Scope / CRT Clocks · Volume 8

Build B — OSC4.4 (owned)

The owned OSC4.4 scope clock, built stage by stage: HV supply, low-voltage and wave-shaper, deflection and logic, CRT wiring, bring-up, and the Wi-Fi/GPS modules

Of the two scope clocks on the owner’s bench, the OSC4.4 is the one this series can support from the inside out, because it was assembled by hand from a bare board and a bag of through-hole parts, and because its designer ships an assembly procedure that is unusually honest about what to test and when. Where the Dutchtronix build of Vol 7 is a polished, well-trodden kit, the OSC4.4 is a builder’s board: a single PCB that carries a ~300 V mains-derived supply, a five-volt logic rail, a transistor wave-shaper, a relay that gates the high voltage, five small-signal transistors forming the X/Y deflection amplifiers, a PIC microcontroller driving a dual DAC through a pair of digital rheostats, and pads for optional Wi-Fi and GPS time-sync daughterboards. This volume reorganises the designer’s step-by-step instructions into a readable build narrative and, at each stage, explains why the circuit works the way it does, leaning on the subsystem engineering already covered in Vols 3 (the HV supply), 4 (the deflection and video amplifiers) and 5 (the timebase and vector firmware). It is written to be read with the original AssemblyInstructionsForOSC4.4 text open alongside — this is a companion, not a replacement, and the original is the authority on any point of conflict.

A word before the soldering iron is warm: the OSC4.4 derives roughly 300 V DC directly from the AC line, stored on capacitors that remain charged after the clock is unplugged, and the relay that gates that rail does nothing to make the stored charge safe. Everything in Vol 12 and the hub-wide safety brief is in force from Part 1 onward. The build sequence is itself a safety feature — it brings the supply up before any logic is installed, so the dangerous rail can be measured and trusted on a nearly empty board.

FIGURE SLOT 8.1 — The owned OSC4.4 populated board, top side, after completion: the transformer and HV electrolytics at one edge, the relay and 7805 mid-board, the IC sockets and set buttons along the logic side, and the ten CRT wires fanning out to the tube socket. Suggested source: owner build photo (staged as figs/osc44_board.jpg).

8.1 Orientation — reading the board before you solder

The single most useful thing a builder can do with the OSC4.4 is spend ten minutes with the bare board, the parts list, and the assembly text, before heating the iron. The designer says as much in the opening line, and the reason is that this board mixes a lethal mains-derived section and a harmless logic section on one piece of FR-4, with no silkscreen warning where one ends and the other begins. Knowing the orientation conventions cold means you will not have to stop mid-solder to puzzle out a footprint.

8.1.1 The four polarity-and-orientation rules

Four conventions govern every part on this board, and all four are stated in the assembly instructions:

  • Transistors. Looking at the front (the flat face), the three leads left-to-right are emitter, base, collector (EBC). Every transistor on the board faces the same way — all the emitters sit closer to the ICs — except Q11, which is reversed (its collector is the lead nearest the ICs). Q11 is the one part you can install backwards while doing everything else right, so it gets its own callout in Part 3.
  • Diodes have a + and a end; the board marks the orientation and the diode’s banded (cathode) end must match.
  • ICs and IC sockets are marked at one end with a small indented notch; the socket notch and the chip notch must agree, and on this board the notch generally points “up,” toward the middle of the board.
  • Electrolytic (cylindrical) capacitors have a + and a lead; reversing an electrolytic on a 450 V rail is how you turn a capacitor into a small grenade.

One more mechanical note that trips up first-timers: on the OSC4.4 all components are soldered on the bottom of the board. Parts are inserted from the top, their leads pushed through, and the joints made on the underside — including the CRT and AC wires, which are fed through their holes and soldered on the bottom. This is the opposite of what the instinct says, so it is worth taping a reminder to the bench.

8.1.2 How the build is staged, and why

The designer’s part numbering (Part 1, 2, 3, 4, 6 — there is no Part 5 in the text) is not arbitrary; it is a deliberate bring-up order that mirrors the subsystem stack from Vols 3–5:

  1. Part 1 — the high-voltage supply. Built first, on an otherwise bare board, so the ~300 V rail can be measured and proven with nothing fragile downstream.
  2. Part 2 — the low-voltage (5 V) supply, the wave-shaper, and the relay control. This adds the logic rail and the transistor circuit that gates HV.
  3. Part 3 — the deflection amplifiers and the logic-IC sockets. The X/Y drive transistors and the sockets for the PIC, DAC, rheostats and shifter.
  4. Part 4 — the ten CRT wires and the pinout/voltage tables for the two supported tubes.
  5. Part 6 — bring-up: jumpers to make a blob, then chips installed one at a time, each power-cycle revealing the next layer of behaviour until the clock face appears.

The genius of the order is that each stage ends in a measurable test, so a fault is caught against a known-good baseline rather than discovered at the end with everything installed.

8.2 Part 1 — the high-voltage supply

The OSC4.4’s HV section is a textbook example of the half-wave, capacitor-input, transformer-derived supply discussed in Vol 3: a small mains transformer steps the line down to a 7–9 VAC secondary for the heater, while a separate winding (or the rectified line behind the relay) develops the few-hundred-volt rail the gun needs through a diode-and-electrolytic arrangement, dropped and divided by high-value resistors to make the focus, anode, grid and brightness potentials. Built on a bare board, it can be energized and measured before anything else exists to be damaged.

8.2.1 The Part 1 component list

Install in the order the assembly text gives, noting orientation on every diode and electrolytic:

RefPartValue / notes
D11N4007 diode1000 V rectifier — note orientation
R1resistor10 Ω, 3 W (brown-black-black) — surge/heater dropper
D31N4007 diodenote orientation
C1electrolytic4.7 µF, 450 V — note + / −
C4electrolytic4.7 µF, 450 V — note + / −
P2potentiometer1 MΩ (105) — anode adjust; do not solder the wiper yet
P1potentiometer1 MΩ — focus adjust; do not solder the wiper yet
P3potentiometer50 kΩ (503) or 47 kΩ (473), top-adjust — brightness; same body as P1/P2
R3resistor470 kΩ, ½ W (yellow-purple-yellow)
neon bulbHV “live” indicator
R37resistor10 MΩ (brown-black-blue)
F1fusesolder-in fuse or holder + bus fuse, 250 VAC 0.5 A
4-pin jumper120 / 240 VAC select

The two 1 MΩ pots, P1 (focus) and P2 (anode), are installed but their wipers are left unsoldered at this stage. The reason is bring-up: the focus and anode wires (board pads F and A) are taken from those wipers, and leaving them free until Part 4 lets you tack the CRT leads on cleanly. R37, the 10 MΩ resistor, is the bleeder-class part on this rail — it is the high-value path that slowly discharges the HV electrolytics; do not omit it, and do not trust it to make the board safe instantly (a 10 MΩ × 4.7 µF time constant is tens of seconds — see Vol 12 on always discharging by hand).

8.2.2 The mains entry, the line jumper, and the transformer

The fuse F1 sits at the line entry. Solder roughly 8 inches of ~18 AWG black wire to the fuse input pad marked “HOT” (the left side, the fuse end closer to pad X1), and ~8 inches of white wire to NEUTRAL. The colour convention matters because it is the only thing telling you which board node is live.

The 4-pin jumper sets the transformer primary for 120 VAC (US) or 240 VAC. On a US build it can be a permanent solder bridge across the marked “120” holes; the header-and-jumper is only needed if the clock might travel between line voltages. Getting this wrong on 240 V with the board strapped for 120 V will overdrive the transformer; on 120 V strapped for 240 V the heater simply runs cold and nothing lights.

The transformer orientation is keyed by which side faces the fuse: the primary side (pins 1,2,3,4) sits next to the fuse, and the secondaries (5,6,7,8) line up with the holes along the board edge. This is a 120/240 VAC primary, 7–9 VAC-secondary part — the secondary feeds the heater dropper and the low-voltage rectifier of Part 2.

8.2.3 The relay temporary jumper and the first HV test ⚠

The HV rail on this board is gated by the relay installed in Part 2 — in normal operation the logic energizes the relay coil to connect the high voltage. But Part 1 has no logic yet, so to test the HV supply you insert a temporary jumper (do not solder) into the lower-left holes of the relay footprint, bridging the contacts the relay would otherwise close. The assembly text draws the exact two holes (and points to Jumper.jpg); the jumper simply fakes a closed relay.

With the jumper in place, connect the black wire to an AC cord hot, the white to neutral, and energize. This is the first live-mains moment of the build — one hand behind your back, no jewellery, and a meter, not a finger, does the probing.

Two things confirm a healthy supply:

  • The neon indicator lights — proof the rail is live.
  • ~300 VDC at C6(−) and R8(+), with the C6 lead negative and R8 positive on the meter.
  • ~300 VDC at G (next to the neon) referenced to C6(−), with G negative and C6(−) positive.

The two ~300 V readings, taken with opposite polarity at the meter, are measuring the rail from the two ends it must serve: the grid/cathode return (very negative relative to the chassis ground that the logic uses) and the deflection/anode side. That the grid node G sits near −300 V while the plate nodes will later read +150 to +250 V is the whole trick of this class of supply — the gun is biased by floating the cathode and grid well below the plate/anode rail rather than generating a single huge positive anode voltage (the approach explained in Vol 3 and confirmed by the per-pin measurements in § 8.5).

If both checks pass, disconnect the power cord and proceed. If you are not fitting the battery-backup option, skip to Part 2; otherwise fit it now while the board is still sparse.

8.2.4 The optional battery-backup sub-assembly

The battery-backup option keeps timekeeping alive across a mains outage. It is a small self-contained sub-circuit added at the end of Part 1:

RefPartNotes
Relay2DPDT relaycoil oriented toward C1 & C4
78L05100 mA, 5 VDC regulator
R10210 kΩ resistor
D2231N4001 diode
3 mm LEDsmall; note orientation
9VDC IN9 V battery leadsnote + / −; easiest to leave off until the build is finished

The 78L05 regulates the 9 V battery down to the 5 V the logic needs; Relay2 switches the logic supply over to the battery when the mains rail collapses, and the 3 mm LED indicates backup state. Note that two later parts change value when the BB option is fitted: C10 becomes the 1000 µF 6.3 V cap instead of 100 µF (Part 2), and C55 becomes the second 1000 µF 6.3 V cap instead of 0.1 µF (Part 3). Both substitutions come from the BB kit.

8.3 Part 2 — low-voltage supply, wave-shaper, and relay control

Part 2 builds the +5 V logic rail, the transistor circuit that shapes the relay-drive and blanking behaviour, the X/Y centring trim pots, and the relay itself. By the end of Part 2 the board has, in the designer’s words, “a high voltage supply, low voltage supply, wave shaping circuit and relay control circuit” — everything except the deflection amplifiers and the brains.

8.3.1 The GPS socket, the display button, and Rfilament

If the GPS module is to be fitted, its 5-pin socket is soldered behind Relay2 now, holes facing up (a custom GPS module may instead want individual flying leads, depending on its pin layout). S3 is the tactile pushbutton that toggles the display on/off.

Rfilament — the resistor next to S3 — is the heater-current dropper and is the one value you may need to tune to your particular CRT. The default is 4.7 Ω, 3 W, which suits most tubes; some European types (the text names the DG7-6 class) want 10 Ω, 3 W to bring the heater voltage down. The procedure is empirical: after the CRT is wired and powered, measure the AC voltage across the filament pins and compare to the tube’s rating (the supported tubes want 4–6 VAC across the heater — see § 8.5). If it reads high, raise Rfilament; if the filament is anaemic, lower it. The DG7-32 datasheet specifies a 6.3 V / 300 mA heater, which the 4.7 Ω dropper from the 7–9 VAC secondary lands comfortably inside.

8.3.2 The wave-shaper and relay-control parts

The wave-shaper is the small-transistor circuit (Q7, Q8 and their resistors) that conditions the relay drive and the blanking logic — it is the OSC4.4’s equivalent of the Z-axis / blanking gating discussed in Vol 4, ensuring the HV relay only latches when the logic is healthy and the beam is blanked during retrace. Install in order:

RefPartValue / notes
R38resistor2.2 kΩ, ¼ W (red-red-red)
Q72N3904 NPNorient correctly — emitter nearest the Q7 marking
R33resistor10 kΩ (brown-black-orange)
B1bridge rectifierfull-wave; note − / + orientation
R14resistor100 kΩ (brown-black-yellow)
C6electrolytic1000 µF, 16 V+; note orientation
R29resistor1 kΩ
D561N4001note orientation
R27resistor10 kΩ (brown-black-orange)
Q82N3904 NPNcollector at the Q8 marking
D91N4001note orientation
P6pot50 kΩ side-adjust — X centring
P7pot50 kΩ side-adjust — Y centring
C22ceramic0.01 µF (103)
C10electrolytic100 µF, 10 V (note orientation); BB option → 1000 µF 6.3 V
R26resistor470 Ω, ¼ W — or 10 kΩ for a dimmer display
5 mm LEDnote orientation; long lead is anode (+)
7805regulator5 V; note orientation

The bridge B1 rectifies the low-voltage secondary; C6 (the 1000 µF cap) is its reservoir, and the 7805 regulates that down to the clean +5 V the PIC and DACs run on. P6 and P7 are the X and Y centring trims — note that, like the size pots later, these side-adjust pots don’t seat all the way through their holes, so it is fine to solder them on the top of the board. R26 sets a baseline brightness floor; swapping it to 10 kΩ is the designer’s recommended fix for a too-bright tube.

8.3.3 The Part 2 test — 5 V, the LED, and the relay click ⚠

Plug in. The 5 mm LED should light, confirming the 5 V rail is alive. Then:

  • Measure +5 V at C44 — the right side of C44 is negative (ground), the left side is the +5 V positive. (C44 is one of the 0.1 µF decoupling caps that gets installed in Part 3, but its pads are the convenient 5 V test point referenced throughout.)
  • Test the relay control by carefully touching a jumper wire between V1 and the positive (left) side of C44. The relay should click, and the neon should light as the HV is connected. This is the manual version of what the logic will later do automatically — V1 is the relay-drive node the wave-shaper transistor pulls, and bridging it to +5 V energizes the coil by hand.

Touching V1 to 5 V proves the relay coil, the relay contacts, and the HV gating path all work before any chip is installed to drive them. If the relay clicks and the neon lights, Part 2 is good.

8.4 Part 3 — deflection amplifiers and the logic-IC sockets

Part 3 is the largest stage. It populates the five small-signal transistors that form the X and Y deflection amplifiers, the resistor network that sets their gain and the deflection load, the size-trim pots, the decoupling caps, and the sockets for the microcontroller, DAC, digital rheostats and shifter. This is the OSC4.4’s realisation of the differential-style high-voltage deflection drive analysed in Vol 4 — but done with discrete MPSA-series transistors rather than a dedicated amplifier IC.

8.4.1 Q11 — the one reversed transistor ⚠

Q11 (MPSA92 or KSP92, PNP) is installed first, and it is the part to get right:

Q11 is the only transistor on the board whose flat side faces the opposite way to all the others — its collector is the lead nearest the ICs, where every other transistor has its emitter nearest the ICs.

The MPSA92 is a 300 V-class high-voltage PNP; it is the high-side device in the deflection or HV-referenced section, which is exactly why its polarity is flipped relative to the NPN small-signal parts around it. Install it backwards and the deflection drive will not work and you may damage the part. After Q11, R36 (47 kΩ, yellow-purple-orange) and R35 (10 kΩ) set its bias.

8.4.2 The deflection transistors and their resistor network

The five MPSA44 NPN transistors (Q10, Q1, Q2, Q3, Q4) — the high-voltage NPN complement to Q11 — form the deflection-amplifier stages, two per axis plus the cathode/video device. They install in the assembly-text order interleaved with their resistors and the size pots:

RefPartValue / notes
Q10MPSA44 (KSP44) NPNcathode-drive / deflection stage
Q1MPSA44 NPN
R99resistor330 kΩ for DG7-32 (orange-orange-yellow); 220 kΩ for 2BP1 (red-red-yellow)
R4resistor330 kΩ for DG7-32; 220 kΩ for 2BP1
R6resistor180 kΩ (brown-grey-yellow)
R8resistor180 kΩ (brown-grey-yellow)
P4pot50 kΩ, 0.12 W (503), top-adjust, next to Q1 — X size (solder on top OK)
Q2MPSA44 NPN
R10, R9resistors10 kΩ each (brown-black-orange)
Q3MPSA44 NPN
P5pot50 kΩ, 0.12 W, top-adjust — Y size (solder on top OK)
Q4MPSA44 NPN
R13, R31, R34resistors10 kΩ each
R5, R11, R7, R12resistors4.7 kΩ each (yellow-purple-red)

R99 and R4 are tube-dependent and this is the most important value choice in the build: they (together with R6/R8) set the deflection-amplifier load and therefore the deflection gain — how many board-volts translate into how much beam travel. The assembly text gives two recipes, which the per-tube notes in § 8.5 restate:

  • DG7-32: R99 = R4 = 330 kΩ, R6 = R8 = 220 kΩ.
  • 6Lo1i: R99 = R4 = 220 kΩ, R6 = R8 = 180 kΩ.

(The Part 3 component list installs R6/R8 as 180 kΩ resistors, matching the 6Lo1i recipe; for a DG7-32 build the per-tube note raises the R6/R8 pair to 220 kΩ. Build to the per-tube note for your actual tube.) The DG7-32 datasheet quantifies why the values differ: its deflection factor is roughly 0.35–0.43 mm/V on the Y axis and 0.22–0.30 mm/V on X at 500 V g2/g4, so a given board swing draws a known amount of beam, and the resistor recipe is what matches the amplifier’s output range to the tube’s sensitivity. P4 and P5 are the X and Y size trims set during bring-up.

8.4.3 The decoupling caps and the logic-IC sockets

The five 0.1 µF (104) ceramics — C11, C222 (the one next to U4), C33, C44, C55 — decouple the logic supply at each IC; C44 is the 5 V test point of § 8.3.3, and C55 becomes a 1000 µF cap if the BB option is fitted. Then the IC sockets go in, all notched and oriented per the board:

SocketPinsHolds
U58”4132” digital rheostat
U48”4132” digital rheostat
U68”12f629” shifter PIC
U320”7528” (TLC7528 dual DAC)
U128”18f26K20” main PIC

Interleaved with the sockets are R25/R22/R24/R23 (47 kΩ each, the rheostat-network resistors), D77 (1N4001), and the trio R20/R21/R18 (2.2 kΩ each, red-red-red). Finally the two remaining set buttons install: S1 (fast set) and S2 (slow set). Together with S3 (display toggle) from Part 2, these are the three tactile buttons that drive the entire user interface described in § 8.7.

The architecture this stage assembles is exactly the timebase/vector chain of Vol 5: the 18f26K20 PIC (U1) keeps time and computes the vector coordinates; it streams X and Y values into the TLC7528 dual DAC (U3); the DAC’s analog outputs feed the discrete deflection amplifiers (Q1–Q4, Q10, Q11); the two 4132 digital rheostats (U4/U5) trim gain and offset under firmware control; and the 12f629 (U6) is the small “shifter” PIC that nudges the image to prevent phosphor burn-in.

8.5 Part 4 — wiring the CRT socket

With the board complete, Part 4 connects it to the tube. Use 20 AWG wire in 8-inch lengths (12 inches works fine if you want the tube mounted further away), one wire per board pad, soldered (from the bottom, like everything else) to the matching CRT socket pin.

8.5.1 The ten board pads

Board padGoes to CRTSource on board
Kcathodenext to Q10
Ffocuswiper of P1
Aanodewiper of P2
Gcontrol gridnext to the neon bulb
X1X1 plate
X2X2 plate
Y1Y1 plate
Y2Y2 plate
ffilament/heaternear R1
ffilament/heater(second heater pin)

The F and A wires are where the unsoldered P1/P2 wipers from Part 1 finally get their connection. K comes off the Q10 deflection/cathode stage, and G off the negative HV node by the neon.

8.5.2 DG7-32 pinout and measured voltages

The Philips DG7-32 is a 12-pin DUODECAL-based, 2-inch electrostatic tube (6.3 V / 300 mA heater, green medium-persistence phosphor). The assembly text gives this socket map:

DG7-32 pinBoard pad
1f
2G
3K
4F
5nc
6Y2
7Y1
8A
9X2
10X1
11nc
12f

Heater voltage across pins 1 and 12 should read 4–6 VAC. With all other pins referenced to ground (0 VDC; the tab of the 7805 is a convenient ground point):

DG7-32 pinElectrodeVoltage to ground
2G (grid)~ −300 VDC
3K (cathode)~ −295 VDC
4F (focus)~ −150 VDC
6, 7, 8, 9, 10Y2, Y1, A, X2, X1~ +150 to +250 VDC

For the DG7-32, R99 & R4 = 330 kΩ, R6 & R8 = 220 kΩ.

8.5.3 6Lo1i pinout and measured voltages

The Russian 6Lo1i (6ЛО1И) is a 14-pin electrostatic tube the board also supports:

6Lo1i pinBoard pad
1f (filament/heater)
2K (cathode)
3G (grid)
4F (focus)
5NC
6NC
7X2
8X1
9A (anode)
10Y1
11Y2
12NC
13NC
14f

Referenced to ground (the 7805 tab):

6Lo1i pinElectrodeVoltage to ground
2K (cathode)−295 VDC
3G (grid)−300 VDC
4F (focus)−150 VDC
7–11X1, X2, A, Y1, Y2+150 to +250 VDC

Heater across the f–f pins should be about 4–6 VAC. For the 6Lo1i, R99 & R4 = 220 kΩ, R6 & R8 = 180 kΩ. These are approximate but should be close; the per-pin pattern is the same on both tubes and matches the supply philosophy of § 8.2.3 — the cathode and grid floated near −300 V, focus around −150 V, and the plates/anode at a modestly positive +150 to +250 V. The choice of tube and where to source one is the subject of Vol 6; the broader CRT data sheets on hand (the Sylvania 2AP1/3AP1/5ADP/5CP/7JP family) show the same electrostatic architecture at higher anode voltages, which is why the OSC4.4’s deflection recipe has to be re-tuned per tube via R99/R4/R6/R8.

8.6 Part 6 — bringing the CRT to life

This is the moment the build pays off, and the designer’s sequence is a careful crescendo: prove the gun can make a dot, then add the chips one at a time so each new behaviour is attributable to the chip that just went in. Complete all of Part 6 before fitting the Wi-Fi or GPS module — the bring-up jumpers and the module pads can interact.

8.6.1 Pre-set the pots and jumper for a blob ⚠

Set P1–P5 to mid-travel. Set P7 (Y centre) fully counter-clockwise (left) and P6 (X centre) fully clockwise (right) — this parks the spot somewhere it can be found. Then, with short hookup-wire jumpers snug in the empty U1 socket, bridge U1 pin 28 to pin 20, and bridge U1 pin 8 to pin 26. These jumpers tie the PIC’s supply and a control line to fixed levels so the analog front end free-runs and the gun unblanks at a fixed point — there is no firmware yet, so the jumpers stand in for it.

Power up. After a few minutes (the heater has to warm) a blob or dot should appear near screen centre. Adjust P1 (focus) until the spot is as round as possible, P2 until it is as sharp as possible, nudging P3 (brightness) as needed for best focus, then re-touch P1. A round, sharp, controllable dot proves the entire HV chain, the heater, and the basic deflection bias are all healthy. Unplug.

8.6.2 Add the rheostats, then the shifter

Install the two “4132” digital rheostats in U5 and U4, notch up (toward the middle of the board). Power up; move P6 and P7 to their mid positions; the spot reappears, now centre it and re-touch P1/P2 for the roundest, sharpest dot. Unplug.

Install the “12f629” in U6. Power up; when the spot reappears, watch it shift slightly left/right and up/down — that motion is the burn-in-prevention shifter working, confirming U6 is alive. Unplug.

8.6.3 Add the DAC and the brain — the clock face

Install the “7528” (20-pin TLC7528 DAC) in U3, correct orientation, and the largest chip “18f26K20” in U1. Power up, and this time press S2 (slow set) once — the clock face should appear. Set the size with P4 (X) and P5 (Y) to taste, then make a final pass on P1/P2 for the best focus at full size. The clock is now drawing vectors: the PIC computing coordinates, the DAC turning them into deflection voltages, the rheostats trimming gain, and the shifter slowly creeping the image — the whole Vol 5 chain working as one.

8.7 Operating the clock and the time-set interface

Once the face is up, the three buttons run everything. On power-up the screen shows a startup digit — “1” for 60 Hz (USA) mains timing or “0” for 50 Hz — and the blue LED lights with an audible relay click. (If the wrong line frequency is auto-detected, unplug, wait for the LED to extinguish, and re-plug.) The set buttons are the pair sitting close together — fast set and slow set — and S3 (toggle) is the separate display on/off button.

The headline interactions, from the operating instructions:

  • Manual time set: from the startup digit, press slow set once; the clock face appears, then the two front buttons set the time. With no GPS, the clock starts at 12:00 AM, so to reach a PM time you must advance past 12:00 into the afternoon hours.
  • Preset on/off times: press fast + slow together for a handless face; fast set increments the hour hand; slow set confirms (colons appear) to set an ON time, then the OFF time, with a second AM-off/PM-on pair available. Erase a preset by setting every entry to 12:00.
  • GPS/Wi-Fi autoset and time zones: the GPS/Wi-Fi firmware (the 18f2685 variant) adds network/GPS sync, a calendar option, 1PPS timing, and a digital/analog/alternating display mode (slow + toggle, then fast to cycle 0 = alternating, 1 = analog, 2 = digital). The time-zone table runs 1 = EST through 13 = AEST.

These behaviours are documented in the owner’s InstructionSetForScopeClock and illustrated in the designer’s YouTube walkthrough.

8.8 The Wi-Fi and GPS modules — what they add

The OSC4.4’s optional daughterboards add automatic time sync, so the clock never drifts and never needs a manual set after a power cut. They plug into the 5-pin socket/header provisioned in Parts 2 and 3, and — per the strict instruction in Part 6 — are only fitted after the clock face is confirmed working on its own.

8.8.1 The Wi-Fi module assembly

The Wi-Fi module is itself a small through-hole build:

RefPartValue / notes
R1resistor1 kΩ
R2, R3resistors10 kΩ each
R4resistor2.2 kΩ
C1electrolytic10 µF, 6.3–50 V
Q12N3904 NPN
U1regulator”1700” 3.3 VDC regulator
LEDstatus
C2ceramic0.1 µF (104)
S1tactile pushbuttonnormally open (reset)
8-pin socketfor the Wi-Fi module proper
5-pin headerplugs into the main board’s 5-pin socket

The module is a 3.3 V Wi-Fi micro (the U1 1700-series regulator drops the board 5 V to the 3.3 V the radio needs) on an 8-pin carrier; install it aligned with the board outline.

8.8.2 Connecting Wi-Fi to the network

The module is a captive-portal device. From the operating instructions: connect your phone or laptop to the Wi-Fi network named “WifiModule” (disconnect from other networks first if needed), open a browser to 192.168.4.1, and the WifiModuleConnect page lets you enter your SSID and password. Submit, and the module joins your network and the clock shows network-synced time; the credentials are saved. To erase saved credentials, press the module’s reset button (S1), then power-cycle the clock after ten seconds.

8.8.3 The GPS module

The GPS module plugs into the same 5-pin socket (soldered behind Relay2 in Part 2) and adds a miniature GPS antenna. It provides stratum-grade time directly from the satellite constellation — and, with the GPS firmware, the 1PPS (one-pulse-per-second) timing option for the most accurate display. The autoset and time-zone selection sequence is the one in § 8.7: from the startup digit, press toggle to enter zone select, fast set to pick a zone from the 1–13 table, slow set to confirm. Either module turns the OSC4.4 from a clock you set into a clock that sets itself.

8.9 If it doesn’t work

The staged bring-up is itself the first-line diagnostic: a fault almost always belongs to the stage whose test just failed. No neon and no 300 V in Part 1 points at the line jumper, the fuse, a reversed 1N4007, or the relay test-jumper not seated. A 5 V rail but no relay click in Part 2 points at the 7805, the wave-shaper transistors, or the relay. A controllable dot that will not become a clock face in Part 6 points at the DAC or main-PIC orientation, or a deflection-resistor value mismatched to the tube. A dim or oversized image is the R26 / R99 / R4 / P4 / P5 territory of §§ 8.3 and 8.4. For the full fault tree — including the HV discharge procedure that must precede any probing inside the board, focus and centring calibration, and the symptom-to-cause table — see Vol 12. And before re-energizing after any change inside the case: discharge the HV electrolytics by hand; R37 alone is too slow to trust.

8.10 References (Vol 8)

  • OSC4.4 designer, Steps to solder components in order for OSC 4.4 (assembly instructions). Held in 02-inputs/OSC4_4/AssemblyInstructionsForOSC4.4.txt. Primary source for this volume.
  • OSC4.4 designer, Parts list for OSC4.4. Held in 02-inputs/OSC4_4/PartsListForOSC4.4.txt.
  • OSC4.4 designer, Instruction set for scope clock (operating and time-set guide, Wi-Fi connect procedure, time-zone table). Held in 02-inputs/OSC4_4/InstructionSetForScopeClock.docx.
  • Philips, DG7-32 — Low voltage cathode ray tube for oscilloscopes (datasheet, 1957–1959). Held in 02-inputs/OSC4_4/DG7-32.pdf. Source for heater rating, deflection factors, and electrode voltages.
  • Sylvania, Picture Tubes data sheets, types 2AP1 through 12KP4 (electrostatic-deflection oscilloscope and special-purpose tubes). Held in 02-inputs/OSC4_4/CRT Data 2AP1-7JP1.pdf.
  • Owner build notes and the designer’s YouTube bring-up walkthrough (youtube.com/watch?v=y8ro4a65wk8).
  • Cross-references in prose to Vol 3 (HV supply), Vol 4 (deflection and video amplifiers), Vol 5 (timebase and vector firmware), Vol 6 (CRT selection and sourcing), and Vol 12 (safety, calibration, troubleshooting).