The Foucault test

Polishing and testing

We should not wait until the mirror is completely polished before trying the first Foucault test. Already after about 1 hour of polishing we can install our mirror on the "optical bench" and see the first results of our work. 
In the meantime we should always remember to perform the polishing movements slowly, keeping a stroke length of a third of a diameter and carefully observe all the precautions described in the preceding pages.
The reason for getting the earliest possible results of a Foucault test is to make corrections as soon as possible. During polishing, corrections proceed much slower than when working with coarser powders. If we wait too long for a first Foucault test and the error is considerable we could loose much time in correction, and we might even have to go back to carborundum powders, such as #1200 or less.



Before installing the mirror on the optical bench we should rinse it thoroughly with water and dry it with a clean cloth. Then we can place it on the console of the Foucault tester and leave it there for about 10 minutes to acclimatize (i.e. to reach temperature equilibrium with its environment). If the rinsing water is very cold this acclimatisation can take a long time just as when you take the telescope from a welly warmed room out into a cold winter evening. In the latter case the instrument is not performing well because all it's the parts are heating the surrounding air and creating convection currents. The mirror itself takes the longest time to reach temperature equilibrium because it is the thickest and the heaviest part of the telescope. 

On the Foucault testing cradle the mirror stands free in the surrounding air and temperature exchange proceeds more rapidly. You will find out soon enough when a mirror is ready for the final measurements.

The radius

Twee bewegingen vd Foucaulttester (1.1Mb)We look towards the mirror through the gap between the edge of the knife and the light source with one eye as near to the edge as possible (not more than a few millimetres). We move the tester in all directions until we see the reflected light of our light source just in the above mentioned gap. At first this can take some time because we do not yet know at what distance and at what angle the mirror holder should be placed.
An easy way to find this mirror image is to use a small pocket flashlight. We position ourselves at about the radius distance of the mirror (in our case about at 2.3 meter). It is not really necessary to point the focus of the torch exactly in the direction of the mirror. Holding the flashlight near your eye you move it about in all directions, except back and forth. When you see the light of the torch flashing by in the mirror you have just passed the main axis of the mirror. Keeping this image in sight ask a colleague to reorient the mirror carefully and possibly to move the Foucault tester until you get the image just behind it.
We now position one eye as near as possible behind the edge of the knife. We look through the gap to the mirror and move the tester to the left and to the right seeking the reflected light of the tester. We should do so very slowly, otherwise the image flashes by and is lost again. At this stage we need to move the Foucault tester as a whole since moving the knife by turning the knobs of the tester is of no use.
An alternative procedure for positioning the reflected light beam near the knife-edge is by catching the image of the light source on a piece of white paper. In that case the room must be completely darkened. The advantage of this procedure is that the help of a second person is not needed. Keeping the lightspot on the paper, the tester can be moved laterally until the spot falls just near the edge. By moving back and forth along the axis of the mirror, the dimension of the spot can be minimised as to fall exactly in the radial point (i.e. the centre of the sphere of the mirror).



Pressing one eye as near as possible against the knife-edge, and looking at the evenly illuminated disk of the mirror, we now slowly move the knife laterally towards the reflected light beam by turning the appropriate knob. This 'cutting' of the light beam should be done very slowly, since it only takes a few tenths of a millimetre to cut the beam to complete darkness. During this most critical phase, gradually
The knife blocks part of the light and the shadows cast on the mirror suddenly reveal precious details about the surface of the mirror. When we observe these shadows we must imagine that the light source seemingly does not fall directly on the mirror but instead grazes it from the side. However, such details are only revealed when the knife lies exactly in the centre of the sphere of the mirror.
Once we have succeeded in observing such shadows, the position of the mirror and of the tester needs to be marked with a pencil to avoid having to repeat the whole searching process during a later test.

If we only observe a vertical black shadow during the "cutting" of the lightbeam and if this shadow moves in the same direction as the knife the latter and also the light source are situated too near to the mirror. If the shadow moves in the opposite direction relative to the moving direction of the knife, the tester is situated too far from the mirror. (In both cases the distance to the mirror can be appropriately corrected with the fine tuning knob installed on the table of the Foucault tester). If the knife lies exactly in the centre of the sphere described by the mirror, the disk gradually darkens until it is completely black. If we don't see any shadows during that darkening, the mirror is perfectly spherical.


When the knife is situated exactly in the centrepoint and if we see some shadows on the disk, moving either to the right or to the left, we must conclude that certain zones on the disk still contain irregularities and polishing should be resumed using strokes of about a third of a diameter. 
To find out whether a shadow means that the corresponding zone is too high or too low we should keep in mind the direction from which the light source comes. In the above illustration (Foucault-gram) the light source is situated to the right. The mirror apparently has an upturned edge, in the 60% zone it has a (circular) bulge and in the centre it shows a smaller hill. If the light source was at the other side we only have to turn the page upside-down to reach the same conclusion (beware of optical illusions!). At a first glance it looks like this mirror is not suitable for a telescope. Nevertheless, this mirror from a Newtonian has been submitted to us for testing out of curiosity. The owner did not realise that anything was wrong with this objective. Probably scores of observers have looked through this instrument without any suspicion. Of course, after careful examination such a mirror is normally disqualified by an experienced observer. This instrument will perform badly especially for observation of double stars and even more so for planets since that kind of mirror flaw seriously lower  the contrast. Such circular bulges are called zonal aberrations or zonal deviations. Another kind of frequently occurring flaw as shown at the left (right?) can be specified as 'local surface irregularities'. Americans call this kind of surface structure 'dog biscuit'. Probably it is caused by performing the strokes too rapidly and by applying too much pression on the pitch tool.




Sometimes the image of the mirror you are testing is not sharp and steady. This could be due to turbulence of the surrounding air in the observing room. Sometimes this turbulence is so slight, that you cannot observe it directly although it is important enough to impede the testing procedure. A very convincing experiment is to hold a burning match a few centimetres away from the mirror.

Turbulentie van een lucifer (0.5Mb)

Turbulentie van handwarmte (0.4Mb)

The air in the neighbourhood of the flame (as seen from the knife-edge) is then turning into a whirling tornado! When someone holds his hand near the mirror you will see, that this is already enough to prevent you from making decent observations. A last surprising but also convincing experiment is to cut the light beam until the mirror image is in half darkness, then to hold one fingertip for one minute on the edge of the mirror. If you look again through the gap you can see a bulge at the edge where you finger has been! After a few minutes this bulge will disappear, but this proves, that due to a temperature difference of only a few degrees you can observe local irregularities of a fraction of a wavelength on the mirror.


Back to polishing

We keep polishing using strokes of a third of a diameter and take care that the angles through which we turn the mirror or the tool are never the same, to avoid building up a systematic zonal irregularity. It normally takes quite a long time to get the area along the edge of the mirror completely polished. It is sometimes a good idea to switch the mirror and the tool and continue polishing with a third of a diameter stroke. We should regularly submit the mirror to the Foucault test, press the pitch disk to match exactly the form of the mirror and, where necessary correct the incisions in the pitch disk. We also regularly check the mirror under the microscope for (scratches or) pits.


Enkele fouten

Ook hier staat de lichtbron telkens links...!

15 cm f/9

Op de bovenste afbeelding zien we een bijna sferische spiegel met centraal een grote berg en daarin een klein putje, maar als we het mes wat dieper in de lichtkegel draaien(onder) ziet hij er opmerkelijk slechter uit.  Bij de reeds aanwezige fouten komt nog een opstaande rand bij en microrippel.  Dit laatste werd veroorzaakt door de viltjes en zal tijdens het polijsten met de pek snel verdwijnen .

15 cm f/8.2

Op de bovenste afbeelding zien we een spiegel met een enorme opstaande rand...!  Ongeveer op de 75% zone zien we een fijne opstaande ring en daarbinnen een verzakt plateau.  Maar het mes staat voor deze test veel te ver naar achter.  Op de onderste afbeelding werd het mes juist geplaatst en we zien een ware metamorfose.  Inderdaad een enorme opstaande rand, maar ook grote berg met daarin (zwak) een verzakt plateau. 
Terug naar 1200

25 cm f/6 - 19 cm f/7

Op de bovenste afbeelding(25 cm) zien we een spiegel die klaar is omte worden geparaboliseerd.  De zeer kleine fouten (opstaande rand,  centraal putje en de microrippel) verdwijnen vanzelf tijdens dit proces.
De onderste spiegel(19cm) heeft een klein beetje astimatisme.  De schaduwen die tijdens het indraaien van het mes schuin invallen ipv horizontaal zijn hier zeer kenmerkend voor.

Onervaren waarnemers zullen (behalve de middelste) geen van deze spiegels als slecht beoordelen.  Sterker nog, de meeste beginnende waarnemers zullen niet eens door hebben dat er iets mis is met deze spiegels.  De 25 cm moet natuurlijk wel eerst zijn paraboolvorm krijgen uiteraard.  Van dit kaliber gaan er, spijtig genoeg, massas over de toonbank.  :-(



afgezakte rand

Commonly occurring irregularities and the way to correct them can be seen in the overview below. Before correcting faults you always should evaluate their importance, otherwise you risk overcorrecting, which implies a new correction procedure in the other direction. In the overview on the next page the mirror disk is not represented as a sphere, but as a flat disk. This is done in all publications because the mirror appears like a flat disk in a Foucault test (although in reality the shadows represent deviations from a spherical shape). Using a straight line instead of a curve in edge-on drawings the deviations are more easily seen.

Click on 'your deviation' to find the solution.