the big white quadrant (whose round back, in white, is visible beneath the DISPLAY board) is connected to the METER board.
The voltages selector (photo here on the side) is revealed as perfectly useless, and not connected to anything. The claim found on some Internet sites, that selecting the wrong voltage may "fry" the device, while absolutely true for most electric and electronic device (therefore, NEVER try it, because among the other things you risk a fire as well as electrocution), in this case is manifestly unwarranted.
It is the writer's opinion that some earlier "Mark", for example, say, the Mark V, was probably wall-powered; using the Mark V chassis to assemble the Mark VI, if the useless voltage selector had been eliminated, there would have been a leftover hole. Usually such holes are covered by appropriate plastic blanks, available in several colours (in this case, black). Possibly for "tradition" or because the blanks weren't available and there was a shipment date to be met (it happens more often than you think; it's called JIT, Just In Time production), the builders had rather leave the old component in place, without connecting it (the connectors have been torn away by sheer force with pliers). ).
The two boards DISPLAY and METER are so different that they might well have been designed and built fifty years apart. What first strikes even the eye of someone unfamiliar with electronics is that while the DISPLAY board is built with several integrated circuits, those small black spidery things, the METER board is a much more haphazard collection of analogic components, back to good old germanium transistors (still in use, however, and much valued by connoisseurs, who boast of them amplifying to a "mellower sound". For the same reason, by the way, there's people extolling the virtues of thermoionic valves). ).
Other very apparent details are the circuit board designs, whereby the DISPLAY board sports straight, evenly spaced traces clearly designed by a PCB layout program, with all components orderly placed on a rectangular grid; while the METER board has curved traces, with components disposed every which way. While the circuit has been etched with a professional printed circuit kit, the design is clearly hand-traced, "old times".
The same can be said for the assembling technique: the DISPLAY board shows professional in-line solder points that smack of a wave fusion solderer (so-called bulk wave soldering machine, nowadays employed by any electronics laboratory above the artisan level), while the METER board has clearly been soldered by hand (you can hardly find two solder points alike) by an amateur or really hasty technician: a couple are reflows over cold soldered joints, many are "fat" - such as those of the lower right rotor - and some others bear witness to a... let's say uncommonly generous use of soldering flux.
A closer look reveals that also the etching of the METER board is of inferior quality, and many traces are oxidized and at risk of detachment (one, on the border, looked on its last leg).
Also, just like in the paintings of the old masters there can be seen the so-called "pentimenti" where the master changed idea or added a retouch, so in the METER board are visible some "enhancements", for example two resistors looking newer than the rest, soldered on the trace side (to avoid drilling the required vias) with four of what must be the fattest soldering points in history (photo on the side).
The DISPLAY board sports a bank of MC-145-238 integrated circuits and one MC145-20B. From the manuals available on the Internet it's a decoder to drive a 7-segment display. An LM324N op-amp, a NAND bank SCL4011BE, and an all-in-one controller for digital clocks complete the picture; the DISPLAY board is again divided in two parts, one being nothing more than a compact clock (or chronometer), the other a numeric state display (the MC145-20B is a decoder for the BCD, Binary Coded Decimal, numerical format).
Without physically disassembling the board (and making it useless), a passive examination of the MC145-20B input and the components placed in the vicinity suggests the whole thing is a counter or incremental toll. Every time the voltage drop on the input resistance goes down, the variation is divided in a certain number of "steps" and their number converted to binary, added through the NAND bank, converted in BCD by the MC145-20B and finally sent to the display by their MC145-238 driver. Depending on the configuration, with such a circuit one could
Since I was told that each "Mark" is touted as more or less a complete replacement, mandatory to all intents and purposes, of its predecessor, sort of like "If your PC hasn't Windows Vista on it, you're a pansy" - and I use with great satisfaction XP Pro SP3, maybe making myself a pansy - I'd like to explicitly comment this thing: the "enhancement" introduced with whatever Mark it was when the DISPLAY board supplemented the METER board (and the boards' designs make me drop-dead certain that it didn't happen straight from Mark One) is without any advantage whatsoever (except having a clock and a counter, which maybe come in handy). It's not at the level of Vista vs. XP (since Vista has quite a lot of features that are really improvements and innovative and worth it - well, after one has sort of acclimatized); it's more along the lines of supplying my faithful XP PC with a replacement keyboard that glows in the dark and saying that now it's a completely different PC, stabler, more efficient and, why not, also faster.
The METER board is also made up, "historically" I'd say, of two sections. The first, on the right, is a Wheatstone bridge allowing for resistance measurements, but supplemented by an amplifying circuit made up of three (and the design looks like a fourth was added later) 2N1303 germanium transistors. A reference resistor bank is selected through a rotary control; the circuit is apparently calibrated using a bank of five (!) multi-turn precision trimmers that seem to control amplification.
While there do exist simpler circuits to obtain the same effect, even without using the more "modern" op-amps of the DISPLAY board but only the good old germanium transistors, the design choice is defensible. But it must be pointed out that this more complex choice will result in a much greater "drift", that is, the readings of two initially identical e-Meters will tend to diverge over time, because the rotary selectors wear out, the components themselves age, traces will oxidize, the "electrode" tin cans oxidize even more, and why not, because there are five trimmers (potential sources of drift) where it's very likely that a single one might have been enough.
Luckily, since what we're interested in is the differential behaviour (i.e., how does the body resistance vary over a short time), all these errors will show as a change in range. So it will be enough to re-center the signal to get again the expected reading (i.e., the needle behaviour).
Even so, should anyone ever try connecting two e-Meters in parallel to the same tin cans held by the same candidate, even with two identical e-Meters built the same day in the same lab, I think he'd be in for a bit of a surprise..
Apart from "virtual" costs incurred through adoption of copyrighted advanced technological solutions (of which, apparently, no trace is to be found) or for especially accurate technical implementation (and here too, with all the good will possible, it's not easy to see any particularly remarkable degree of proficiency in the DISPLAY board production - a good product, to be sure, but hardly remarkable - and it's pretty much impossible to see it in the METER board's shoddy workmanship), the main costs should be due to design, implementation and materials.
The DISPLAY board is a professionally constructed board and those services that produce (and often design or co-design) such boards usually have "all included" prices. This specific board could cost some 50-80 EUR apiece for small lots including components, assembly and testing. The initial design could have costed, at a generous guess, one thousand EUR; but it would have been more than enough to pin an announce on the appropriate bulletin board and lots of electronic engineering students would have been more than willing to devote two or three evenings to the project.
It is true that such boards, very specialized boards for alarm systems and so on, are indeed sold for 300-500 EUR or more, exploiting what is to all intents and purposes a monopoly (when you fork out EUR 2,000 for an alarm system and the mainboard goes south, you either cough up those 500 EUR or you have to change the whole central unit, and maybe the whole system altogether): but the board itself costs, to the seller, 50 EUR, tops. Often they're partially recycled or "refurbished" boards and don't reach a third of that figure.
Considering that in the past devices such as the DISPLAY board required machinery that at the time was proportionally more expensive (the old "lambda wave" soldering stations used for the first printed circuits are almost not even used anymore; soldering stations will use "double waves", or even "variable waves" adapted to the components to be soldered), let us recap and say we may try and justify a cost of 150, 200 EUR for the DISPLAY board.
The METER board is actually a Wheatstone bridge - cum - amplifier, and you can find dozens of such circuits on the Internet (and on the journals, hobbist magazines, etc). The components were and are very easy to find - "old" components obsolesce more slowly than "modern" ones, to the point that there's still demand for germanium transistors, and the analog components are still very much in use - and they may cost from 10 to 30 EUR depending on the supplier. Assembly and printed circuit production may cost more or less the same. A relatively expensive component (anywhere from 20 to 50 EUR) is the needle quadrant, which may be also quite difficult to find, modern components tending to be smaller and more compact.
All in all, and excluding the designing costs, a single piece test run, built from scratch, will surely never exceed 400-500 EUR, and this is an "all-top" figure: for runs of anywhere more than a few pieces, the cost per single piece will easily drop to maybe 100 EUR (also ammortizing fixed costs such as design). Also if you just replaced the less common components with their off-the-shelf equivalents and employing a standard container and display, the cost apiece will surely at least halve.
If one replaced all the components with "modern" equivalents, with a custom multi-resistor bank, a standard chassis and multiple display, one would get something not too different from the picture below, that would cost (again, one-piece runs) around a hundred EUR, and would probably be much more precise and durable.
Actually, I fancy that, should anyone decide to build such a thing nowadays, it would be much more practical to build an audio converter capable of transforming ohmic resistance into sound frequencies. Almost any modern computer is equipped with a mic input capable of sampling the 1-15 kHz range with a 12-bit precision (4-5 Hz or better!) with a frequency of 22050-44100 samples per second. From there on, it is straightforward to implement the whole emeter logic in software. There're guys that from a microphone input have built an oscilloscope, so a simple Wheatstone bridge with band selection and amplifying should be a piece of cake. zeitnitz.de/Christian/Scope/Scope_en.html. .
First of all it must be said that this approach is conceptually wrong, because not only is the electric resistance maximal (and therefore the variations percentually minimal) between the palms of the hands, where evolution - or intelligent design - has placed pads of nonconducting, grippy horny tissue; but the resistance is influenced by the efficiency of the contact itself, and therefore from how much a person grips the electrodes. The basic muscle tone in the hand muscles will therefore show as a difference in signal amplitude. How to discriminate a tiny variation in an atonic person from a wide variation in a well-toned (or just tense) person - when the muscle tone is a consistent, but unknown, part of the very signal that we are going to measure? Also, a steady pressure variation during the measurement may produce a completely spurious "signal" which could be easily avoided by using appropriately designed contact pads (not coincidentally, the type employed in all medical instruments - see on the right).
At this point, from the theoretical point of view, it must be also noted that of those "signals" that the device pretends to measure, clear of the various disturbances listed above and which are far from secondary, signals that depend from deep muscle tension, surface muscle tension and haematic conductivity, none is completely involuntary.
The "claim" of the device - to be able, in the hands of a qualified operator, to give indications on the interior status of a subject - while theoretically (barely) defensible for devices of this kind - (trans)cutaneous ohm-meters - is completely without basis in practice for this model of device; the main reasons being, from a practical point of view, the measurement being taken not in the proper places and not in the proper way; from a theoretical point of view, the fact that the interpretation is unreliable to such a point that it cannot be conceivably compensated for by an operator however qualified.
It may well happen that an operator with sufficient experience and sensibility may intuit the interior status of the subject in a manner completely independent of the E-meter reading, and therefore be able to "interpret" the subject's ohmic behaviour in such a way as to lend authority to this interpretation. Unfortunately, it is just as likely that an operator simply delude himself and buttress in good faith a completely wrong reading.
Moreover, the device has a price tag which appears grossly exaggerated, unless one wants to invoke some kind of placebo effect induced by the exorbitance of the request - it was Woody Allen, if memory serves, to say that the price of a psychiatry session is a fundamental part of the therapy.