FR2459470A1 - DEVICE FOR MEASURING THE DEFLECTION OF A LIGHT BEAM - Google Patents

Abstract

A REFRACTOMETER INCLUDES IN PARTICULAR AN OPTICAL BENCH 10 WHICH IS PLACED INSIDE AN OVEN AND IS CONNECTED TO A LIGHT SOURCE 16 BY AN OPTICAL CABLE OF FIBERS 18. THE INLET TUBE 20 AND THE OUTLET TUBE 22 OF THE SAMPLE LIQUID WHICH IS INTRODUCED INTO THE REFRACTOMETER ARE THERMALLY COUPLED SO THAT THE TEMPERATURE OF THE FLUID ENTERING THE SAMPLE CELL IS AS CLOSE AS POSSIBLE TO THE TEMPERATURE OF THIS CELL.

Description

The present invention relates to beam deflection instruments

  light, like refractometers. In these instruments, a light source typically directs a beam towards a detector (for example a photoelectric cell, by making it pass through a sample cell. A variation of a certain physical quantity (for example the refractivity) displaces the beam light relative to the detector. One difficulty is that a change in the position of the light source (resulting for example

  filament movement in an incandescent lamp) cannot

  cannot be distinguished from a change in the physical quantity being measured, since the two changes produce movement

  of the light beam relative to the detector. The plaintiff discovered

  green we could stabilize the position of the light source in relation to the area (for example the sample cell) in which the measurement is made, by modulating the position of the light beam (i.e. by making it perform an alternating cyclic scan) with an amplitude chosen in advance which is independent of the movement of the light source or of other movements of the light beam of which one is not master (such as the movements which result from thermal vortices). The movements of the light source only vary the phase of the modulation (for example the instants at which the cyclic scanning begins and ends), and the electronic circuits which process the output signal of the detector can easily be designed.

  so as to ignore these phase changes, such as determining

  undermining the average position at which the beam reaches the detector.

  The invention makes it possible to greatly reduce the sensitivity to movements of the source and it can be implemented economically in the preferred embodiments. A second aspect of the invention

  uses an unfocused light beam which is directed perpendicularly

  rement to the direction in which the light beam moves during a measurement. The light which comes from many points of the light source can thus be distributed uniformly in this

  unfocused direction, which makes the measurement insensitive to varia-

  the light source brightness in this direction.

  The invention relates to devices, the operation of which is based on precise measurements of the variations in the refractivity of a circulating fluid, such as for example in liquid phase chromatography. Because temperature affects refractivity, the temperature of the circulating liquid must be carefully regulated. In refractive index detectors that are used in liquid chromatography, we typically compare in a cell

  the refractivities of a sample current and a reference current

  rence. One technique for equalizing temperatures in the sample and reference streams in the cell is to enter the fluid into the cell after passing it through sample and reference inlet tubes which are placed so that there is heat exchange between the tubes and with a large metal block. In the ideal case, the two currents are at equal temperatures before entering the cell. The Applicant has discovered that the temperature can be better controlled by carrying out

  a heat exchange condition between the sample input current

  lon (to the cell) and the sample output current (from the cell). The equipment required is simple and economical. It is thus possible to obtain an excellent temperature equalization

  sample and reference fluids, and heating and cooling

  very quick setting up of the device.

  The Applicant has discovered that it is possible to increase the light flux passing through the cell of a refractometer by placing a layer

  reflective integrated in the cell, in the cell or on the latter

  so as to reflect the light beam to send it back through the cell. The problems of parallax between the chambers of the circulation cell and the reflecting surface are reduced. There is

  less area exposed to ambient conditions, which reduces the per-

  your doesàd accumulation of dust and surface reflection. Manufacturing is simplified by the fact that the number of elements required

is reduced.

  The applicant has also developed an improved technique

  which is relatively insensitive to variations in the

  light density and which, in preferred embodiments, sup-

  takes precedence over optical zeroing. Generally, we introduce in the

  system output signal a main term which depends on the different

  between the two optical measurements and an offset term which

  depends on at least one of these measures. Any change in intent

  The luminousness affects the main term and the offset term almost equally, which maintains the precision of the zeroing. The Applicant has discovered that it is possible to increase the light flux passing through the cell of a refractometer by introducing into the cell a curved surface integrated into the cell making

  lens function to focus the light beam. The problems

  the parallax between the chamber of the circulation cell and the surface acting as a lens are reduced. There are fewer surfaces exposed to ambient conditions, which reduces losses due to dust accumulation and surface reflection. The

  manufacturing is simplified by the fact that the number of elements required

  saires is reduced. According to one aspect of the invention, an opaque mask is introduced inside or in the circulation cell.

  The invention also relates to circuits which are intended for

  analog-digital and digital-analog version and to opera-

similar.

  The digital circuit operates faster than the analog circuit, and the time required to complete each sequence of operations is therefore defined in. a large measure by the response time of the analog circuit. For example, if the analog circuit includes a filter to smooth out ripple and noise, it may take a few seconds for the analog circuit output signal to reach a new value. When these few seconds are multiplied by the number of repetitions of the sequence, the waiting time lost can be considerable. The Applicant has discovered that the time lost waiting for the reaction of the analog circuit can be shortened by dissipating (for example using a switch connected in parallel) the load from the charge storage device (for example a capacitor) which is part of the analog circuit. After dissipation, the analog circuit and the charge storage device can react normally.

  The invention will be better understood on reading the description

  which will follow of an embodiment, given without limitation.

  The following description refers to the accompanying drawings in which

  Figure 1 is a perspective view of an embodiment of the invention; Figure 2 is a partially sectional representation of

this embodiment.

  Figure 3 is a section along line 3-3 of Figure 2 which

  shows the photocell end of the optical bench.

  Figure 4 is a section along line 4-4 of Figure 2, which shows the end of the bench side circulation cell, as well as the outer insulating screens and cylinder, with the heat shield

  internal / light deflector 77 removed.

  FIG. 5 is a section along line 5-5 of FIG. 2,

which shows the light source.

  Figure 6 is a diagram of the heat exchanger piping. 7a and 7b are respective sections of the heat exchangers

  sample and reference heat.

  Figure 8 is a section along line 8-8 of Figure 4

  which shows the structure of the circulation cell.

  Figure 9 is an elevational view of the rear surface of the flow cell, in the direction indicated in 9-9

in figure 8.

  FIGS. 10a and 10b are diagrams of the optical paths in the embodiment considered; Figure It is a block diagram of electronic circuits

  which process the photocell output signals.

  Figures 12a and 12b are diagrams of electronic circuits.

  ques which cancel the photocell output signal and process the canceled output signal to achieve display and integration. We will now consider FIG. 1. An optical bench 10 is supported inside an oven on 4 insulating feet 12 which are fixed to the lower horizontal wall 14 of the oven. The

  light source 16 for the bench is placed under the bench and outside

  laughing of the oven. A frame of optical fibers 18 routes the light from the source 16 to the bench. The sample liquid which comes from the outlet of a chromatography column (not shown) which is inside the oven enters the optical bench through the inlet tube 20 (with an inside diameter of 0.23 mm), and it exits through the outlet tube 22 (with an internal diameter of 1 mm). A small diameter sample inlet tube is used to minimize band spread in the chromatogram. Similarly, the reference liquid enters and leaves the bench through the inlet tube 24 (with an inside diameter of 0.5 mm) and through the outlet tube 26 (with an inside diameter of 1 mm). The four tubes are made of stainless steel and have

  1.6 mm outside diameters. The outlet tubes have diameters

  very internal superior to those of the inlet tubes to reduce the back pressure and its effect on the refractivity. The outlet tubes 22, 26 are connected together downstream from the optical bench to equalize

  sample and reference pressures in the circulation cell

  tion. Electrical wires 28 which come from photocells

  52 (figure 2) connect the bench to the processing circuits

  which are shown in Figures II, 12a and 12b.

  We will now consider Figures 2 to 4 which show that the optical bench 10 is constituted by an inner cylinder 32, in

  which we pass a light beam B, and through an external cylinder

  concentric laughter 34 which defines an insulating air space 36. Two flat screens 38 (Figure 1) delay the thermal radiation towards and from the bench, and they act as legs (Figure 4) to support the cylinders on feet 12, by means of 4 bolts 40. End caps 42,44 close each end of the outer cylinder 34 and end caps 46,48 close each end of the inner cylinder 32. The cap end 46

  supports the extended outlet 50 (1.3 mm wide by 8.9 mm high

  fiber optic cable 18, as well as the photocell

  electric 52. The end cap 48 supports the circulation cell

  lation 54 by means of cell bridges 56,58 which are fixed to the cap and are fixed to each other by screws and

  epoxy. The sample inlet and outlet tubes 20, 22 terminate

  are located at the bridge 58, while the reference tubes 24, 26 are

  end at the trigger guard 56. A cavity 60 in the cap of ex-

  end 48, behind the bridges, contains about 4 turns of the sample inlet tube 20. Notches 62, 63 which are formed in the

  inner cylinder 34 define inlet passages for the

  bes. The end caps, cylinders and screens are all-

  aluminum, to accelerate the heating of the bench, while insulating

  the bench thanks to the air space 36 between the cylinders.

  Now considering FIGS. 8 and 9, it can be seen that the circulation cell 64 comprises two hollow chambers 70, 72 which are respectively intended for the sample liquid and for the reference liquid. Each chamber (Figure 8) has a triangular cross section (approximately 45 x 45 x 90 degrees) and 1.57 mm for each

  short side, and it is connected to the inlet and outlet tubes res-

  pective by internal passages 73. The height (or dimension

  vertical in Figure 9) of the chambers is approximately 13 mm. We manufacture

  as the circulation cell by melting together, without adhesive, elements of borosilicate glass. Teflon joints 75 which the bridges compress against the cell seal between the sample and reference tubes, on the one hand, and the internal passages 73 of the cell, on the other hand. The front surface 74 of the circulation cell is rectified to form a unitary lens which has a curvature in horizontal planes, but not in vertical planes. The back surface of the cell

  has a reflective surface coating 78, in gold, to constitute

  kill a mirror reflecting light backwards towards the photocell 52, through the chambers 70, 72. The line

  focal length of the lens is positioned at the photocell

  52 and the distance between the mirror 78 and the photocell

  electric 52 is about 15 cm. As shown in Figure 9, the mirror coating 78 is limited approximately to the area which

  located directly behind chamber 72, to limit principally

  light reflection through the triangular chambers

  laires. The other light is absorbed by a black epoxy coating

  76 which is applied to and around the mirror coating

  deny. The mirror coating is slightly larger than the cham-

  to account for variations in the internal size of the chambers

  bres 70, 72. The coating stops before the top of the chambers 70,72 (Figure 9) so as not to reflect the light passing through the

  top of the chambers, where bubbles may form.

  To reduce heat transfer by radiation and convection to the circulation cell from inside the

  optical bench, a blackened disc 77 provided with a rectangular aperture

  the area 79 for the light beam (just large enough for the circulation cell to be exposed to the beam) is placed in front of the circulation cell. The disc also functions as

  light deflector and it is tilted by 100 (see Figure 2).

  The sample liquid and the reference liquid are introduced

  duits in the circulation cell passing through counter-current sample and reference heat exchangers, 90, 91 (FIGS. 1, 6 and 7a, 7b). Each heat exchanger is formed by fixing the corresponding inlet and outlet tubes together inside a tubular jacket. Each pair of tubes attached together is then routed along a multi-zone path that begins at

  outside the bench and ends in the circulation cell.

  As shown in Figures 7a, 7b, the heat exchanger of heat

  tillon 90 is produced by placing the tubes 20, 22 inside a tubular copper braid 80, by shrinking under the effect of heat a polyethylene tube 82 on the outside of the copper braid , and by filling the interstices between the braid and the inlet and outlet tubes with epoxy material 84 (Stycast 3051), which has a low viscosity and is moderately conductive of heat. Reference tubes 24, 26 are fixed without copper braid

  by inserting the tubes inside a Teflon tube and filling

  the tube with the same low viscosity epoxy material as for

  the sample tubes. The braid is removed because the trans-

  heat of heat which is necessary for the reference is lower, because there is no circulation during the measurement but only one

purge between measurements.

  We will now consider Figure 6 which represents schematically

  the path to several zones followed by the heat exchangers. The first zone for the sample heat exchangers as a reference begins outside the optical bench and extends along the outside of the bench, in longitudinal direction, between the external cylinder 34 and the screen. 38 (the total length of the area is about 20 cm). The sample heat exchanger 90 is placed on the side of the bench which is closest to the center of the oven, where the temperatures are best defined. At the end cap 42, the two heat exchangers rotate at 1800 and enter the space 36 between the cylinders 32, 34, passing through a slot (not shown) in the end cap. The second area for the sample and reference heat exchangers extends along the space 36 (its total length is about 18 cm). The reference heat exchanger enters directly into the cylinder 32 from the space 36, passing through the notch which is formed at the end of the cylinder. The inner cylinder 32 as well as the reference inlet and outlet tubes go directly to the circulation cell

  54 via the trigger guard 56.

  The sample heat exchanger 90 continues on its way by entering a third zone beyond the end cap, in which it is curved to form a coil 92 which comprises 4 turns (the total length of the turns is about 61 cm) which are

  positioned in the space behind the end cap

  mite 48. The last turn is adjacent to the rear face of the hat

  end. From the coil, the sample heat exchanger

  lon enters the cylinder 32 through the notch 33. Inside

  After the cylinder, the sample outlet tube 22 is directly connected to the circulation cell. The sample inlet tube 20 is rolled up to form another coil 94 (with a total length of about 30 cm) before entering the circulation cell. The coil 94 is positioned in the cavity 60 and it is coated with a

  heat conductive epoxy material to ensure good conduct-

  activity with the end cap and cell jumpers.

  Considering Figures 2 and 5, we see that the light source

  neuse 16 includes an incandescent lamp 100 (PhUiips lamp type 6346, base H 3, 6 V, 55 W operating at 4.8 V) with a filament 101 directed vertically; a concave mirror 102 which concentrates the light (glass coated with gold) and a rotating prism 104. The prism is made to rotate at approximately 50 to 60 rpm along an axis parallel to the axis of the filament, at l using an asynchronous motor with a phase shift ring 106, which

  also drives a fan 108 which provides cooling air

  flashing. The prism 104 is approximately 9.4 mm high, made of glass and has a rectangular cross section. Two opposite surfaces 110 of the prism are transparent and have a width of about 7.6 mm. The other two surfaces 112 are opacified with

  opaque silicone rubber, and have a width of approx.

  6.4 mm. The optical fiber inlet 114 is round (with a diameter of about 3.8 mm) and it is placed on the side of the prism which is opposite the lamp. The mirror 102 is placed so as to form an image of the filament 101 on the face of the inlet 114. The emission peak of the lamp 100 is in the near infrared, at a length

about 1000 nm.

  The fiber optic cable 18 is internally subdivided

  in sub-beams and the sub-beams are intentionally dis-

  ordered at an entry to randomize the light path between

entrance 114 and exit 50.

  The photocell 52 has two photocells

  double triangular and adjacent voltaic 180,182 (silicon with

  gold connections) which are arranged so that their large dimensions

  sions are directed horizontally, that is to say in the direction of movement of the light beam. Each triangle has a length of about 3.8 mm and a height of about 1.3 mm. The spacing between the triangles is approximately 0.20 to 0.25 mm. The shunt impedance is maximized at operating temperature (around 1500C) and it

  the same is true for sensitivity to long wavelengths.

  The oven in which the bench is located is heated by

  electric resistance elements with proportional control. Inside

  In the oven, the temperature variations can reach 5 to 7t from one point to another, but they are much smaller (for example 0.30C) at the same point over time. We vary the

  duration during which the resistance elements are energized pro-

  proportionally to the difference between the actual temperature of the oven and the desired temperature, and proportionally to the integral of this difference. To make the oven temperature less sensitive to

  variations in the voltage of the AC grid, we also

  vary the duration inversely proportional to the square

  of the network voltage, because the heat given off by the elements

  is proportional to the square of the grid voltage. The elements are controlled by thyristors and they are not energized or

  de-energized only when the current passes through zero.

  FIGS. 11, 12a and 12b show the electrical circuits which process the output signals from the photoelectric cell 52. The

  figure It represents all the circuits in the form of a syn-

  optical. The input signals 118 which come from a panel (for example

  example the gain of the recorder) are applied to a central processor

  tral 120. The central processor triggers the automatic electrical zeroing of the photocell output signals and it sends signals to the circuits which are represented in FIG. 32b, via the buffers 122 and the gain flip-flops 124 , to adjust the gain for displaying the chromatograph on a

  recorder. An analog data acquisition voltage is con-

  vertically in digital form and it is sent by the processor to a display panel 126 via the input buffer memories.

  Figure 12a shows the electrical zeroing circuits.

  than. The current output signals (alternating signals) of the

  The photoelectric cells 180, 182 are brought by a screened cable to the current-voltage converters 130. The alternating voltages A, B produced by the converter are summed and amplified with a gain of 2.2 by the amplifier 132 to form the expression -2, 2 (A + B) called SUM. Amplifier 134 subtracts voltage A from voltage B and adds the difference to the sum of three voltages: SUM, ZERO END, and ZERO GROSSIER. These last two voltages are obtained by multiplying the term SUM by a negative scale factor. It is thus possible to express the output signal of the amplifier 134 (OUTPUT SIGNAL SET TO ZERO) in the form: [B - A] - 2, 2 [0.33 - 0.67KC - 0.0033KFi [A + B ] where KC is the coarse zero scale factor and KF is the fine zero scale factor. Scale factors Kc, KF are fixed

  between about zero and about one by the digital circuits of the sub

  set 136 each time a signal is applied to the conductor AUTOMATIC RESET ORDER. We normally perform the update

  zero before producing a chromatogram, but it can be done

kill at any time.

  The above expression for ZERO OUTPUT SIGNAL can be presented in a simplified form

[B - A] -K [A + B]

  where K is the overall scale factor. The expression is inde-

  pending variations in the overall brightness of the light beam reaching the photoelectric cell 52, since the setting term there

  at zero (K [A + BJ) is not a constant but, like the term of diff-

  ference (B - A), is proportional to the brightness of the beam. By

  example, if the brightness increases by 10%, the term of diff-

  ference as the zeroing term similarly increase by 10%, and the full expression therefore remains zero. When there is effectively a deflection of the beam, under the effect of

  changes in refractivity, the zeroing term remains approxi-

  matively constant because of the complementary shape of the two

  lules 180, 182, which have approximately the same vertical height

  combined with any horizontal position.

  Two successive approximation registers 138, 140, atta-

  quent two digital-analog converters 142, 144, to form the signals ZERO FINE and ZERO GROSSIER. Each converter 142, 144 multiplies the SUM signal by a scale factor which is fixed by the digital output signal from the registers 138, 140. The registers

  138,140 follow a classical algorithm of successful approximations

  sives to choose digital output signals or scale factors. About once a second, the registers receive a clock pulse which comes from chip 148, which produces a

  slow clock from the high / fast clock signal of the process

  sister. At each clock pulse, the output signal from a register is adjusted in accordance with the output signal from comparator 146 which indicates whether the ZERO FINE / COARSE signal which is applied is too large or too small. The input signal of the comparator 46 is the CONTINUOUS OUTPUT SIGNAL signal which is produced by the amplifier.

  filter (see figure 12b) .We use a RESTAURA- connection

  FILTER TION between the zeroing circuit and the filter amplifier 150, during the zeroing process, to discharge the capacitors in the filter and to restore the CONTINUOUS OUTPUT SIGNAL to zero. This allows to accomplish faster

  the automatic zeroing sequence. Register 138 acts first

  mier to fix the coarse scale factor KC, then the register fixes the fine scale factor KF. The central processor uses the AUTOMATIC ZERO ORDER signal to initiate the automatic zeroing sequence. The AUTOMATIC RESET signal is used to alert the central processor that the refractometer is performing

the zeroing sequence. We will now consider Figure 12b which represents the circuit of

  processing of the ZERO OUTPUT SIGNAL. The amplifier

  52 raises or lowers the signal level as a result of gold signals

  dre 151 which come from the central processor 120 via the flip-flops 124. The demodulator 154 (helped by the phase calculation sub-assembly 153) converts the alternating signal into a

  continuous signal and the filter amplifier 150 smoothes the continuous signal.

  The switching sub-assembly 156 operates during zeroing so as to make the RECORDER and INTEGRATOR signals disappear. This sub-assembly is also used to change the polarity of the

  continuous signal under the effect of a POLARITY signal which comes from the pro-

  central terminator 120 via data flip-flops 124.

  Downstream of the subset 156, the continuous signal is processed by the amplifier.

  ficitor 158 and it is applied to an integral output conductor

  tor. The continuous signal is also processed by an attenuator 160, under the control of the central processor which is exerted via

  signals 162. The attenuator produces a recording output signal

  processor 164 which is applied to a recorder output terminal and to an amplifier 166, and a data acquisition voltage which is applied to the central processor for display on the panel. The sub-assembly 168 applies a marking signal to the recorder under

  the effect of the AUTOMATIC ZERO ORDER signal, to indicate on the timer

  the point at which the sample injection takes place.

  central stopper sends the AUTOMATIC ZERO ORDER signal when

sample injection.

  The operation is carried out as follows:

  under tension the oven which surrounds the optical bench and the chroma-

  tographic, and a heating time of approximately one and a half hours is observed to allow the temperature inside the bench to stabilize. After heating, a solvent is pumped into the

  reference and sample found inside the bench.

  When the solvents are changed, a sufficient time is observed for

  allow a purge of the two circuits. We then interrupt the circu-

  lation in the reference circuit (but the reference chamber 72

  remains filled with reference liquid). We then inject a sample

  lon in the sample column. The electrical output signal of the refractometer is zeroed by triggering the setting sequence

  with automatic zero which was described previously. The sample tra-

  pours the chromatographic column and enters the optical bench. Of

  in general, the variations in refractivity of the sample provo-

  quent a movement of the light beam relative to the photoelectric cell 52, which causes a change in the electrical output signal, which is plotted as a function of time on a recorder to

  scrolling, thus producing a chromatogram.

  The temperatures inside the chambers 70, 72 of the circulation cell are maintained at around 0.0001% C from each other

  during operation to minimize the error. A differ-

  rence of temperature between the two circulation cell chambers

  tion results in a difference in refractivity. We achieve equalization

  tion of the temperature by establishing good thermal insulation around the circulation cell, in the form of the air space 36 between the interior * and exterior cylinders, screens 38 and and blackened disc 77; by surrounding the circulation cell with a thermal mass in the form of bridges 56, 58 and

  end cap 48; and directing the circulation of the exchange

  incoming turret so that it passes through a heat exchanger at

  counter-current, very effective, which brings the temperature of the sample

  lon to that of the circulation cell. The incoming sample which is upstream of the heat exchanger has a temperature difference characteristic of 1'C (sometimes reaching 2 to 30C) compared to the circulation cell, due to the spatial differences in the temperature of l oven and because of the heat which is released by viscous heating inside the inlet tube. This temperature difference is gradually reduced over the length of the heat exchanger, by thermal conduction between the tubes.

  entry and exit. At the end of the heat exchanger, the difference

  The very low temperature that can remain is minimized by heat transfer between the end cap 48 and the coil.

  94, just before entering the traffic cell.

  The sample heat exchanger is divided into three zones, to increase its efficiency, and each successive zone has better thermal stability and is closer to the temperature of the circulation cell. The heat exchanger is constructed so as to ensure good thermal conduction between the tubes but

  a very weak conduction in the direction of circulation of the tubes.

  There is significant heat transfer between the tubes and the surrounding air. The thermal interaction between the heat exchanger and

  the surrounding region must therefore be taken into account. The pre-

  zone, between the outer cylinder 34 and the screen 38, ensures

  gradual temperature transition before the heat exchanger

  their only enters the optical bench. The length of this area is more than 10% of the length of the sample inlet tube inside the bench. In the absence of the first zone, i.e. if the inlet and outlet tubes were joined just outside

  from the entry in the end cap 42, there would be a transition

  steeper temperature along the heat exchanger, and much of this temperature transition would occur

  along parts of the heat exchanger which are inside

* laughter of air space 36, which would produce heat transfer

  undesirable to or from the bench. With the configuration

  tion preferred using a first zone outside the bench, the temperature of the heat exchanger is closer to that of

  bench when entering the air space.

  The heat exchanger enters the air space from the outside

  from the bench on the side of the photocell, whereby any heat transfer to or from the bench that actually occurs takes place in a well-located location

  separated from the circulation cell.

  This same principle consisting in passing the heat exchanger

  their by regions with increasing thermal stability is also applied to the second and third zones. In the second zone, the sample heat exchanger is directed along the air space 36, from the end of the photocell side to the end of the circulation cell side, where the temperature stability is most high. In the third zone, the sample heat exchanger is wound in a serpentine behind the end cap which corresponds to the circulation cell, and each successive turn

  is closer to the end cap and the flow cell.

As a final phase, the sample inlet tube alone is coiled in the cavity 60 of the end cap 48 to minimize the small temperature difference which may possibly

remain between the incoming sample and the circulation cell.

  Because the reference solvent does not circulate during a measurement, the reference temperature exchanger is less perfected. It does not have the third serpentine zone and it does not have a heat-conductive copper braid surrounding the inlet and outlet tubes. A limited heat exchange is established on the reference side to maintain an approximate temperature equalization during the purge of the reference circuit, which shortens the time which is necessary for the stabilization of the temperatures after the purge.

  The optical elements of the refractometer are shown diagrammatically

  tick in Figures lOa and lOb. For clarity, the optical path is shown in an unfolded form, the mirror 78 being considered as a window. Figure 10a shows a horizontal section of the

optical path while FIG. 10b represents a vertical section.

  Considering FIG. 10a, it can be seen that a single light ray B is shown to illustrate the movements of the beam. The surface

  lens 74 which is located on the circulation cell 54

  on the photoelectric cell 52, the light which emerges from the fiber optic cable outlet 50. The image which is focused on the photoelectric cell is shown diagrammatically on the left side of the figure. To illustrate the effect of rotating the

  prism 104, the figure shows four views of the prism (a to d) in different

annular angular positions, as well as the corresponding positions

  of the light beam on the photocell. The light that travels

  pours the chambers 70.72 is deflected proportionally to the difference

refractive index of liquids in both chambers

bres. FIG. 8 shows that the chambers are constructed in a conventional manner so that the surface 190 of the chamber 70 is parallel to the surface 188 of the chamber 72, the surfaces 184 and 186 being parallel

  in the same way. These four surfaces are the four surfaces at the

calf from which light is deflected by refraction. If the two Cham-

  bres contain a liquid with the same refractive index, the light is deflected by the same value on all the corresponding parallel surfaces and it emerges from the circulation cell along a path B which is practically unaffected by the changes in refractivity which are common to both rooms. If the liquid in the two chambers does not have the same refractive index, the light is deflected differentially at these parallel surfaces and it emerges along a trajectory which is inclined with respect to the trajectory of equal refractivity . The light ray B 'on the

Figure lOa illustrates such a condition. We measure the value of the incli-

  nation or deflection of the light beam in the circulating cell

culation by detecting the position of the image of the beam at the level of the photocell 52. The difference between the electrical signals

  output of the two triangular cells 180, 182 makes it possible to determine

ner the horizontal position of the light beam with very fine resolution. Alignment imperfections of the photocell

with the circulation cell and other tolerances of the pro-

  characteristically evoke an inequality of these electrical signals

  output of the two cells, even when the sample liquid

and the reference liquid have the same refractive index. This diff-

  initial electrical ference is canceled by the update procedure

  automatic zero which is described above.

The position of the light beam on the photoelectric cell 52 should theoretically only be a function of the difference in refractive index between the sample and the reference (and it should not be a function of the position of the lamp filament 101). To achieve this, the distribution of the light intensity on the output 50 of the optical fibers must have spatial stability over the chromatography period considered (one second to several hours). This requires that the distribution of the light intensity which is applied to the optical fibers is stable. The apparent position of the lamp filament 101, / of the inlet of the optical fibers 114, varies under the effect of the deformation of the filament and of the thermal vortices in the path in the air between the filament and the inlet. The movements of the filament in the direction of the length thereof (vertical direction on the

Figure 2) are relatively non-critical. Similarly, the chan-

  The distance between the filament and the entry of the optical fibers is not observable and is therefore not critical. Along the third axis of movement (vertical direction in the figure), the apparent position of the filament, seen by the entrance of the optical fibers, can be subject to spatial stabilization podr that the position of the beam at the level of the cell. photoelectric either

  independent of the position of the filament. To achieve this stabili-

sation, an optical modulator of spatial homogenization is used, being in the form of the rectangular prism 104 in the light path between the filament and the entry of the optical fibers. The prism establishes an offset of the optical path which is a function of its position of rotation. When the prism turns, the filament seems

  perform an optical scan of the entrance face of the op-

ticks 114. In position a, the prism is oriented so that the

light coming from the filament 101 is deflected out of the angle of

  the fibers of the cable 18, and the light which is transmitted to the

  bench is negligible. In position b, the prism has turned sufficiently

  so that the light is transmitted by at least some of the fibers of the cable. In position c, the image of the filament scanned the face of the entrance of the optical fibers under the action of the prism; In position d, the prism has brought the image into a position which is beyond the acceptance angle of the fibers, and the transmitted light is again negligible. When the prism turns even more, the beam begins to reappear beyond the angle of acceptance of the

fibers, as in position a, then another scan begins.

The scanning action, including the negligible light transmission period, takes place twice each turn of the prism, or approximately

times per second.

If the filament 101 moves or seems to move, this has the effect of changing the instant at which the beam begins and ends its scanning of the input of the optical fibers. Thus, the movement of the filament only modifies the phase of the movement of the beam. The electronic circuits described above calculate the average or median position which is scanned by the image. The electronic circuits are insensitive to these phase shifts or time shifts, and the

  unwanted effects of filament shift are therefore minimized.

The apparent position of the light source is further stabilized by using the beam of distributed optical fibers.

  random, 18. In an optical beam of fibers distributed by-

  randomly, the fibers adjacent to one end of the bundle are distributed randomly at the other end. So increasing the light on one side of the beam entrance and decreasing

  the light on the other side does not cause any change in the distribution

  Light at the exit end of the optical fibers In practice, the random distribution in a beam is not perfect, and a certain change does appear at the exit end. However, the use of random distribution cptic fibers still effectively reduces the effect of filament movement.

on the movement of the beam at the photocell 52.

  As shown in Figure lObles fibers do not focus the beam on the photocell in the vertical direction, as is the case in the horizontal direction. Instead, the light emerging from the outlet 50 of the cable 18 remains unfocused in the vertical planes, which produces for each light point at the outlet a vertical line of light on the photoelectric cell. The vertical height of this line is limited by the vertical height of the mirror 78, which acts as a mask. The light rays which come from individual points, for example points X and Y, on the exit 50 of the cable diverge, but only the rays which are inside

  limit radii X1, X2 (Y1, Y2 for point Y) reach the cell

the photoelectric (the other rays are not reflected so as to pass through the photoelectric cell). The vertical heights of the mirror 78, of the photocell 52 and of the cable outlet 50, as well as the spacing between the ends of the bench which correspond to the circulation cell and to the photocell are chosen so that the limiting radii for all the points that are on

the cable outlet fall entirely above and entirely above

  below the triangular cells 180, 182 of the photocell

  than. The limit radii for point X and point Y, at the upper and lower ends of the cable outlet, are shown in Figure lOb. So each point on the cable outlet produces a line

of uniform intensity on the photocell. These lines overlap

all of the photocell, which ensures uniform vertical intensity over the entire extent of the photocell, regardless of the vertical variation in intensity which may exist at the cable outlet (for example under the effect of a variation filament intensity in the vertical direction). The final result is constituted by the light intensity profile which is represented on the left side of FIG. 10b. Over the vertical height of the photocells, the intensity is uniform; outside the cells

  photoelectric, the intensity decreases to zero. Vertical uniformity

  wedge of light intensity on photoelectric cells is necessary to linearly determine the horizontal position of the light beam on triangular cells 180, 182 (a vertical variation in intensity would be indistinguishable from a movement

horizontal of the light beam).

  The invention may give rise to other embodiments.

  For example, reflective coatings other than gold could be used (e.g. aluminum, silver or

multiple layers); one could use an anti-reflective coating

  health instead of black epoxy coating 76, with a light trap placed behind the cell and outside of it

  to absorb the light passing through the coating; we could replace

  place the borosilicate glass used for the circulation cell with quartz glass or the like; and the glass elements of the circulation cell could be joined together by

  fixation by diffusion or with an adhesive.

  It goes without saying that many modifications can be made.

  tees to the device described and shown, without departing from the scope of the invention.

Claims (10)

  1 Measuring device of the type comprising a light source which directs a light beam towards a detector by passing it through a measurement area, in which the movement of the beam in a measurement direction, relative to the detector, gives an indication of the measurement which is carried out in said zone, characterized in that it comprises: means which modulate the position of the beam relative to the measurement zone, in a modulation direction and with an amplitude chosen in advance, this amplitude being independent movements of the light beam at the light source or in the beam path between the source and the means of modulation, whereby you can make the measurement practically inde- pending beam movements in the modulation direction.   2. Device according to claim 1, characterized in that the   modulation direction is the measurement direction.   3. Device according to claim 1, characterized in that it further comprises second means which modulate the position of the beam in a second direction of modulation relative to the   beam, thanks to which the measurement can be made practically independent   during beam movements in the direction of modulation   first and in the second modulation direction.   4. Device according to claim 1, characterized in that the modulation causes the beam to perform a cyclical scan between first and second positions along the direction of modulation.   5. Device according to claim 4, characterized in that   the modulation means include means which produce practical only a momentary extinction of the beam at the first and second positions. 6. Device according to claim 1, characterized in that the modulation means are located on the path of the light beam, upstream of said zone. 7. Device according to claim 6, characterized in that the modulation means comprise optical modulation means to modulate the beam. 8. Device of the type comprising a light source which directs a light beam towards a detector by passing it through a measurement area, in which the movement of the beam in a measurement direction, relative to the detector, gives an indication of the measurement which is carried out in said zone, characterized in that it comprises means which allow the light beam to pass, without being focused in selected planes, from the light source towards the detector, by crossing the measurement zone. 9. Device according to claim 8 characterized in that said means comprise: a lens structure which, in general, only focuses the light beam in the measurement direction; and masking means which block certain parts of the light beam so that light from individual points on the source is spread at the level of the detector in overlapping lines of light significantly on the detector, whereby a variation in the   light source density along the unfocused direction   does not significantly affect the measurement made at the detector. 10. Device for measuring variations in refractivity   of a fluid which circulates in a cell whose temperature is regulated   lée, characterized in that an inlet tube and an outlet tube des- intended to convey the sample fluid respectively to and from the cell are mounted so that there is exchange of heat between them.   he. Device according to claim 10 characterized in that the tubes extend side by side against the current so that the temperatures of the fluid passing through the inlet tube gradually approach the temperature of the cell, passing through at least   a first heat exchange zone.   12. Device according to claim ll characterized in that the cell is mounted inside an enclosure and the first zone is   found inside the enclosure.   13. Device according to claim 12, characterized in that the tubes extend in a second heat exchange zone, at least, which is located upstream of the first zone, this second zone being inside the enclosure.   14. Device according to any one of claims 10 or 13, characterized in that at least one of the tubes is coiled in a serpentine.   15. Device according to claim 13, characterized in that the tubes extend in a third heat exchange zone which is   located outside the enclosure.   16. A refractometer of the type comprising a cell in which there is a sample fluid and through which a beam passes luminous, characterized in that the cell comprises a reflective layer   which is integrated into the cell in order to reflect the   light beam backwards by making it cross the cell again. 17. Refractometer according to claim 16 characterized in that the reflective layer consists of a reflective coating which   is applied to an exterior surface of the cell.   18. A refractometer according to claim 17 characterized in that the reflective layer covers a limited area of said surface and therefore only reflects part of the light beam which passes through the cell, whereby the reflective coating acts as a   mask to block certain parts of the beam.   19. A refractometer according to claim 17 characterized in that the area covered by the reflective layer is chosen so that a percentage less than or equal to 50.% of the reflected part of the beam is constituted by light which does not pass through l 'sample which is in the cell.   20. Device for measuring the deflection of a light beam   by detecting this beam using means which carry out two optical measurements and which provide an electrical output signal which depends on the difference between these two measurements, characterized in that it comprises an electrical circuit which includes in the signal of output an offset term depending on the value of at least one of the measurements, and of the means which effect a zeroing of the device by adjusting the   shift term so that the output signal is equal to a value their predetermined basis. 21. Device according to claim 20, characterized in that the   offset term depends on the two measurements. 22. Device according to claim 21, characterized in that the output signal depends on the difference between quantities A and B, which are themselves respective quantities corresponding to the two measurements; and this device is of the type in which the sum A + B is   practically insensitive to variations that appear in measurements   res, while the shift term depends on said sum so   to maintain the precision of zeroing when the measurements change   nice. 23. Device according to claim 22 characterized in that the shift term is the product of said sum by a constant which   is determined by the zeroing means.   24. A refractometer of the type comprising a cell in which there is a sample fluid and through which a light beam is passed, characterized in that this cell has a curved surface integrated into the cell which acts as a lens   so as to focus the light beam.   25. Refractometer according to claim 24, characterized in that   the curved surface is the first surface that the beam crosses.   26. A refractometer according to claim 2 5 characterized in that said surface has a curvature in a first family of parallel planes, but does not have a curvature in a second family   of planes orthogonal to the planes of the first family.   27. A refractometer of the type comprising a cell in which there is a sample fluid and through which a light beam passes, characterized in that this cell comprises a part of   opaque masking which is intended to absorb certain parts of the luminous skin.   28. A circuit of the type comprising an analog circuit which is   aimed at producing an analog output signal under the effect of a signal   input, and a digital circuit to sample the output signal   tie, this analog circuit comprising a charge storage element which can be charged under the effect of the input signal, characterized in that it comprises charge dissipating means which are intended to dissipate the charge which is accumulated in charge storage element during an interval preceding signal sampling   output by the digital circuit.   29. The circuit as claimed in claim 28, characterized in that the digital circuit and the analog circuit are connected in a loop and the digital circuit comprises means which apply the input signal to the analog circuit and which periodically change the signal   input due to changes in the sampled output signal.   30. Circuit according to claim 29, characterized in that the digital circuit comprises means which change the input signal   so that the output signal approaches a desired value.   31. Circuit according to claim 30, characterized in that the   desired value is zero.