Visitors to the AAT are not normally expected or encouraged to hypersensitise (hyper) their own plates, though they may do so if they wish. The following section was initially prepared so that the hypering techniques which we have evolved over several years were adequately documented and so that other staff members could use the procedures. However, since the sensitometric properties of hypered plates are changed by hypering and are strongly recipe dependent, and the observing procedures themselves can be dictated by plate pre-treatment, full details of the processes used at the AAO are described here.
Plates are received at Siding Spring between 5 and 6 months after ordering from Eastman Kodak in Rochester, USA. An elaborate procedure is followed to ensure that their transit time is minimised and that the packaging is regularly replenished with dry ice at intervals along the way. The plates arrive by `Consolidated Air Freight' and are customs cleared by Kodak in Melbourne. After trials using charter aircraft we now ship plates by road from Melbourne, often sharing a consignment with the UK Schmidt. Apart from one or two breakages our plates have always arrived in good condition. They are tested for chemical fog level (an indication of in-transit warm-up) as soon as possible after arrival.
At the AAT, plates are stored in a walk-in deep freeze kept between -12 and -14°C on the office floor level. To avoid wastage and to maintain an efficient stock rotation system only AAO staff have access to this store. The condition of the plates is checked regularly by developing samples to determine the chemical fog levels. We have found that the fine grain IIIaJ and IIIaF emulsions store very well, a rise in chemical fog of ~ 0.02 density units per year being typical, while the fog level of IIa and 098 plates rises by about twice this amount annually. We naturally try to minimise time before use of all of our plates but the long order lead time and erratic usage rate makes such a goal difficult.
Emulsion types IIaO, IIaD, 098-04, IIIa J, IIIaF, and IV-N are always in stock in the commonly-used 10 x 10 x 0.06 inch sizes. Some of these emulsion types are also stocked in the 5 x 7 x 0.04 and 4 x 10 x 0.03 inch formats, but the smaller sizes are in much less demand and stocks are gradually being reduced so that they will eventually no longer be available.
The following discussion refers to all types of plates used in astronomy (generally known as `spectroscopic' plates) except type IV-N, which is the subject of a separate section.
The purpose of hypering is to increase the effective long-exposure photographic speed of the emulsion without adversely affecting its imaging and contrast characteristics. The aim is to improve the gain or DQE of the detector and reduce telescope exposure times. With the procedures described below an increased sensitivity to low light levels is derived from three quite distinct mechanisms which give a reduction or elimination of low intensity reciprocity failure (LIRF) as well as a real increase in speed which is not time-dependent. All of these mechanisms operate in the usual hypering at the AAT and it is important to appreciate the underlying reasons for the sequence of events.
The procedures in use at the AAT are based on the experiences of a number of observatories where hypering is carried out, modified and improved to suit our requirements and facilities. Basically, the techniques are designed to remove oxygen and water, naturally present in the gelatin component of the emulsion and to follow this by reduction sensitisation in a hydrogen atmosphere. Oxygen and water are known to be a principal cause of low intensity reciprocity failure (LIRF) in most emulsions, including those specifically manufactured to minimise this problem which are identified by a lower case `a' after the emulsion type code number in the Kodak literature.
These volatile components are flushed out by baking the plates in a flow of dry nitrogen at 65°C. This has the additional effect of completing the so called `ripening' stage of emulsion manufacture which is deliberately left unfinished by Kodak to improve the storage properties of the materials. Even without gas flushing, baking in an enclosed box of air gives a significant speed increase due to the ripening effect alone. Prolonged room temperature nitrogen flushing (the UKSTU system, Sim 1977) and the application of vacuum without heat are therefore not strictly equivalent to nitrogen baking. This difference is evident in the poor keeping qualities but higher speed of some of our hypered plates (particularly IIIaF) when compared with non-bake processes, and seems to be an inevitable consequence of approaching the ultimate sensitivity of these products.
The second stage of the hypering process involves `seeding' each of the silver halide grains with a few atoms of metallic silver to form sub-latent image ( i.e. not developable) sensitivity specks. In the ideal situation such grains need only 1-2 more silver atoms at the same site to produce a true latent image capable of development. The intention is that these should come from interaction of the grains with image photons. In the absence of the sub-latent image specks each grain would require to interact productively with 4-6 image photons to render it developable. Reduction sensitisation thus lowers the detection threshold by a mechanism conceptually similar to the well established preflashing technique but without the contrast reduction inherent in that method.
In all these procedures (including preflashing) the gain in speed is limited by the inevitable growth of non-image fog which lowers the output signal-to-noise ratio of the processed plate. An important advantage of chemical sensitisation is that the grains which are most susceptible to over-hypering ( i.e. those which give rise to chemical fog) are not necessarily those which are most sensitive to light, whereas with preflashing the opposite is true. This rather subtle difference means that it is possible to obtain a useful gain in detectivity by preflashing fully hypered plates, but only if the telescope exposure is chosen so that no sky background is recorded. This approach is useful in narrow-band interference filter photography.
Control of the chemical fog level is a critical part of the hypering process. The ratio of the bake time to hydrogen soak must be carefully determined by a series of experiments to be described later. In general, too much baking gives high chemical fog without further gain in speed while too much hydrogen (after an optimum bake) increases both fog and speed. Hypering recipes are designed to yield maximum speed at a fog level not exceeding a density of 0.3 (ANSI diffuse).
Baking and hydrogenation each give speed gains which are almost additive for long exposures, but it is important to note that hydrogenation in the presence of oxygen and water is not as effective as it is in an emulsion free from these contaminants, quite apart from the lowering of LIRF given by outgassing. Presumably, this is why baking in forming gas, a dilute hydrogen in nitrogen mixture is less effective than separate nitrogen baking followed by pure hydrogen.
The increased speed derived from gas hypering can be quickly lost if the fully hypered emulsion is exposed to normal air before or during exposure to light. Dry gelatin is very hygroscopic and rapidly absorbs moisture from its surroundings, partially nullifying the effects of hypering. The latent image is also subject to regression if an exposed plate is left in air. (See Malin 1978). The procedures described below have been designed to produce and preserve the maximum possible speed compatible with convenient operation.
Plates are kept in a cold store in 12-plate cardboard boxes as received from Eastman Kodak, the most usual sizes being 5 x 7 and 10 x 10 inches. This description refers to the larger plates but is generally applicable to all sizes normally stocked. The plate packaging consists of three boxes, each one fitting inside the other to make a compact easily opened, light-tight container. Note that the boxes are not square and can only be replaced within each other in two ways, 180 apart. The inner box containing the plates has a small section cut out of one side to simplify removal of plates after the first few have been used. A piece of cardboard is inserted between the plates and the cut-out slot to prevent light leakage when the box is assembled. In newly opened boxes the tape joining the cardboard flap to the box wall must be cut with the thumbnail before the cut-out is revealed. This piece of cardboard must be replaced before re-assembling the box.
Plates are packed with emulsion side alternating throughout the box with the top plate emulsion down. They are kept apart by cardboard frames resting on (and fogging) the edges of the plates. The cardboard frames are about 1cm wide and great care must be taken to avoid dropping them on to the emulsion surface. As plates are removed one at a time the cardboard separators can be discarded but a separator must be left on the last plate remaining to prevent the emulsion surface contacting the lid when the package is re-assembled.
Occasionally, the first plate in an already opened box is a cut piece remaining after taking samples for test. This piece always has one dimension such that it cannot slip through a separator on to the plate below. The cut plate is always emulsion down in the box and must be replaced in that orientation on top of a separator after removing the required number of plates.
Figure 4.1 Arrangement of slot and removable insert in the Kodak Full Frame Packaging (FFP) plate box.
Figure 4.2 Arrangement of contents with Kodak Full Frame Packaging.
Full boxes of 10 inch square plates require at least 4 hours to warm up after removal from the cold store before they can be opened. It is normal practice to remove from the store all the unhypered plates required for a run and leave them at room temperature for a day or so. Rise in fog level in this time is negligible. However I-N, 098-04 and IIaD's are more sensitive in this regard than other types and their time at room temperature should be minimised.
Pre-cut 2 inch square samples of all current batches of plates are kept in aluminium containers along with cardboard boxes in the cold store. These should be removed (and returned) to the store at the same time as the bulk packages to maintain a similar thermal history.
All the stainless steel hypering boxes in use at the AAT are fitted with a removable lid containing a one-way outlet valve and Neoprene sealing gasket. The box itself has an inlet valve at the bottom and should contain a rack for either 12 plates 10 x 10 inches or 24 plates 5 x 7 inches. The racks for the larger plates also accept an aluminium tray fitted with grooves to hold up to eight 2 inch square samples. These samples should, of course, be from the same batch as the large plates in the box.
A standard way of loading the hypering boxes has been evolved which minimises the chance of misidentification and damage to plates. Although several types of emulsion may have similar hypering recipes, only one type of emulsion is normally loaded into any one box. The plate racks are more easily loaded when they are removed from the boxes. Plates are fitted into the racks with the emulsion side facing you and from front ( i.e. nearest you) to back. Since plates are often hypered in batches of 6 or less this leaves a gap at the rear of the rack. In the last or next to last groove the sample tray is installed, again with plates emulsion out, facing forward. This arrangement permits rapid removal of samples for test without any danger of damaging the large plates.
The loaded rack is lowered into the hypering box so that with the emulsion side of the plates facing you the gas inlet on the box is on your right. The lid is clamped, again with the gas outlet to the right. In this orientation a label is affixed which details the box contents, hypering history and test results.
The laboratory oven in the plate baking laboratory on the first floor is set to run at 65°C and should have been switched on at least an hour before use. The oven contains a substantial coil of copper tubing through which the nitrogen passes to pre-heat it before it enters the hypering box. Before putting the boxes in the oven they should be flushed with about 5 x their volume ( i.e. 60 litres) of nitrogen. A gas meter is provided to measure the outlet volume. The oven is fitted with adjustable flow meters and tubing so that two boxes can be baked independently. After placing the box(es) in the oven and attaching the gas inlet tube to the bottom nipple, the gas flow is adjusted with the flow meter so that the box receives one tenth of its volume of nitrogen per minute. The large boxes require 1800 mL per minute, the small sample boxes 600 mL per minute. This flow rate is entirely arbitrary, but enables these conditions to be reproduced in any sized hypering box and any oven.
The baking cycle is timed from the insertion of the boxes into the oven until their removal. During cooling, a nitrogen flow of about 2 litres per minute from a portable flow meter is bled into the box to prevent ingress of air. The large boxes require at least 60 minutes for their contents to reach room temperature. This desirable equilibrium is more certain if the box is stood in the draught of the fume cupboard vent alongside the oven.
The plates must be taken to the UK Schmidt to receive hydrogen. Access to the Schmidt hypering facility is strictly limited to specified persons who have received detailed instructions on its operation. The procedure will not be described further here.
In the absence of a designated AAO staff member, UKSTU staff may arrange for the boxes to be placed on their system but this should only be necessary in an emergency. Note that hydrogenation should be started as soon as possible after cooling so baking must be planned so that UKSTU staff are available to both install and remove boxes from their hydrogen system. Plate boxes removed from the UKSTU system receive a final automatic purge of nitrogen to displace the hydrogen.
Plate hypering boxes have been tested for light leaks in bright sunlight and found to be satisfactory. However carrying a full box of plates by the handle may cause sufficient flexure to distort the container and admit light. It is suggested that a dark cloth be used to cover boxes carried in bright sunlight.
To maintain hypersensitised plates in fresh condition they should be sensitometrically tested immediately the hypering cycle has finished and transferred to the cold store as soon as possible.
A laboratory version of the sensitometer fitted to the prime focus camera is available for plate testing in the 5th floor darkroom. This device supports the sensitometer tube and lamphouse in a box which is provided with facilities for exposure in a nitrogen atmosphere to closely mimic conditions on the telescope. A 2mm Schott BG34 filter in the optical train converts the ~2900K tungsten source to a colour temperature of approximately 6000K so that a 15 minute exposure produces a useful range of sensitometric spots with all our commonly used emulsion and filter combinations. A series of 50mm square filters, 2mm thick, is provided for insertion in the filter drawer on the instrument. As on the telescope these act as the transparent side of a cell from which air is displaced by nitrogen before exposing a hypered plate. The sensitometer is well made and a continuous purge of nitrogen is not necessary.
Standard emulsion and filter combinations for the sensitometry of hypered plates are:
IIIaJ or IIaO + 2mm GG 385
IIaD + 2mm GG 495
IIIaF or 098-04 + 2mm RG 630
IV-N or I-N + 2mm RG 695
Satisfactory spot densities are obtained with a 15 minute exposure and an aperture of 1.6mm set on the calibrator lamp house aperture plate.
The laboratory calibrator is designed to take 50mm square plates and, by means of a simple adjustment, films or plates up to 4 x 5 inches. Once the sensitive material has been loaded and the appropriate filter is in place room lights can be turned on.
To avoid ingress of air, 50mm sample plates should be removed from hypering boxes with a vigorous flow of nitrogen entering the box. A nitrogen line is provided in the 5th floor darkroom for this purpose (see plate loading, section 5.10). The box of hypered plates should be taken or returned to the cold store as soon as possible after purging.
After the sensitometric exposure the plate is processed, preferably immediately. However, it is possible to store up to three exposed plates emulsion down beneath the lid of the calibrator. The plates are processed in the small (2 litre) gas burst tanks in the special processing rack provided. Developement is for 5 minutes in D19 at 20°C ± 0.2° with a 1.2 second gas burst every 8 seconds. Inlet gas pressure should be set at 50kpa. It is important to ensure a uniform bubble pattern by levelling the tank. Development is followed by a 30 second rinse in the stop bath and fixation in rapid fixer for 2 minutes, again in a gas burst tank.
Sensitometer samples are retained for a few days only so a 5 minute wash is adequate. A brief rinse in diluted Photo-flo (NOT in the large tank for full sized plates) aids uniform drying, speeded by putting the test plate on the bottom shelf of the film dryer set to 350W, with the blower fan on. Plate densities must not be measured with wet plates, as density changes markedly on drying.
After drying, the step wedge densities on the sample plate are measured on the Macbeth densitometer on the first floor. The fog level is also noted. Densities above fog are plotted against a set of standard curves (kept with the densitometer) with the log E axis marked in minutes of exposure time for the triplet corrector. The telescope exposure time necessary to reach a density above fog of 0.6 (for IIa and 098-04 emulsions) or 1.0 (IIIa emulsions) can be read off. These figures have been gathered over a long period and give an accurate indication of exposure time to reach a given density in a dark sky through the appropriate filter. They are however dependent on both the output and colour temperature of the calibrator lamp remaining constant. Regular testing with specially reserved batches of IIaO (B) and 098-04 (R) over many years has shown that the output and colour of the lamp, while changing detectably, is adequately constant for our purpose. The laboratory results for batches of plates used on the telescope correlate well with the actual sky fog densities obtained in practice and the drift in sensitometer characteristics is sufficiently slow and small for us to detect the change in night sky brightness with solar cycle.
Typical triplet exposure times for properly hypered plates are listed in Table 4.1. Doublet exposure times are about 15% shorter. All these values have been obtained with a fog level of 0.3 or less, but recent (since 1984) batches of IIIa plates have been systematically slower than previously. Discussions with Eastman Kodak now (1989) appear to have overcome this problem.
|Limiting Exposure||Limiting Sky|
|IIIaJ||GG 385||385-550||70||~ 15 hrs||1.0-1.2|
|IIIaF||RG 630||630-700||80||~ 24 hrs||1.0-1.2|
|IVN||RG 695||695-900||90||~ 10 days||1.0-1.2|
*ANSI diffuse density above fog, 5 mins D19 development, dark sky
If the plates are under-hypered (low fog, low speed), warm the boxes to room temperature and give an additional hydrogen soak. If they are seriously over-hypered and a high fog level cannot be tolerated, start again.
Sky limited exposures in moonlight are difficult to calculate with any precision since the night sky brightness changes markedly across the sky and with the phase of the moon. However, the following table, based on figures provided by the UK Schmidt Unit, is offered as a rough guide to what can be achieved with some of the more usual combinations of plate and filter.
The table lists the exposure (in minutes) needed to give a sky background density of 0.6 or 1.0 above fog at various phases of the moon, using an f/3 telescope and hypersensitized plates.
|Emulsion+ filter||Exposure Time (min)||Background|
|IIIaO + UG1||90||30||20||12||-||0.6|
|IIaO + GG385||30||12||7||-||-||0.6|
|IIaO + GG495||30||25||12||7||-||0.6|
|098-94 + RG630||40||30||20||10||-||0.6|
|IIIaJ + GG385||70||30||15||5||-||1.0|
|IIIaF + RG630||90||70||40||15||5||1.0|
All emulsions exhibit marked batch-to-batch variations in their response to hypering. IIIaJ and F are particularly bad in this respect though recent batches (since mid 1978) have been more consistent. IIaO and 098-04 are much more predictable. The only general trends are for the IIIa emulsions to require much more baking than the others, up to 8 hours in some cases, compared with a maximum of 2 hours for IIaO's, and for a fall in obtainable maximum speed since 1975, when accurate records were first kept. This latter trend seems at last to have been reversed.
To establish useful recipes in a reasonable time the following procedure is adopted. Using the small sample boxes containing a sample rack, a single 2 inch square test plate is put in the oven to bake. An hour later another plate is added to the one in the box and so on until a range of 5 or 6 hours is covered. Baking continues until the box contains plates which have received, for example 4, 5, 6, 7, 8 and 9 hours bake in the small box at the standard 10% per minute nitrogen flow rate. After removal from the oven, the box is allowed to cool to room temperature with a slow bleed of nitrogen to prevent ingress of air and the plates hydrogenated for, say 4 hours at 20°C. Each plate is then tested as described in Section 4.7.
From this series of experiments will emerge a feeling for how the batch responds to baking. If the plate with a fog of just under 0.3 is nowhere near the optimum speed listed in Table 4.1 and if the speed gain, irrespective of fog, levels out at some well defined bake time, repeat the experiment, possibly with fewer samples and give 8 hours hydrogen. With two experiments it should be possible to tell roughly what the optimum bake time is. Next, bake two samples in separate boxes for the best bake time found earlier and give them two different hydrogen times, say 5 and 7 hours. From these results it should be possible to extrapolate a reasonable recipe which should, however, be confirmed with samples in a large box.
An alternative method, which avoids much plate testing, is to include samples of the new batch in the test rack of a box of plates with a known recipe being prepared for use on the telescope. This quickly establishes if the old recipe works for the new plates. On this basis an abbreviated set of tests can be designed to refine the optimum bake/hydrogen ratio for the new batch.
In general IIaO and 098-04 plates work well with short bake times (1-2 hours) followed by 6-10 hours hydrogen whereas the more recent batches of IIIa plates have needed 5-8 hours bake and 6-8 hours hydrogen. However one recent batch of IIIaF's required 8 hours bake and 1 1/2 hours hydrogen and another batch of IIIaJ's are still not up to speed after 10 hours bake and 8 hours hydrogen. A 1983 batch of IIIaF's were excellent with 3 hours bake and 3 hours hydrogen!
This near-infrared (700-900 nm) sensitive emulsion requires a hypering technique fundamentally different from that employed for plates sensitive to visible light. Essentially this involves bathing the plates in dilute silver nitrate solution, rinsing in water, drying under cool conditions, and then exposing in the telescope as soon as possible.
The mechanism of this process is not fully understood but the remarkable gain in speed (a factor of about 400 compared with unhypered plates) mainly in the infrared region, indicates that the treatment primarily affects the sensitising dye adsorbed on to the grain surface. This idea is supported by the fact the hypersensitised but completely dehydrated plates are slower than those exposed at ambient humidity. One effect of the silver nitrate is to remove excess bromide ion which acts as a fog restrainer for storage and transport of unhypered plates. In practice, unhypered plates store extremely well at -15°C, with an undetectable rise in chemical fog over many years.
The equipment and materials necessary for preparing infra-red plates are kept in the 4th floor coudé (east) darkroom which is used exclusively for this purpose. It is essential that great care be taken to ensure that the room is free from dust, that the air temperature does not exceed 17°C and that everything which will contact the plates or solutions is washed with de-ionised water before use. The process is probably best described in stages:
a) Ensure that the room temperature is below 17°C (see NOTE). Take 20-30 mL of a silver nitrate stock solution containing 10 g/L Ag NO3 and dilute it to 2 litres with de-ionised water. This gives a solution with a concentration of 0.10 to 0.15 g/L Ag NO3 . The exact concentration used is batch dependent. Pour the solution into the tray rocker and ensure that its temperature is 20 ± 0.5°C. Use a fresh lot of solution for each plate.
b) Fill the two white plastic trays to a depth of 1cm (equivalent to 950 mL) with distilled water. To one of the trays add a few drops of PhotoFlo 60 (diluted PhotoFlo 600) to produce a just-stable foam.
c) Set up a domestic fan about 2 meters away from the wooden drying rack and place a hand-towel flat on a bench so that it can be found in the dark.
d) Before putting out the lights ensure that the double entrance doors are labelled with warning signs and that these doors are closed so that the darkroom can be vacated without fogging the drying plates.
e) Load a plate into the Perspex processing frame and lower it into the silver nitrate solution, at the same time switching on the tray rocker (foot switch). Start the timer. Process for 4 minutes.
f) During the 30 seconds or so before the end of processing, wash the hands in de-ionised water without using soap and without drying them.
g) A few seconds before the end of the processing time stop the tray rocker in a horizontal position, remove the plate from the tray and drain it briefly vertically, holding the plate so that run-off from the hands does not flow over the wet plate.
h) Lower the plate into the first tray of de-ionised water and agitate vigorously (rocking motion) for at least 30 seconds. Drain, and wash again in the second tray containing dilute Photoflo, again for at least 30 seconds.
i) Remove the plate from the final rinse avoiding contaminated water from the fingers reaching the emulsion. Place the plate, backing side down, on a hand towel and move it around to remove as much water from the back of the plate as possible.
j) Stack the plate corner-down in the wooden drying rack and turn on the fan. Orient the rack so that the emulsion faces the fan flow at an angle of about 45°. Drying takes about 40 minutes.
We have recently found that plates fog badly if dried at 20°C but are satisfactory when dried at 17°C. Ensure that the room temperature is lowered before you start work.
While loading plateholders may seem a trivial aspect of prime focus photography it is in fact rather important since only at this time are hypersensitised plates exposed to room air which rapidly de-sensitises them. The procedure for 10 x 10 inch plates will therefore be described in some detail.
Hypered plates are kept in nitrogen in stainless steel boxes; the racks within the boxes are loaded as described in section 5.3. We now make no attempt to clean plates before exposure since Kodak's full frame packing (section 5.2) is much cleaner than the previous system and particles on the emulsion surface are rarely a problem. We do however clean the plate holders before use, in particular the surface on which the edge of the plate rests and which defines the focal plane. Small particles of dust or glass on this surface can displace a plate sufficiently for it to be out of focus.
The required number of plate holders are placed on the bench, their backs opened with a push and twist of the central locking lever. they are brushed out and blown clean with nitrogen. The nitrogen nozzle is then fitted to the inlet of the plate box and nitrogen allowed to flow briskly through it. With the lights out and door locked the box can be opened and a plate removed. The purpose of the nitrogen flow is to discourage air from entering the open box, so replace the lid as soon as possible. If the inlet valve is to your right the plates should have their emulsion towards you and be near the front of the box. A rack containing samples has (unlike the plates!) rounded corners for identification in the dark and is at the rear of the rack. Be sure to confirm that you know which is emulsion side of the plates (a moistened lip sticks to the emulsion!) and write the plate holder number and emulsion type in a corner in pencil. This can be important if many plates of the same object are intended.
Turn the plate emulsion down and fit it into the plate holder, finally pressing down quite firmly and moving the plate around on the support surface to be sure that it is free from gritty particles. Replace the plateholder back and purge the plate holder with at least 5 litres of nitrogen to dispel the entrapped air. The plate holders are well made and the nitrogen atmosphere should survive for several hours.
The plate box should receive 60 litres of nitrogen to displace any air which might have been admitted when removing the plate. This of course can be done with the box sealed and room lights on.
The prime focus camera is fitted with a nitrogen inlet tube which blows nitrogen at about 2 litres per minute into the space between plate and filter. All major leaks are sealed with foam plastic but there is no attempt to pressurise the area, we only need to displace ambient air. Nitrogen normally flows all night, except when IV-N plates are being exposed.
The overall effect of these procedures is not so much to increase plate speed as to prevent speed loss by interaction of the hygroscopic emulsion with the atmosphere, which can often be very humid at Siding Spring. The following table gives results as a percentage speed loss for hydrogen hypered plates which were allowed to stand in still air at 20°C and 50-60% relative humidity for various periods. The results underline the importance of rapid plate handling and the use of a nitrogen atmosphere to prevent pre- and post-exposure image loss.
TABLE 4.3: Fall in Sensitivity of Hypersensitised Plates
This work is described in detail by Malin (1978)
The pressure is applied to the scoring wheel by a relatively long spring, set by the screw and lock nut on top of the scoring head. The length of the spring makes the cutter relatively insensitive to changes in plate thickness. The cutter pressure has been set to cut backed plates from 0.03 to 0.06 inches thick and must not be re-set. Any increase in cutting pressure will cause flaking of the glass along the score and reduce the life of the scoring wheel.
The height adjustment (lower screw) is set so that the full pressure of the cutting wheel is applied to the plate without the wheel chipping the edge of the glass at the end of a score. This setting is not critical but the height may require re-adjustment for 0.06 inch plates.
With care, a carbide wheel will make many hundreds of satisfactory cuts without replacement. Should failure occur replacement cutters can be found taped to the underside of the platen. Old cutters look just like new ones. Be sure to put the old cutter in a waste bin.