Shutter body — also “the crate” (6, Fig.128)

DSCN6415

The brass or aluminium-alloy box that the whole curtain assembly sits inside. It carries the frame window and the restrictive gear, and bolts to the assembly plate with three screws. Everything else in this glossary is either inside it, bolted to it, or driving it.

Kuni and Visser both call it the shutter crate. Same thing.

Frame window

DSCN6415

DSCN6419

The 24×36 mm brass rectangle inside the shutter body — the hole your picture is taken through. The curtains travel across it.

The gear at the lower edge of the second frame is the restrictive gear. Those two parts sharing a frame is the whole shutter in miniature: the hole, and the thing that decides how far the curtains may travel across it.

Assembly plate (3, Fig.116)

DSCN6280

The long flat plate the shutter body screws up against. The drum and both bushing sleeve shafts pass through holes in it; the gears and the speed mechanism live on and around it. When Maizenberg says “connect the shutter body with the assembly plate” he means the single most awkward move in the whole reassembly — expect several attempts.

The two curtains (5 and 8, Fig.129)

DSCN6273

Two light-proof cloths, originally vulcanised silk. They are not interchangeable and they have proper names:

The bushing sleeve curtain — the first curtain, the short one. Glued to a bushing sleeve. It is the one that moves first when you press the release.

The drum curtain — the second curtain, the long one. Glued to the drum. It follows, after a delay set by the speed mechanism.

You will see all four names used for the same two objects — first/second, short/long, sleeve/drum. They pair up: first = short = sleeve, second = long = drum.

Javier's note on material: originals run 0.3–0.45 mm thick; he uses 0.2–0.3 mm and considers the thinner cloth an improvement. Replace both even if only one has failed — two different cloths in one shutter behave differently.

You measured both originals. The figures are in the notebook and are set out in section B below. Nothing else in the record comes close to being that useful.

Curtain drum (7, Fig.127; 11, Fig.129)

DSCN6276

The tube the long curtain is glued to, carried on a shaft with two pulleys. Winding the shutter reels the long curtain onto the drum.

There is a hole through the drum wall that the coupling pin drops into. Maizenberg says it comes three ways: one round hole, one elongated hole, or both. Which one yours has is still unsettled. The frame above shows a single round hole — but on this camera the drum is captive, so that tube may be a bushing sleeve rather than the drum. Check it by eye before you rely on it. The notebook does not settle it either: at DSCN6276 you were logging “tiny washer” and never named the tube.

Why the drum is captive, in your own words. The notebook, at the end of the roller work: “I had to give up trying to remove the main drum and the sprocket roller. There's brass plugs in the set screw holes making it impossible to remove them. I lost the witness mark on the main drum. I just started cleaning the main drum in place.”

Two things follow that the recovery plan had been treating as guesses. The plugs are in the set screw holes, and there are two of them — main drum and sprocket roller. And there was a witness mark: it went to the cleaning rag, not to the teardown. That is worth knowing because it means the mark was a mark, not a scratch you imagined, and Maizenberg's absolute substitutes are replacing something real.

A discrepancy to resolve at the bench, not now. For an elongated hole, the recovery plan says to leave only the coupling pin's width above the pulley disk; the Maizenberg text as scanned reads as positioning that part of the hole under the rest of the pulley disk. Look at his Fig.133 and Fig.134 and trust the drawing. This only matters if your hole turns out to be elongated.

Bushing sleeve (3, Fig.129; Fig.130)

DSCN6277

A roller with a spiral spring inside it. There are two, and with the drum that makes three tubes in the shutter — which is what you are looking at in DSCN6275.

One sleeve has the short curtain glued to it. The other has the drum curtain's ribbon ends glued to both of its sides; that one is slightly narrower, which is how you tell them apart.

The spring's inner end is fixed to the shaft, the outer end to the sleeve. That is why winding stores energy: the sleeve turns against its own spring. To take a sleeve apart you remove the side wall with the spring end attached — but there is no reason to, and Maizenberg warns against flushing a spring and reassembling it dry.

Two short tubes and a small spacer come out of the opening-curtain sleeve. Yours are photographed in DSCN6274. Do not lose them — the spacer sets that sleeve's height.

DSCN6274

But check that identification. The recovery plan's Phase 0 assumes those loose parts came from the opening-curtain sleeve. Your notebook at that frame says something else: “some random parts fell off. I think they're part of the upper end caps of the smaller rollers.” Sleeve spacer and roller end cap set different things — sleeve height versus roller endfloat — so decide which it is with the parts in front of you before you fit them.

Lath (6, Fig.129)

DSCN6275

The metal strip across the free end of each curtain — the leading edge. The two laths overlapping is what makes the shutter light-tight when it is closed, and the gap between them is what makes the exposure when it is running.

Two rules that come up later. A warped lath opens a light gap, so check both are flat. And the long curtain's lath must cover one half to three quarters of the short curtain's lath — that single measurement is the target that replaces the reference marks you never made.

Kuni makes replacement laths from a coping saw blade: grind the teeth off, cut to length, grind notches for the ribbons.

Ribbons — also “tapes” (7, Fig.129)

The two narrow straps at either end of each lath. They are what actually pulls the curtain across. Their loose ends attach to the pulleys; the drum curtain's ribbon ends are glued to both sides of a bushing sleeve.

DSCN6275

DSCN6423

Kuni calls them curtain tapes; the Zorki 4 manual calls them ribbons; your notebook calls them shutter straps. Four names, two pairs of straps.

Lengths, measured off yours, are in section B. All four are 100 mm — one length, four pieces, nothing to keep straight.

The figure-of-eight rule. Three sources agree, so treat it as settled: the second (long) curtain runs over the first curtain's lath, and the second curtain's ribbons run under the first curtain's sleeve shaft.

Pulleys and the pulley disk

DSCN6421

Two pulleys ride on the drum shaft. The ribbons run over them, which is what keeps the curtains moving smoothly instead of snatching. Larger pulley towards the bottom of the shutter body, smaller towards the assembly plate.

The pulley disk is the flange beside them. Part of it has been cut away, and the remaining piece is a landmark: setting the drum so the drum hole sits above the remnant is how you find the wound position without any marks. Watch during winding for a ribbon climbing off a pulley towards the gear teeth — if you see that, stop.

The slit

Not a part — the gap between the two laths as they cross the frame window. That gap is your shutter speed. About 40 mm wide at 1/25 sec, down to roughly 3 mm at 1/500. Every tensioning and timing adjustment later is really an argument about the width of this gap.

No photograph — the slit only exists in a shutter that is assembled and running, and this one has been in pieces since 7 August. Maizenberg draws it; you will see it for the first time in section 8.

Restrictive gear (1, Fig.128)

DSCN6419

The gear inside the shutter body that decides how far the curtains are allowed to travel. It has a groove machined in it, and a screw in the body wall runs in that groove; when the groove runs out, the movement stops. A ledged screw is its shaft, with a locking nut on the far side of the body to stop the screw working loose.

It is the middle link in the chain: release gear » restrictive gear » curtain pulley gear. Check it turns freely in the empty shutter body before the curtain assembly goes in — it is awkward to reach afterwards.

Curtain pulley gear (6, Fig.127)

DSCN6421

The gear on the pulley side that meshes with the restrictive gear.

This frame is worth more than it looks. It is the record that the pulley-gear-to-restrictive-gear mesh was never broken on this camera, because the drum never came out — which is what turns section 7 from reconstruction into verification.

This is the gear Maizenberg has you shift by one or two teeth when the cam is not centred. That adjustment is the whole recovery method: you are not copying a lost mark, you are converging on it. The restrictive gear does not move — only this one.

Release gear (3, Fig.131)

Pressing the release button pushes its shaft against this gear, disengaging it and letting the curtains run. Grease its spindle on reassembly so it stays put instead of dropping out while you are trying to mate the body to the plate.

DSCN6262

DSCN6263

DSCN6264

The release button itself, which the notebook names simply “shutter button”, with its screw and spring alongside. (Corrected 19 Aug: this page and INDEX.md had both put the release button at DSCN6246, read from the photograph alone. The notebook has 6245–6246 as the winding shaft assembly.)

Coupling, with its pin (Fig.135)

A short fitting that turns in a bushing in the assembly plate and carries a pin. The pin drops into the hole in the drum. That is the only thing connecting the drum to everything on the outside of the plate.

Two things to remember when the time comes. Turn the coupling so its pin is at the top before you bring body and plate together. And if you wind the shutter afterwards and only one curtain moves, the pin went into the drum hole below the pulley disk instead of above it — separate them and do it again. That is a known failure with a known fix, not a disaster.

No frame in the set shows the coupling. It is Fig.135 in ocr/Maizenberg-Russian-Cameras-searchable.pdf, where the PDF page number is the printed page number.

Cam 7 (Fig.126)

Sits on the outside of the assembly plate, fitted on the square end of the coupling, so it turns with the drum. It works against a lever.

Cam 7 centred in the lever slot is your pass/fail test. Wind the shutter, refit the lever, look. Centred means the drum and pulleys are in the right relationship and you may tighten the third body screw. Not centred means shift the pulley gear a tooth or two and try again. If the cam is loose on the coupling square it cannot be adjusted at all — check that first.

No frame in the set shows cam 7 — it sits on the outside of the assembly plate, and the teardown photographs work from the other side. Maizenberg Fig.126.

Exposure disk (4, Fig.126)

A disk fixed on the curtains drum shaft, pierced with one hole per shutter speed.

This is why the mashed plugs are worth leaving alone. The disk and the drum are fixed to the same shaft, so the angle between them is your speed indexing — and on this camera it has never been disturbed. Drilling the plugs out would throw away the last original reference you still have.

No frame identifies the exposure disk with certainty. One candidate worth checking with the part in your hand: DSCN6244, a black disc with a keyway that your notebook records as having “dropped out from under the top plate”. If that is it, the frame is worth more than this paragraph. Maizenberg Fig.126.

Exposure regulator (1, Fig.126)

Turns with the exposure disk, with a pin seated in one of the disk's holes — which hole is which speed. At its far end another pin strikes the lever, which releases the cam and with it the second curtain. That is the delay: the gap between the first curtain going and the second being let go.

Held by a screw with a spiral spring underneath, so if you ever undo it, hold it — the spring will fire the screw across the room.

No frame in the set shows the regulator. Maizenberg Fig.126, alongside the exposure disk it turns with.

Engraved speed knob (8, Fig.126)

DSCN6226

The knob you actually turn to set a speed. It screws onto the exposure lever and is fixed by an M1.7×0.35 set screw.

That set screw is the contingency worth knowing about. If at the end the knob's index will not line up with the arrow on the accessory shoe, the standard fix is to slacken it and rotate the knob — and a mashed plug blocks that. It should not arise here. If it does, drilling comes back on the table; not before.

Set the knob against B when checking the index, because B is the easiest speed to judge by eye.

Shield — the protective set square (10, Fig.127)

A thin guard inside the shutter body, held by the same two screws that fasten the release spring. A bent edge on it is a leading cause of jammed ribbons and of curtain coating scraped off. Straighten it before assembly, and when it goes back, its edges must touch neither curtain.

No frame in the set shows the shield on its own. Maizenberg Fig.127, part 10 — worth looking up before section 2, since straightening it is on that list.

Shaped nuts (1, Fig.132) — left-hand thread

They hold the bushing sleeve shafts. They tighten anticlockwise. Turning them the way your hands expect will undo them, and if you lean on it, damage them. Worth a note on the bench.

No frame shows the nuts clearly. Maizenberg Fig.132, part 1. The shafts they hold are the ones in DSCN6277.

Frame counter

DSCN6257

The disc engraved 0 to 35 in fives that sits under the film wind knob, held by a circlip. Nothing to do with shutter speeds — worth saying, because at a glance it looks like a speed scale.

Vulcanite

DSCN6214

DSCN6228

The hard pebbled black covering on the body shell. Brittle with age and unforgiving of prying.

Yours is intact, and the first frame is the record of that. Your notebook calls it Step 0 — The Camera. It is the before photograph, and the standard the finished job gets held to.

The sneaky screw — your name, and it earned it

DSCN6419

Not a Maizenberg term. At DSCN6419 the notebook records “screw & a sneaky screw & little washer”, with the bracketed detail that matters: the little washer goes on top of the gear. A second, non-obvious screw next to an obvious one, and a washer whose side is invisible once assembled.

This group is already back in the camera. The notebook's cleaning pass records it reinstalled — gear, hub screw, sneaky screw and washer, lightly oiled, with a touch of Loctite on the threads. It is the one sub-assembly that is not in a numbered bag. When the inventory comes up short by a gear and two screws, this is where they are.

Lens register

The distance from the lens mount flange to the film. 28.8 mm, ±0.02. Set the camera-side distance 0.03–0.05 mm larger than the measured lens working distance to allow for the film bowing. Adjusted with paper shims under the lens mount, measured at all four mount screws.

DSCN6217

DSCN6218

Your shims — if there were any — are in these two frames, taken as the mount came off. Visser found his had worked loose by reassembly time and glued them down first.

Register comes before rangefinder calibration, never after. And always set the lens to minimum focus before mounting or removing it.

First curtain — short, bushing sleeve

Cloth — 72 mm long × 33 mm wide.

Ribbons — 100 mm, both sides.

The width was written 34, struck out, written again, struck out, and settled at 33. Take 33 as the reading and confirm it against the original before cutting.

Second curtain — long, drum

Cloth — 101 mm long × 33 mm wide.

Ribbons — 100 mm, both sides.

Same 33 mm width as the short curtain, which is what you would expect: both cross the same 24 mm frame window with the same overlap either side.

All four ribbons are the same length

100 mm, every one. That is the reading off the recovered silk ribbons themselves, and it is the one to cut to. One length, four pieces — there is no left and right to keep track of.

The notebook sketch reads differently, and it is worth knowing why before you second-guess yourself mid-cut. It has the short curtain at 100 and 105 and the long curtain at 101 and 101 — a 5 mm outlier on one ribbon and a millimetre of rounding on the pair. Measuring a limp silk ribbon under slight tension will do that. The parts win.

The cloth figures above have not been re-measured. They still rest on the notebook alone, so check them against the originals too while you have the parts out.

Backed out four turns

At DSCN6250 the notebook reads “lock screws for roll tension”, with a sketch of the two screws and a curved arrow, and underneath: “backed tensioners out @ 4 turns”.

This is the one teardown step that can be exactly undone. Four turns back in and the curtains sit at the tension the camera actually ran at before you opened it. Everything the sources say about tensioning — add a turn if the second curtain lags, take one off if the slit chokes at 1/500 — you now apply as a correction from a known starting point rather than as a search from zero.

Wind them in with the shutter uncocked and count out loud. Four turns, then stop and test before touching them again.

How the camera was left

At DSCN6229: “Note: shutter set to 1/25”. At DSCN6234: “shutter knob @ 1/25 when cocked”.

So the speed mechanism has been sitting at 1/25, cocked, since 31 July. That is the state the exposure disk and regulator are indexed in, and it is the speed to set the knob to when you first check that the mechanism still agrees with itself. (For the accessory-shoe index check, use B instead — easiest to judge by eye.)

It also means the curtain springs have been under load for three weeks. Not a disaster on a shutter this simple, but expect the first few cycles to be sluggish and do not read a slow 1/25 as a fault until it has been run twenty or thirty times.

WD-40 is in the winding knob

At DSCN6247: “I sprayed some WD-40 into the set screw hole because it's stuck.”

It worked — the knob came off. But WD-40 creeps, it is not a lubricant, and it will carry grit into the wind mechanism and eventually onto a curtain. Degrease that area before final assembly and oil the shaft properly.

The rangefinder shim is loose somewhere

At DSCN6254: “rangefinder and a little shim that came off early in the process.”

Early in the process, so it is not in the bag it belongs to. Find it before you start on the rangefinder — a missing shim there shows up as a calibration you can never quite bring in, at one distance or the other.

Rangefinder shim located

All four eyepiece bezels are already marked

The notebook records the front left as “slightly damaged it” and the other three as “maybe a slight mark from the pliers”, at DSCN6230 through DSCN6233.

Pre-existing. When you find them at reassembly, you did that on 31 July and 2 August, and there is nothing to fix.

Two screws that do not come out

At DSCN6272, of the tensioning screws and lock washers: “the screws don't come out”. They are captive. Do not keep turning them looking for a thread to end.

And the “big important screw” (DSCN6266) that would not shift on the first attempt did come out at DSCN6270–6271 with its spring and collar. Grease it and seat it carefully going back.

Bottom plate lock (DSCN6424–6427)

screw » washer » plate + tiny stop screw, and a tiny washer on the other side.

DSCN6427 has ЗАКР (closed) and ОТКР (open) and their arrows legible — that is your orientation reference for the key.

The lens flange stack (DSCN6217–6219)

Coming apart it went: flange screws (four), flange, lens mount gasket, backing plate (bottom), backing plate (top).

Against the top backing plate the notebook has a sketch of a curved tapered strip and your own question: “which way does it go?” Still unanswered. The sketch is the only record of its shape, so work it out from the part and the seat, and write the answer down when you have it.

Register comes before rangefinder calibration — see lens register.

Three things nobody has settled

Which drum hole variant. Round, elongated, or both. DSCN6276 shows a single round hole in a tube; the notebook logs only “tiny washer” at that frame and never says which tube it is. Confirm by eye before relying on it.

Which way the top lens flange backing plate goes.

Where the DSCN6254 rangefinder shim is.

The two mashed brass plugs are not on this list. They are mashed, they are in the set screw holes of the main drum and the sprocket roller, and they do not come out. That is settled — not by a photograph, by having tried.

The camera is not in a neutral state

The shutter is cocked, and has been since 31 July — set to 1/25 (N, DSCN6229 and DSCN6234). The curtain springs are loaded.

The roll tensioners are backed out four turns from where the camera ran (N, DSCN6250). That is your starting number in section 8 — you are not searching from zero.

The sneaky screw group is already installed — gear, hub screw, sneaky screw and washer, oiled, with Loctite on the threads (N, DSCN6419). It is not in a bag. If you need that gear out for access, know now that it is threadlocked and do not mistake it for seized.

Both brass plugs stay. Decision of 18 Aug, and the reason still holds: the exposure disk is fixed on the curtains drum shaft (M 114), so those plugs are guarding the drum-to-disk angular relationship that indexes your shutter speeds — the last original reference on this camera. The only thing that reopens the question is section 9's speed-knob index, and not before.

Shutter released and the springs let down before any dismantling of the crate

Bench and tools

Drivers to fit — including a 1.2 mm for the rangefinder infinity screw (J). Two tools for tensioning: one to turn the adjusting screw, one to hold the locking nut down, because the nut will not stay put (J). Small pliers with the jaws lined in heat-shrink for the rangefinder sensor head (J).

Naphtha or lighter fluid, watch oil, sewing machine oil, silicone grease. A needle for oiling the drum shaft through its access hole (V) — that is your mitigation for the bearing surfaces the captive drum keeps you away from.

Double-sided tape, in small pieces (K). Cement — Cemedine Super X, chosen 20 Aug on Aki-Asahi's recommendation. See the adhesive note in section 3 for what that choice changes. The book answers it replaces: thin shellac, the factory material (M); rubber contact cement, Pliobond or Rugby, Javier's (J).

Fasteners. On threads use Loctite 222, the purple low-strength grade — it is specified for small screws and soft metals like brass, and releases with low-torque hand tools. Blue 242/243 is for medium-to-large steel fasteners and may want heat to break, which a camera cannot be given. The bottom plate already has blue on it, done 19 Aug; leave it, but if one of those screws ever fights you, that is why.

For witness marks, a brittle inspection lacquer — torque seal, Dykem Cross Check — cracks visibly when a fastener moves. On this camera in particular, a mark that tells you a screw has shifted is worth having.

Visser's holder. A Styrofoam block with a slit cut in it, wrapped in aluminium tape so it cannot shed particles into the camera. You will want it in section 8, holding the camera upside down (V §16). Make it now, not then.

The full kit list, covering all six projects rather than just this one, is in materials.md — lubricants and the rule against over-oiling, solvents, what not to use, and the depth gauge that lens register actually requires.

Styrofoam holder made and taped

Curtain material to hand, 0.2–0.3 mm

The two records to keep open

zorki_1_images/INDEX.md — and remember what it is for. Reassembly order is the photographs in reverse: DSCN6429 » DSCN6214. Filename order is chronological, verified on all 88 frames. When you cannot remember which way a part faced, the frame is there.

zorki_1_images/270818-zorkiNotes-transcription.md — your own notebook, keyed to the same numbers.

And keep writing. zorki-1-work-log.md has a baggie register that is still an empty table, and five open items at the bottom of the recovery plan — teeth shifted, final spring turns, register at each of the four screws. Those are the numbers the next camera needs.

Lay it all out before anything goes together

Drum, both bushing sleeves, both laths, the shield, the coupling, the restrictive gear, the release gear. One numbered bag per group, and a line in the work log for each.

Everything laid out and photographed

Baggie register in zorki-1-work-log.md filled in

Four things to find specifically

The rangefinder shim (DSCN6254) — your note says it “came off early in the process”, so it is not in the bag it belongs to. A missing shim here reads later as a rangefinder you can never quite bring in.

The two tubes and the ring (DSCN6274). The recovery plan's Phase 0 calls these the spacer and tubes from the opening-curtain sleeve; your notebook calls them upper end caps of the smaller rollers. Those set different things — sleeve height versus roller endfloat. Decide which with the parts in your hand.

The lens flange top backing plate (DSCN6219), and your own question beside its sketch: which way does it go? Work it out from the part and the seat now, while there is no hurry, and write the answer in the log.

Both laths. Check they are flat. A warped lath opens a light gap between the two overlapping laths, and you will not see it until the film comes back (M 123).

Rangefinder shim located

Loose parts from DSCN6274 identified

Both laths confirmed flat

Checks

The restrictive gear must turn freely in the empty shutter body. Verify it before the curtain assembly goes in — awkward to reach afterwards (M 119).

Straighten the shield. A bent edge on part 10 is a leading cause of jammed ribbons and of coating scraped off a curtain (M 121, 123). Do it now; it goes back in section 7 with its edges touching neither curtain.

Restrictive gear turns freely

Shield straight

Kuni's two preconditions

The in-camera method needs room, and the drum cannot simply be spun because its rotation is gear-limited. So, before curtains:

Remove the closing curtain latch shaft and spring, and the slow speeds retard rod if fitted. Kuni calls this easily met (K p.2).

Check the lath will pass around the drum. If it will not, grind a little off the ends (K p.2).

Latch shaft and spring out

Lath clears the drum

Clean and lubricate

Bushing sleeve interiors, drum spindle and groove, and the crate. Naphtha to flush, watch oil to lubricate (Z4 pp.38–43). Do not flush a spring and reassemble it dry (M).

For the surfaces the captive drum keeps you from: thin oil on the drum shaft through its small access hole, on a needle, kept off the curtain; sewing machine oil on the bottom drum shaft pivot; silicone grease on the shutter latch and the rewind (V).

Much of this is already done — your notebook records the crate, top plate, rollers, bottom plate, take-up spool and laths all cleaned (scans 7–9). This is the gap list, not a repeat.

Degrease the winding knob set screw area. WD-40 went in there on 4 August (N, DSCN6247). It creeps, it is not a lubricant, and it will eventually reach a curtain.

Drum shaft needle-oiled through the access hole

WD-40 residue cleaned out

Your numbers

Short / first / bushing-sleeve curtain — cloth 72 × 33 mm.

Long / second / drum curtain — cloth 101 × 33 mm.

All four ribbons — 100 mm.

Full provenance and the superseded notebook readings are in section B. Cloth 0.2–0.3 mm; thinner than the 0.3–0.45 mm originals, and Javier considers that an improvement (J).

A cross-check worth having

The Zorki 4 manual's generic figures, which the recovery plan carried as a fallback: long curtain 10 cm, short curtain 7.5 cm, both 3.2 cm wide, ribbons 10 cm (Z4 p.37).

Against yours: 101, 72, 33, 100. Every figure lands within a couple of millimetres of a manual you did not measure against. Two independent sources agreeing is the best evidence you are going to get before you cut, and it is a particularly good sign for the 100 mm ribbons, which is the figure that changed.

The one real gap: your short curtain reads 72 mm against the manual's 75. Three millimetres is inside the range of two people measuring a limp cloth differently, but it is the largest disagreement on the sheet — so measure that original once more before cutting, and if you are going to err, err long. Cloth can come off; it cannot go back on.

Cutting

Cut the ends at precise right angles to the edges (K). A skewed end is a warped curtain, and a warped curtain is a light gap.

Kuni's check on width, if you want to confirm 33 mm independently: the curtain width is simply the length of the central closing-curtain part of the drum, and it varies a little by model. Measure that and you have it from the camera itself.

Replacement laths, if either is past saving: coping saw blade, teeth ground off, cut to length, notches ground for the ribbons (K).

Short curtain re-measured before cutting

Both curtains cut, ends square

Four ribbons cut at 100 mm

The adhesive — Cemedine Super X

Chosen 20 Aug 2026 on Aki-Asahi's recommendation. They sell the curtain cloth, so their recommendation is the one with something at stake.

It is a one-part silyl-group polymer that cures by absorbing moisture from the air. Three properties matter here:

Solvent-free. This is its real advantage over both book answers. Shellac carries alcohol and contact cement carries a hydrocarbon solvent, and both of those are things you would rather not have wicking around rubberised silk and freshly oiled bearings. Super X brings none.

Elastic when cured — elongation at break around 200%, flexible from -60 °C up. The curtain joint has to bend around a small-diameter drum thousands of times; a joint that cures rigid is a joint that eventually cracks. Every practitioner source says the same thing: flexible, not strong.

About ten minutes of working time, then it goes pressure-sensitive. Roughly 30 minutes to handling strength. Complete cure is 24 hours — Cemedine's own test data quotes seven days for full properties.

If your tube is Super X Gold No.777 rather than plain Super X, the handling time drops to one or two minutes. Check the tube before you open it, because that is a different job.

What Super X changes about the method

Maizenberg's instruction to begin reassembly immediately, before the glue sets (M 124), no longer applies. That advice exists because shellac stays movable, so he can correct a bad ribbon position while it is still wet. Super X gives you ten minutes and then it is committed, and it does not re-dissolve. Do not follow M 124 here.

Which turns Kuni's dry run from a good idea into the method. Prove the geometry on double-sided tape, all the way through the cam 7 check in section 7. Cement only to positions that have already passed.

And cure before you tension. Tensioning loads every glued joint permanently. Give it the 24 hours. Cement in the morning, tension at the end of the day at the earliest — overnight is better.

One consequence of moisture curing: it wants ambient humidity. In a dry room it will take longer, not less time. Do not read a slow set as a bad joint.

Removal, if it comes to that, is mechanical — peel and scrape. There is no solvent that will politely undo it. Another argument for proving the geometry first.

Tube checked — plain Super X, not Gold

Curtains to laths

Glue the short curtain to its lath first — Javier's order, and the joint that needs the most strength since nothing is sewn (J).

The curtain folds so its tip faces the film side of the shutter (J). Viewed from the back on a correct assembly, the folded edges wrapped around the laths are the ones you can see.

These lath joints can be cemented properly now. It is the ends that attach to the drum, sleeves and pulleys that stay on tape.

Ends onto the rotating parts — on tape

Small pieces of double-sided tape bond the curtain and ribbon ends to the drum and rollers temporarily (K).

Wrap the closing curtain, its lath and its tapes twice around the drum after the end is attached. Drum rotation is gear-limited, so you cannot wind it on the way you would on a bench (K).

No warp, and equal distance from both edges of the drum and of the sleeve (M 124).

Only one coil ever gets cemented — the wrap does the holding, the cement only anchors (M 124, J).

The figure of eight

Three sources agree, so treat it as settled (M 118, J, Z4 p.35):

The second (long) curtain runs over the first curtain's lath.

The second curtain's ribbons run under the first curtain's bushing sleeve shaft.

Pulley orientation: larger pulley towards the bottom of the shutter body, smaller towards the assembly plate (M 118).

Figure-of-eight confirmed by eye before going further

The target

Set the drum to the shutter-wound position: the hole in the drum sits above the remaining part of the pulley disk, and the curtain lath is right next to the drum (M Fig.133, p.125).

Then position the short curtain's ribbons on the pulleys so that the long curtain's lath covers one half to three quarters of the short curtain's lath (M 125).

That single measurement is the whole specification. Maizenberg is explicit that only two marks ever mattered and that this replaces them: “do not make any other marks because the positions of the short and long curtain ribbons are not important.”

Long curtain ribbon ends onto bushing sleeve 1, with the sleeve strictly parallel to the curtain lath (M 124).

If your drum hole is elongated

Still unsettled on this camera — DSCN6276 shows a single round hole in a tube, but the notebook never says which tube, so it is an inference. Check by eye.

Round hole: nothing to decide.

Elongated: position it so only the part equal to the coupling pin's diameter sits above the pulley disk (M Fig.134).

Both holes: the coupling pin goes in the round one.

There is a documented discrepancy here between the recovery plan's wording and the Maizenberg text as scanned, over whether the elongated portion sits above or under the pulley disk. Trust the drawing — look at Fig.133 and Fig.134 in ocr/Maizenberg-Russian-Cameras-searchable.pdf, where PDF page equals printed page.

Dry run, then cement

Wind and release by hand. Watch the lath overlap and the travel. Reposition. Repeat.

Then go and do sections 6 and 7 with the curtains still on tape. This is the part of Kuni's method that earns its keep: you can take the whole camera to the cam 7 check, get a pass, take it apart again, and only then commit glue to a geometry you have already proved.

When it passes: cement the joints, remove the double-sided tape piece by piece and replace it with cement (K).

Maizenberg's advice to reassemble before the glue sets (M 124) is written for the bench method, where the dry run is not available — and for shellac, which stays movable. With Super X it is actively wrong: ten minutes of working time, then committed, and no solvent will undo it. See section 3. You have the better version of his advice anyway. Take the certainty instead of the hurry.

Then leave it alone for 24 hours before section 8. Tensioning puts permanent spring load on every joint you just made.

Lath overlap is 1/2–3/4, by eye, wound

Dry run through section 7 passed

Cured 24 h before any tension goes on

Cemented, tape removed and replaced

Installing

Turn the restrictive gear clockwise all the way and leave it there (M 118, J). Javier's phrasing: install the drum with the restricting gear at its rightmost limit seen from the top.

Turn drum and pulleys counter-clockwise to reel the long curtain onto the drum and the short curtain's ribbons onto the pulleys, exactly as during winding. Install the drum shaft in that position and link the gears (M 118).

Fit both bushing sleeve shafts and screw on the shaped nuts. Left-hand thread — they tighten anticlockwise (M 118, Z4 p.46). Leaning on them the way your hands expect will undo them and can damage them.

Don't forget the spacer and the two tubes from section 1.

Javier's check: with the drum correctly installed it should not turn clockwise any further, but should turn with the curtains going toward the spring-loaded rollers.

Mating — “perhaps the most difficult step”

Javier's words, and Maizenberg says expect several attempts. Three things have to land at once: the coupling pin through the drum hole, the restrictive gear into the advance and release gears, and both spring roller ends into their holes.

Pulleys clockwise all the way; drum hole above the pulley disk remnant. Turn the coupling so its pin is at the top before you start (M 119).

Grease the release gear's spindle so it sticks in place instead of dropping out mid-attempt (J).

Right hand on the shutter body, thumb pinning drum and pulleys. Left hand on the assembly plate.

Be very patient. That is a quotation, not encouragement.

The diagnostic that saves an hour

Wind it. If only one curtain moves, the coupling pin went into the drum hole under the pulley disk instead of above it (M 119).

Separate and repeat. That is a known failure with a known fix, and it means nothing about the curtain geometry — do not start repositioning ribbons because of it.

Both curtains move when wound

The loop

Tighten the left body screw. Fit both bushing sleeve shafts into the plate holes. Tighten the right screw. Leave the third (M 119).

Wind both curtain springs 3.5–4 turns to the left and tighten the locking screws.

Wind the shutter, refit lever 5, and look at cam 7.

Centred in the lever slot » the drum and pulleys are in the right relationship. Tighten the third body screw.

Not centred » remove the body and shift the pulley gear one or two teeth, either way. The restrictive gear does not move. Repeat.

You are not recalling a lost position. You are converging on it, and the loop tells you when you have arrived.

Cam 7 centred

Teeth shifted, and which way, recorded in the work log

Check first, before you start shifting

Is cam 7 loose on the coupling square? If it is, it cannot be adjusted at all and no amount of tooth-shifting will centre it. Rule that out before you take the body off even once.

Backlash. Drum and pulley shaft backlash should be barely perceptible, and the same between drum and pulleys. Shim with washers between the pulley gear and the shutter body if needed (M 119). Maizenberg singles out the curtain drum shaft as the one to pay attention to — speeds only adjust easily when the backlash is right.

Then the shield, and a lath cross-check

Fit the shield so its edges touch neither curtain (M 119).

As the shutter winds, the first curtain's lath should progressively overlap the second's with no gap, and when released neither lath should be visible in the gate (Z4 p.50).

Stop winding immediately if you see a ribbon run along a pulley and then turn toward the gear teeth (M 121). Maizenberg calls a ribbon jammed in the gears one of the most common shutter failures, and a deformed shield is the usual cause — which is why the shield was straightened back in section 2.

Shield clear of both curtains

Lath overlap correct through a full wind

Where to start

You have a number the books don't: four turns. That is how far the tensioners were backed off from a camera that ran (N, DSCN6250). Wind them back in, count out loud, and stop.

For reference, where the sources would have started you: Maizenberg 3.5–4 turns on both (M 119) — the factory-trained figure, and reassuringly close to yours. Javier ends up at 4–5 on the first roller and about 3 on the second, and prefers minimum tension for smoother, quieter running (J).

They agree on the shape of the answer: a snappy first curtain and a rather slow second one. Too much on the second makes it catch the first — banding, or outright capping.

They also agree on direction of travel: fix problems by rebalancing the two springs, not by cranking both up.

Adjusting without pulling the crate

Visser's route, and the reason for the Styrofoam block: access is through the shutter crate's bottom cover with the camera held upside down. Both tension screws are reachable there — second curtain top, first curtain bottom (V §16).

Remove the small stopper. Loosen the centre screw while gripping the adjustment screw with pliers. Counter-clockwise to increase tension, in quarter or half turns, re-securing the centre screw each time. Then tighten the centre screw and refit the stopper.

Javier's limit: no more than two or three turns of added tension total beyond your starting point. Past that you get noise, vibration, a heavy wind and a heavy release.

The three speeds that settle it

Adjust B first, then set the exposure disk and regulator, then the rest (M 119).

Then check 1/25, 1/50 and 1/500. Maizenberg: “if these speeds work correctly, so will all other speeds” — the shutter runs at one velocity at every speed, only the slit width changes. About 40 mm of slit at 1/25, about 3 mm at 1/500.

1/25 — watch both curtains. Either one sluggish, give that spring half a turn more.

1/50 — watch the drum curtain near the body wall. Slight overlap snags it, delays a fraction, and gives uneven exposure. Usually increase the second curtain's winding.

1/500 — look at the sharpness of the frame window's edge lines. Sharp is correct. Left border blurred » wind the first curtain's spring more, the second is catching up. Half a turn, recheck, repeat. Orientation for that call: control knobs at the bottom, release spring on top, cam 12 facing you (M 120).

Recheck all speeds after any change.

B correct

1/25, 1/50, 1/500 correct

Final spring turns on each roller recorded

Order

From here, reassembly is the photographs in reverse — DSCN6429 back towards DSCN6214 — with the notebook alongside. Two orders it records that are easy to get wrong:

Bottom plate lock: screw » washer » plate + tiny stop screw, with a tiny washer on the other side (N, DSCN6424–6427). DSCN6427 has ЗАКР and ОТКР legible for key orientation.

Lens flange stack: flange screws (four), flange, mount gasket, backing plate (bottom), backing plate (top) — reversed (N, DSCN6215–6219).

Two of Javier's traps

The lower light baffle — “the next most difficult step” after mating. It is clamped by the long release spring's retainer. Glue the retainer to the baffle so it stays put to receive the spring's bolts.

The lens mount screws are hard to start because the retainers they thread into fall into the body. Glue the retainers in with rubber cement.

Both are five-minute preparations that prevent half-hour recoveries.

The one contingency that reopens the plugs

When the top plate goes on: does the engraved speed knob's index line up with the arrow on the accessory shoe? Set the knob to B to judge it — easiest speed to read by eye.

If it lines up, you are done with the question forever.

If it does not, the standard fix is to slacken the M1.7×0.35 set screw and rotate the knob on the exposure lever (M 114–115) — and a mashed plug blocks exactly that. Then, and only then, drilling comes back on the table.

It should not arise. The drum-to-exposure-disk relationship has never been disturbed on this camera, which is the whole reason for leaving the plugs alone. Do not drill pre-emptively against it.

Speed knob index aligns at B

Lens register

28.8 mm, ±0.02 — acceptable 28.78–28.82. Measured from the surface of the pressure plate as it sits on the film rails under its two leaf springs. The indicator must not push the plate out or the reading is wrong (M 133).

Set the camera working distance 0.03–0.05 mm larger than the measured lens working distance, to allow for the film bowing inward. This is Maizenberg's refinement and Javier does not mention it (M 132).

Paper shims 0.03–0.3 mm under the mount. Measure at all four mount screws and find the position where the readings match all round; tolerance ±0.03 mm across them. Short of 28.8 » add shims; over » remove a layer. A part-shim under one screw corrects one corner — Javier raised a corner about 0.02 mm that way.

Visser's warning: he found his lens mount shims had worked loose by reassembly time, and glued them down first. Yours are in DSCN6217 and DSCN6218.

No indicator? The step-ladder method — five books as steps 5–10 cm apart, photograph from 1.25–1.5 m focused on the middle step, see where best sharpness falls. Behind the target » move the lens out; in front » remove a shim. 1 mm of index shift on the range scale = 0.02 mm of axial lens shift (M 135).

Register within 28.78–28.82 at all four screws

Before and after readings recorded in the work log

Rangefinder — four adjustment points

Infinity, horizontal. Remove the large screw beside the VF window; the tiny adjustment screw is in the hole beneath. 1.2 mm driver, minute increments, against a genuinely distant target. Image short of lining up » clockwise. Past it » counter-clockwise. Caution: some collapsible Industars travel past the infinity mark and lock there — check the pointer really is at infinity first.

One metre, horizontal — the step a Leica doesn't need. Javier is emphatic: calibrating at infinity alone does not guarantee close focus on a FED or Zorki. Camera on a tripod, drawn cross on paper on a wall, camera back exactly 1 m from the target. Focus until the images coincide and read the scale — the index should point exactly at 1 m. Off by more than 0.5 mm from the “1” and focus will be wrong from minimum out to about 3 m.

Adjust at the RF sensor head on the tip of the RF arm. Reads closer than 1 m » turn the head up. Reads farther » down. Small pliers, jaws lined with heat-shrink. Recheck infinity after every single change — moving the tip disturbs it.

Vertical, last. Left RF bezel off to reach the optical wedge collar; two notches to turn it. Often glued — alcohol or acetone on a cotton wad. It can disturb horizontal, so recheck that too.

When infinity and 1 m are both right, lock the sensor tip with a drop of shellac or cyanoacrylate where it meets the RF arm.

The handling rule that protects all of this: always set the lens to minimum focus before mounting or removing it. Mounting at infinity with the cam jutting out is what knocks the sensor off.

Infinity correct

1 m within 0.5 mm of the index

Infinity rechecked after the 1 m work

Sensor tip locked

Test roll

Cheap negative film, and shoot it in a way that tells you something.

A frame at each marked speed, even light, plain subject — looking for banding and capping

Several at 1/500, the speed that exposes tension errors first

Wide open at minimum focus, then at 1 m, then at 3 m — that is the rangefinder test, and wide open is the only way it shows

A frame at infinity, wide open

Half the roll wound on in bright sun with the back to the light — the light-leak test

Note frame numbers against what you were testing, before you forget

What the failures look like: a dark band across the frame is curtain timing, go back to section 8. Progressive darkening towards one edge is the second curtain catching the first — the same fix. Everything soft at 1 m but sharp at infinity is the sensor head, not the lens. Edge fog is the shield, the baffle, or a warped lath.