Click to email me

M73 Compass Repair

I have an M-73 Francis Barker military compass that I built from OEM parts more than 30 years ago.  The compass is generally in good order except that the luminous marker has detached from the card, there is a significant air bubble in the bowl, and the card no longer floats horizontally.  This may be due to the detached card marker, which is still rattling around in the bowl, but there are other potential causes, and the card may always have been unbalanced.



Figure 1.  M73 Francis Barker Military Compass


I’m going to attempt to repair the compass.  If you choose to follow this procedure then please read it in its entirety before you start.  No special tools are needed other than a set of jewellers screw drivers (in good condition), a magnifying glass, some non-magnetic tweezers and a hot plate.  Note that the compass screws are very small with fine threads in brass or aluminium.  Remove any paint from the head slot to ensure that the appropriately sized screw driver will seat properly in the slot.  When refitting screws turn them gently counter-clockwise to feel for the start of the thread before engaging them, and seat them positively without undue force.

Removing the air bubble will be easy through the filling port, but repairing the luminous marker and balancing the card will be a bit more challenging.



Figure 2. Bowl Assembly
(Note unbalanced card and missing luminous marker.)


The luminous markers on this compass are not based on an active tritium source.  They comprise a green/yellow power in a plastic envelope (with open ends).  I figure that they contain an activated zinc salt, but they may contain other nasty chemicals.  My original plan was to gain access to the card and simply re-adhere the original marker but subsequently I have decided to replace all of the markers with activated strontium aluminate.

There is no ready access to the card in the bowl.  The front glass is adhered firmly to the brass bowl body with an adhesive.  Unfortunately I don’t know what this adhesive is.  However it is alcohol resistant and perhaps slightly brown in colour suggesting urethane, or epoxy (probably the former given the vintage of this compass).  Note that de-soldering the bellows on the rear of the bowl is not an option as the card is held in place by the limited clearance between the pivot and the front glass, and refitting the bellows would be difficult.

I started the repair by loosening the dial clamp, removing the two grub screws that secure the dial ring, removing the dial ring and easing the bowl from the compass body.



Figure 3.  Dial Ring (marker never fitted)


Next, I removed the aluminium dial retaining ring (eight aluminium machine screws).

Getting the dial retaining ring off the front of the bowl was difficult because the parts were very closely fitted and adhered by a very thin layer of putty which had set solid over time.  I could not insert a scalpel blade into the joint so I made a very fine file cut on the edge of the brass bowl (this will not be visible on re-assembly).  Then I used a scalpel blade to gently pry off the dial retaining ring.



Figure 4.  Dial Retaining Ring

Next I removed the putty from the retaining ring and the edges of the glass by gentle scraping with a sharp plastic tool.

I removed the bowl filling plug (on the side of the bowl) and emptied the remaining ethanol from the bowl by gently depressing and releasing the bellows (at the rear of the bowl).  Note that the bellows must be handled with care throughout the repair to avoid any permanent distortion or damage.



Figure 5.  Bellows on the Back of the Bowl


The filling plug red fibre sealing washer was clearly worn.  This is the likely cause of the air bubble and it will need to be replaced with a suitably sized O ring.

Whatever I do I don’t want to damage the card, the bowl or the glass.  While the glass is probably replaceable, damage to the card or bowl will probably render the compass unserviceable.



    Figure 6.  Filling Plug Washer (and two luminous markers)


Glass Removal

In order to get access to the card I need to remove the front glass.  I tried a range of solvents hoping that these would either dissolve, soften or swell the adhesive.  These included water, ethyl acetate, acetone, toluene, turpentine, and white spirits (note that there is no point in trying alcohols because ethanol is used as the damping fluid).  The solvents were all applied externally (that is they were not placed inside the bowl) for 24 hours at room temperature.  They didn’t work.  Then I tried heated solvents to their respective boiling points for about 30 minutes with the bowl inverted in a beaker with sufficient solvent to immerse the glass.  Still nothing was attacking the adhesive.

Finally I settled on directly heating of the bowl on a hotplate (glass face down) on a piece of aluminium foil to allow for rapid secure removal of the bowl, glass and card without having to handle hot components.  The heat does two things: it softens the adhesive and expands the brass bowl.  The combination should free the glass.



    Figure 7.  Removal of the Glass from the Bowl


The duration of heat application was gradually increased in 15 second increments to 90 seconds.  Then the hot plate temperature was increase half a digit and the time increments were repeated.   After each trial the bowl was allowed to cool back to ambient temperature and all components were inspected for any sign of impending heat damage.  At a hot plate temperature settling of 4 the glass and card fell free from the bowl after 75 seconds.  The aluminium foil worked a treat for rapidly removing the bowl, glass, card and the stray luminous marker from the hotplate element.  There was no apparent heat damage.



Figure 8.  Glass
(Note ring in centre.  This is residual powder and not glass abrasion from card bearing wear)



Figure 9.  Card and Luminous Marker



    Figure 10.  Bowl


With the card free from the bowl I placed the parts under an upturned beaker while I contemplated the repair.

I figure I need to inspect and clean everything.  Ethanol and cotton buds will be the preferred cleaning method although I may polish the brass components (including the inside of the bowl) with metal polish followed by a thorough wash in ethanol.

Then I’ll need to identify an appropriate adhesive to reattach the luminous maker to the card. It needs to withstand long term immersion in ethanol.

I’ll also need a replacement bedding adhesive for the glass to bowl seal, again resistant to ethanol.  A liquid tight seal is desirable but I figure that the adhesive also acts to seat the glass evenly on the bowl which reduces the potential for stress cracking.

And I’ll need a putty replacement.  Regular white oil based putty should be adequate.

I’ve settled on a two part clear epoxy resin for the card and for the body seal.  This has excellent resistance to ethanol at anticipated ambient temperatures.  However it does soften above about 180C and will be attacked by acetone to allow future glass removal if required.  Epoxy is also relatively viscous allowing controlled application and it has a relatively long cure time allowing parts to be aligned and secured before the adhesive cures.  Epoxy also forms strong glass to metal bonds.



    During  inspection the North luminous marker fell from its recess on the yoke (internal to the bowl).  The adhesive had degraded to a spongy white goo.  The marker is still in good order and will need to be re-secured to the yoke recess.

    The inside of the bowl has a thin white film covering all surfaces.  I figure that this residue is from the failed luminous marker adhesive.  It wipes off quite easily with a cotton bud.  There is also a lot of flux reside on the inside of the bowl from the original soldering of the bellows and the yoke to the cast brass body.  This will be removed.

    The main hinge on the compass body is a little too tight and will need to be lubricated or adjusted or worked free.  Do not lubricate the hinge under any circumstances!  If it is too loose then the lid will not sit vertically for sighting.

    A luminous marker is missing from the cover sight and I suspect that this was never fitted.  I’ll need to make a new marker for the cover sight.

    There are burrs on the bowl face beside the eight mounting screw female threads.  This is a typical deformation caused by tap formed threads.  The burs need to be removed.

    The file cut that I had made to remove the dial needs cleaning up and polishing.

    The luminous markers are no longer performing as well as they once did. They are no longer very bright and have hardly any persistence.  I tested the originals against activated strontium aluminate and they need to be replaced.   This will require some additional work.  I have ordered some activated strontium aluminate powder to make the new markers with.  Note that the sintered composite is soluble in water but this shouldn’t be a problem because it will be embedded in epoxy and the card and bowl markers are immersed in ethanol.

    The dial is slightly eccentric.  I figure it has a good knock at some point in time (not really surprising given that the compass was used in military service).

    There are a few paint abrasions on the body (specifically the dial clamp mount) and some paint chips on the thumb ring.

    Everything else looks in good order and there are no significant scratches or dents in the metal work, or scratches or chips in the glass.  The card pivot is clean and sharp, located centrally in the body and vertical.


Clean Up

On with the cleaning of the brass bowl, bellows and yoke using metal polish, cotton buds, a needle to clean recesses, and small linen rags.  The bowl came up nicely including removal of the flux residues from original manufacture.



    Figure 10.  Clean Bowl
    (Note slight flat at 10 O’Clock from my file cut.)



Figure 11.  And Clean Bellows
(The cast body will not take a high polish.)


I progressed on to the dial ring.  Care is required to avoid damaging the painted graduations.  I protected them with masking tape.  Do not polish the the inside of the brass that presses over the dial retaining ring.  This has tight tolerances and polishing will prevent the dial ring from staying on the dial retention ring.  If the fit is too loose due to wear then the dial retaining ring will require re-anodizing (which increases its diameter by up to 0.1 mm).



Figure 12.  Masking Tape to Protect Graduations



Figure 13.  Clean Dial Ring


The dial retaining ring is anodized aluminium.  It was cleaned with detergent and distilled water taking care not to polish the anodized surface.



Figure 14.  Clean Dial Retention Ring


The card was carefully cleaned with ethanol and cotton buds, removing a white film, the adhesive remnants where the luminous marker used to sit, and some stray adhesive globs that had attached to the back of the card.  The various brass components and the magnet were not polished to avoid damaging the card surface.  The needle bearing recess was flushed with ethanol using a syringe to ensure it was free of dust and dirt.  The clean was not perfect as there is still white residue around the perimeter of the brass mounting boss.  I don’t want to remove this because it will change the card to magnet alignment.



Figure 15.  Clean Card


Card Balance

With everything clean it was time to investigate the card balance issue.  The luminous marker weight is insignificant and makes almost no difference to the level of the card when placed in its original location.  I placed a small length of coiled solder on the card to estimate the required weight to level the card.



Figure 16.  Solder Balance Weight on Card North Index


There is a small brass counterweight attached to the magnet on the rear of the card.  I figure from my solder experiment that the brass counterweight is actually at the wrong end of the magnet!  Note that turning the magnet around won’t fix the problem and will make the card orientate South instead of North.



Figure 17.  Card Brass Balance Weight at South (6 O’Clock)


I removed the magnet from the card assembly (two brass screws), gently pressed out the brass counterweight, moved it to the opposite end of the magnet, pressed it back into place and secured it with a single drop of cyano-acrylate adhesive from the tip of a needle.  The magnet was refitted to the card assembly which now sits almost perfectly flat in the bowl.



Figure 18.  Card Brass Balance Weight Moved to North (12 O’Clock)



Figure 19.  Card Sitting Central and Level


I can’t proceed too much further until I can make some new luminous markers.  However I did give the case a good clean and foolishly lubricated the cover hinge with CRC and worked it a few times.  While the hinge action is now positive and the cover closes flush with a clean snap action, the cover no longer sits vertically.  Do not lubricate this hinge - simply work it free.  I cleaned the oil from the hinge and re-pressed the hinge pin to tighten it back up.

I’ll be putting this repair aside for a week or so until the activated strontium aluminate arrives for the new markers.

The luminous powder from TechnoGlow arrived today (just a few weeks after my last post on this page).  It will take a few days to sort out making the new markers for the M73 but here is a photo showing just how effective this glow stuff is after just a minute of exposure to a florescent lamp.



Figure 20.  Activated Strontium Aluminate


The persistence is excellent (hours).  In Figure 20 the original compass markers are between the EcoGlow strip (which is understood to use the same technology) on the right, and the 450 g bag of ‘Green Glow in the Dark & UV Powder’ on the left.  The original markers glowed dimly for just a few seconds before fading into obscurity.  Also note the good contrast of the writing on the bag.  The M73 compass uses the markers for either alignment or contrast.  This stuff will be a significant improvement over the original markers.

Time to experiment with making the markers.  First I 3D printed a mask.  A late thought was that I could incorporate a reference line in the print so I made a second mask with the additional feature.



Figure 21.  Two Masks Taped to Polypropylene (PP) Sheet
Masks are 0.8 mm thick


The masks were cello-taped to a piece of PP sheet because epoxy doesn’t stick very well to this material.  The masks are expendable and if necessary I will cut them clear of the epoxy.

Next it was off to the local hardware store to purchase some ultra-clear two part epoxy.  This has a short working time of perhaps 5 minutes so I’ll need to work quickly.  I added about one third powder to the epoxy by volume and gently stirred it together trying to avoid crushing the luminous powder and getting air in the mix.  I’m using wooded sticks to stir with to avoid any possibility of metallic contamination of the powder.  The final paste felt quite gritty.  Note that the powder is actually relatively heavy compared with epoxy so the final face of the markers is against the PP.  If the powder settles in the epoxy it will be at the visible face of the marker.



Figure 22.  Mixed Paste Remnants


The paste was scooped into the mask recesses using a small stick and roughly levelled.  The mask was then covered with plastic cling-wrap and a microscope side, and topped off with a weight.



Figure 23.  Filled Masks with Cling-Film and Glass Slide



Figure 24.  Weighted Masks


After 30 minutes the masks were peeled off the PP and cling-wrap.  The results are not perfect as there was a bubble in one of the strips and my mixing doesn’t look like it was thorough enough.  I’ll leave these to cure for 24 hours before cutting the markers from the mask, measuring them and seeing how they stand up to alcohol immersion.

I’m already contemplating a further attempt, perhaps with slightly thicker markers (1.25 mm as opposed to 0.8 mm) and maybe using a slow cure epoxy.



Figure 25.  Filled Masks Glowing in a Darkened Room


The masks were trimmed away from the new markers using a razor blade.  The markers were still slightly soft and flexible which is not what I was expecting.  However this made cutting quite easy.  They need to be handled with care in this soft stage as they will pick up finger prints and dirt.  They  can be bent (and conversely straightened).  They hardened up over the next few days.

The long thin makers were not very successful but these were readily replaced by cutting one of the spare larger rectangular markers into strips.  The use of cling-wrap as a backing resulted in an uneven upper surface.  I’ll be replacing the cling-wrap with Sellotape on a glass slide.

The markers actually perform much better once they are cut from the mask as the sides are then also exposed to energising light.  I placed my first-attempt markers on the compass (no adhesive as yet) to see how they look.  The case and card markers could be slightly thicker.  The dial marker could be fractionally longer (the original was also short) and its North index (3D printed filament) could be slightly thinner.  The through-card illumination (mounted on the bowl yoke) did not work very effectively as it is in a poor position for light charging.

Note that there is no dial marker on the original compass.  While this is not necessary for operation at night I am going to add one.



Figure 26.  Markers on Compass
(Original markers at bottom centre.)


I’ll be remaking the masks and mixing up another batch of epoxy and glow powder using slower setting marine epoxy.  This time I’ll be weighing the epoxy and powder to achieve an 80% w/w mixture.

I made two more sets of markers, one 0.8 mm thick and the other 1.25 mm thick, each with 0.25 g of epoxy resin, 0.05 g of hardener and 0.23 g of glow powder.  While the epoxy constituents were weighed, counting drops (5 of resin to one of hardener) provided almost exactly the right ratio.



Figure 26.  New Markers


After 24 hours the markers were cut from the mask with a safe edge razor blade. The new markers are certainly superior to my first effort.  The thick markers do not seem to perform any better than the thin ones.  I now have a bunch of good replacement markers and some offcuts so I can experiment with my adhesive technique.  The card and dial markers need to be placed with good alignment and without getting everything covered with adhesive.  While the markers were initially quite soft and flexible they hardened up over a couple of days.  They can still be cut with a razor and the edges can be sanded flat and true to size.

The new markers have been carefully cut and sanded to size, and adhered into position using slow setting marine epoxy resin.  Note that the markers that fit in recessed require that the substrate be smooth, clean and preferably reflective.  Epoxy was applied to the base of the recess and the markers were gently pressed into place making sure that there are no air bubbles beneath the markers.   Any excess epoxy was wiped off with lint-free tissue.  The surface mounting markers on the card and the dial glass were pressed into a smear of epoxy and then placed in position.  The epoxy will need a few days to set solid but the new markers look great.



Figure 27.  Markers in Action


Before I can set the bowl glass I need to make sure that the markers all align correctly.  This requires gentle bending of the pivot yoke (vigorous or extreme bending risks breaking the yoke solder joints) while ensuring that that the pivot needle is vertical.  Alignment by eye is difficult due to parallax, and any ferrous tools near the compass make this task more difficult.  A razor blade within 100 mm of the side of the compass causes several degrees of card rotation.  This is why using a compass near large metal objects is problematic.

Make sure that the dial glass N datum line aligns exactly with the centre of the pivot and the tip of the small notch soldered to the inside of the South side of the bowl.  This is absolutely critical to the repair.  Check it again.

With the markers secured and everything in alignment it’s time to give everything a final clean and adhere the glass bowl cover to the bowl.  I’m going to be using an integrated circuit lifter (in essence a rubber suction cup with a controlled release) to manipulate the glass.  I won’t need too much epoxy adhesive for this operation (about three drops of mixed slow setting marine epoxy).  With the card removed from the bowl, the epoxy was spread evenly around the glass seat with a wooden splint (actually a bamboo kebab skewer) taking care to avoid getting epoxy into the bowl.  With the card still out, the glass was fitted and rotated a couple of times to ensure the epoxy was spread evenly.  The glass was removed, the card was then place on the pivot, and the glass was finally fitted and aligned.

The compass was then set aside under an upturned beaker for a few days for the epoxy to fully cure.

From time to time I manage to get stuff wrong.  While it would be easy not to mention my mistakes here and present a clean error-free procedure I’d hardly be sharing an honest learning experience.

Fools’ learn from their own mistakes.  Wise men learn from fools’ mistakes.

I started assembling the compass today and proceeded as far as filling the bowl with isopropyl alcohol and fitting the bowl in the case.  One look through the optical card reader and it became apparent that I had fitted the bowl glass exactly 180 degrees out of alignment!  The only solution is to strip everything down, remove the glass using the stove top procedure above, re-clean everything and start the assembly again.



Figure 28.  Dial Glass Adhered Beautifully
(and exactly 180 Degrees out of Alignment!)


The stove top procedure for removing the bowl glass worked very well with no heat damage to any parts (including the new markers).  The epoxy was then scraped off the bowl glass and the matching recess with the back end of a scalpel blade, polished and cleaned.  I am ready to begin adhering the bowl glass again.

Note that I have decided to use isopropyl alcohol as opposed to ethanol for the damping fluid.  While the freezing point of isopropyl alcohol is somewhat higher than ethanol  (cf. -88 and -114C) I don’t expect that the compass will ever be subjected to these extremes.  The reason for the change is that ethanol forms a binary mixture with water at about 95% (which can only be reduced by distillation from benzene).  I am concerned that the water may dissolve any exposed activated strontium aluminate in the markers

The bowl glass has been re-adhered to the bowl with correct alignment (within 2 Mil or 0.1 degrees).



Figure 29.  Vacuum Lifter for Handling the Bowl Glass
(Note idiot’s marker on the bowl to ensure correct alignment this time.)



Figure 30.  Bowl Glass in Place
(with no bubbles or defects in the epoxy.)


I set about final assembly today.  This starts with putting a new O ring on the filling plug and filling the bowl with isopropyl alcohol using a syringe.  Keep the fill port vertical and fill to overflowing.  Gently depress the bellows a few times and tilt the bowl slightly to the left, right, forward and back to ensure that any air bubbles migrate to the filling port.  Top up to overflowing again before screwing in the plug with a new O ring.



Figure 31.  Fill Plug and New O Ring



Figure 32.  Filling with a Syringe



Figure 33.  Full to Overflowing with Port at the Top


Now for the glass retention ring.  Make a 120 mm long cylindrical roll of putty (I used white BostiK plumbers putty seal) about 2 mm in section diameter.  This stuff is pretty sticky so roll it on some grease proof paper with very clean hands.



Figure 34.  Putty Sausage on Grease Proof Paper


Mould the putty into the recess between the bowl and the bowl glass and press it into place.



Figure 35.  Putty Pressed in Place on Bowl Assembly


Now orientate the dial retention ring and press it into place.  Excess putty will slowly ooze onto the bowl glass.



Figure 36.  Dial Retention Ring Pressed onto Bowl Assembly


Now fit the eight aluminium mounting screws and progressively tighten these in diagonal pairs to keep the mating surfaces parallel.  The screw heads must sit below the dial retention ring surface without protruding burs that might scratch the dial glass.  If any screw head are sitting proud then remove them and make sure that the seats are clean and free from putty.

Now remove the excess putty from the glass without undercutting the dial retention ring.  Note that a ball of putty can be used to pick up putty from the glass.  You should be left with an thin even putty line between the glass and the dial retention ring.  Clean the bowl glass to remove any putty residues.



Figure 37.  Thin Putty Line Between Dial Retention Ring and Bowl Assembly


Now fit the dial ring.  Make sure that the clamp is slackened off, align it with the dial retention ring and simply press it into place using moderate hand pressure distributed around the circumference..

Now insert the bowl assembly into the case.  Align the notch in the bowl with the dial clamp, insert the bowl assembly into case, and rotate it slightly to align the bowl assembly retention screw holes.  Then fit the two tiny screws.



Figure 36.  Bowl Notch Aligned with Dial Clamp



Figure 37.  Bowl Retention Screw Holes Aligned


Now is a good time to check out the function and cleanliness of the optics.  While holding the compass level and looking through the sight you should see the current heading in Mil.  Sighting North you will also see the pip mark on the bowl that should align within one graduation of 6400 Mil.



Figure 38.  Optics All Good and Aligned


We are almost done but the only way to clean the cover glass effectively is to remove the protector bars.  The glass is seated between the top of the cover lid and a rubber gasket.  Note that the painted datum line on the glass is on the top surface (under the protector bars).  Remove the three protector bar retention screws, remove and clean the glass and reassemble.  The datum line needs to aligned with the bowl glass datum line by rotating the cover glass and utilizing the play between the glass at the lid recess.



Figure 39.  Lid and Protector Bars Removed for Cleaning the Cover Glass



Figure 40.  Rubber Cover Glass Seating Washer


While the repair is essentially complete I have decided to remove the new dial glass marker.  The reason for this is that it actually glows down onto the card causing an annoying reflection in darkness.  In any case the new marker isn’t actually necessary for using the compass at night.  I pried of the dial ring and carefully cut the marker down in layers using a scalpel.  The final thin layer was gently peeled from the glass take care not to damage the painted rectangle.  The dial glass was re-cleaned before being re-assembled onto the dial retention ring.  If nothing else this exercise demonstrated how well the new markers are adhered.



Figure 41.  New Markers in Darkened Room



Figure 42.  M73 Compass Repair Complete


Although there is still some minor cosmetic paint damage on the compass, and in particular on screw heads, the thumb ring and at the base of the dial clamp, these are signs of fair wear and tear and I’ll be leaving them as they are.

Job done, but if you have any critical comments or suggestions regarding this repair than I’d love to hear from you.  Click here to email me.

[Turtles are Here] [Site Map] [750Z Housing] [Construction] [A20 Housing] [Drawings] [Galapagos] [Data Logger] [Making PCBs] [PCB Drill] [CD Welder] [Current Monitor] [Proportional Controller] [Lock In Amplifier] [NAD C 520] [Kenwood Tape Deck] [Tape to Computer] [Reflow Oven] [Designed to Fail] [M2 3D Printer] [Coil Winders] [Changing a Light Bulb] [Squash Racquet Repair] [Dive Photos] [Sand Casting] [Oxy-MAPP] [Lathe Maintenance] [Thermal Adhesive] [M73 Compass Repair] [Calcium Block for Turtles] [Infinite R Network] [ST7820 LCD] [Northern Lights] [Links] [Contact Me]