The High-Voltage Power Supply

Everything else in a scope clock is an exercise in steering a beam that already exists; this volume is about making the beam exist in the first place, and it is the part of the project most likely to hurt you. An electrostatic cathode-ray tube is a multi-electrode vacuum device, and each electrode wants a different voltage with respect to the cathode: a low-voltage heater to boil electrons off, a slightly negative grid to gate them, a focus electrode at a few hundred volts to squeeze the beam to a point, and one or more accelerating anodes that pull the beam toward the screen at anywhere from a few hundred volts to several kilovolts. The high-voltage supply is the circuit that manufactures that little stack of rails out of either the AC mains or a 12 V brick, and because the anode current is small the whole thing can be built compact and quiet — but “small current” at these potentials still means lethal, and a filter capacitor will hold a killing charge for minutes after you pull the plug. Treat this volume as the engineering reference and Vol 12 as the law: read the safety brief before you energize anything described here.

3.1 The rails an electrostatic CRT needs

Before designing a supply you have to know what it is feeding. An electrostatic gun (the physics of which is Vol 2’s job) presents the supply with five or six distinct loads, and the defining feature of all of them — except the heater — is that they draw almost no current. The beam itself is microamps; the focus and grid electrodes draw essentially nothing (they steer fields, they do not pass current); the accelerating anode A2 sinks the beam current plus a trickle through the internal divider. That is why a 300 V supply rated for a few milliamps runs a whole clock, and why the danger is voltage, not heat.

Here is the rail-by-rail picture for a typical small electrostatic tube (the 2”–5” oscilloscope class these clocks use), with the sign convention referenced to the cathode where the datasheet uses it and to chassis ground where the clock’s supply does:

Rail	Typical magnitude	Sign vs. cathode	Current	What it does
Heater (filament)	6.3 V AC (many tubes); 2.5 V for older 3” types	n/a (floats near K)	150–600 mA	Heats the cathode to thermionic emission
Cathode (K) bias	sets the K–G operating point; often a few hundred V below ground in clock supplies	reference (0)	µA (beam)	Source of the electron stream
Control grid (G1)	a few volts to ~50 V negative of K	negative	~0	Brightness / blanking — more negative = dimmer/off
Focus (often G3 / A1-region)	a few hundred volts; can sit below cathode in some guns	varies	~0	Electrostatic lens — focuses the spot
Accelerating anode A2	a few hundred V to several kV	positive	beam µA + divider	Accelerates the beam toward the screen
Post-deflection accel. (PDA)	+2 kV to +10 kV (PDA tubes only)	positive	µA	Adds brightness/energy after deflection

A few consequences fall straight out of that table. The heater is the only high-current rail, so it gets its own low-voltage winding (§ 3.7). The grid and focus draw no current, so they can be driven from a high-impedance divider off the main HV rail through a potentiometer — that is exactly what the OSC4.4 does with its 1 MΩ focus and anode pots and its brightness pot (§ 3.3). And A2 is where the beam current actually flows, so A2 is the rail whose stiffness and ripple you can see on the screen (§ 3.5).

Which tube wants which A2 voltage is a sourcing question, not a supply-design question, so the full compatibility table lives in Vol 6. The short version: 2” tubes (2AP*, 913-class) want their A2 kept below ~1 kV or you risk the tube; 2BP* tubes tolerate up to ~2.75 kV; and any tube whose type carries a PDA designation needs the multiplier of § 3.4 to reach its rated 3 kV-plus. Pick the tube first, read its A2 ceiling, then size the supply.

3.1.1 Why the polarities look upside-down

Newcomers trip on the sign conventions, so it is worth stating plainly. A CRT accelerates electrons, which are negative, so the screen end (anode) must be positive relative to the gun end (cathode). There are two equally valid ways to arrange that potential difference: put the anode at a big positive voltage and the cathode at ground (the textbook picture), or put the cathode at a big negative voltage and run the anode and deflection plates near ground. Clock supplies almost always choose the second scheme, because it lets the deflection plates and amplifiers (Vol 4) swing around a comfortable mid-rail near ground instead of floating at multi-hundred-volt potentials. That is why the OSC4.4 test points read the cathode at roughly −295 V to −300 V and the grid even more negative, while the deflection plates, focus, and anode sit at modest positive voltages of 150–250 V referenced to ground (§ 3.3). The difference across the gun is still a few hundred volts; the supply has simply slid the whole stack down so the fast analog electronics can live near zero.

3.2 Two ways to make high voltage

There are two architectures in common use for hobby scope clocks, and the choice between them colors the whole build — isolation, efficiency, size, and exactly how it can kill you. The first is line-derived: a mains transformer steps the AC up (or just rectifies it), a diode or bridge rectifies it, and a filter cap smooths it to DC. The owned OSC4.4 is this kind, producing ~300 VDC. The second is low-voltage-derived boost/flyback: a 12 V supply feeds a switching converter that steps up to 800–1200 V, optionally followed by a voltage multiplier for PDA tubes. The open-source TubeTime crt-driver 1 kV board is this kind.

The trade space:

Property	Line-derived (OSC4.4)	LV boost/flyback (TubeTime 1 kV)
Input	AC mains (120/240 VAC)	12 VDC from an external brick
Isolation	Depends on transformer; OSC4.4 uses a mains transformer (isolated secondary), but the supply is referenced to mains and still demands an isolation transformer on the bench	Inherent — the 12 V brick provides mains isolation; the HV is referenced to the 12 V ground, which is isolated from mains
Top voltage	Whatever the transformer + rectifier give (~300 V here); multiplier needed for more	800–1200 V out of the box; multiplier section on-board reaches 3 kV+
Efficiency	Fine — a 50/60 Hz transformer is ~90%+, but it is heavy and runs continuously	High at light loads; the switcher only draws what the beam needs
Size / weight	Dominated by the iron transformer	Small — a CCFL transformer and an inductor the size of a thumbnail
Ripple frequency	100/120 Hz (full-wave) — in-band for the eye, shows as 100 Hz screen waver	Tens of kHz switching — easier to filter, and above visual flicker
Danger	Mains-referenced; lethal and shares a ground with the wall	HV-isolated from mains, but still “several mA at 1 kV stops your heart” (the repo’s own warning)

Neither is “safe.” The line-derived supply adds the hazard of a mains-referenced chassis, which is why an isolation transformer on the bench is non-negotiable (Vol 12). The boost-derived supply removes the mains-reference hazard but the README is blunt that “the circuit can source several milliamps … voltages this high with even just a few milliamps can stop your heart.” The honest summary is that the boost approach is easier to make safe and far easier to package, while the line approach is simpler to understand and is what the owned, fully-documented OSC4.4 actually does.

3.3 The OSC4.4 HV supply in detail

Because the OSC4.4 is owned and its assembly instructions are in hand, it makes the best worked example of a line-derived supply. What follows is reconstructed from the OSC4.4 assembly instructions (Part 1, “High Voltage supply”) and the documented test points; the full build walk-through is Vol 8, but the supply theory belongs here.

3.3.1 Topology and the parts that make it

The OSC4.4 front end is a mains transformer feeding a half-wave voltage-doubler arrangement of two 1N4007 diodes (D1, D3) and two 4.7 µF / 450 V electrolytics (C1, C4), with a 10 Ω / 3 W resistor (R1) in series to limit the inrush surge into those caps at power-on. The two stacked 450 V caps are the giveaway that this is a doubler: each cap charges on alternate half-cycles and they sum to give roughly twice the peak secondary voltage — the ~300 VDC main rail. From that rail a high-impedance network sets the gun electrodes:

• P1 (1 MΩ) — focus pot; its wiper goes to the CRT F (focus) pin.
• P2 (1 MΩ) — anode pot; its wiper goes to the CRT A (anode) pin.
• P3 (50 kΩ / 47 kΩ) — brightness; sets how negative the grid sits.
• R3 (470 kΩ, ½ W) and R37 (10 MΩ) — bleeder / divider elements in the HV string (R37’s 10 MΩ is also a safety bleeder — see § 3.6).
• A neon bulb across the rail as a HV-present indicator: if it glows, the rail is live. (It is not a substitute for a meter, but it is a cheap “do not touch yet” lamp.)
• F1 — a fuse (or fuse holder + buss fuse) in the mains input.
• A 4-pin jumper that selects 120 VAC (U.S.) or 240 VAC operation by reconfiguring the transformer primary; if you never need to switch, you solder a wire into the marked 120 or 240 holes for a permanent connection.
• The transformer — primary pins 1–4 next to the fuse, secondary pins 5–8 toward the board edge.
• B1 — a full-wave bridge rectifier, used in the low-voltage section (Part 2) that derives the +5 V logic and heater supplies, not the HV rail.

   MAINS HOT ─[F1 fuse]─┐
                        │   120/240 jumper selects primary tap
                    ┌───┴───┐
   MAINS NEUTRAL ───┤ XFMR  │ (primary 1-4)
                    └───┬───┘
                 (secondary 5-8)
                        │
              ┌── D1 ──┬── C1 (4.7µF/450V) ──┐
   [R1 10Ω 3W]┤        │                      ├──► ~+300 VDC main HV rail
              └── D3 ──┴── C4 (4.7µF/450V) ──┘     (voltage-doubler output)
                        │
                        ├─[ neon HV indicator ]─ (glows when rail is live)
                        │
                  ┌─────┴───────┬────────────┬──────────┐
              [P1 1M focus] [P2 1M anode] [P3 bright]  [R37 10M
                  │             │              │         bleeder]
               CRT F         CRT A          CRT G        │
                                                        GND

3.3.2 The documented test points

The assembly instructions give two go/no-go measurements that confirm the HV section before you wire the CRT. With the board powered from the wall (through an isolation transformer, per Vol 12) and no tube connected:

Test	Where	Expected	Polarity
Main HV rail	across C6− and R8+	~300 VDC	C6 negative, R8 positive
Grid-side rail	at G (next to the neon) referenced to C6−	~300 VDC	G negative, C6− positive

The neon should light when power is applied; if it does not, the HV is not coming up and you stop and debug before going further. Once both readings are good you disconnect the cord, let the caps bleed down, and proceed to the low-voltage section. The as-wired CRT voltages (cathode at ≈−295 V, grid at ≈−300 V, focus at ≈−150 V, deflection/anode at 150–250 V) are the Vol 8 / Vol 6 numbers, reproduced in the rail table of § 3.1, and they confirm the “slide the whole stack negative” scheme of § 3.1.1.

3.3.3 The battery-backup option

The OSC4.4 offers an optional battery-backup sub-circuit so the clock keeps time through a mains outage. It is a small, entirely-low-voltage addition built around a 78L05 (a 100 mA 5 VDC regulator), a 10 kΩ resistor (R102), a 1N4001 diode (D223), a 3 mm LED, and a 9 V battery connector wired with attention to polarity. The diode isolates the battery from the main 5 V rail so the battery only takes over when mains power drops, and the 78L05 drops the 9 V to the 5 V the timekeeping logic needs. It does not keep the CRT lit — there is nowhere near the energy for the HV rail or heater in a 9 V cell — it only preserves the clock’s count so the time is correct when mains returns. The assembly note suggests leaving the 9 V connector off until the rest of the build is done, to avoid an accidental live lead during assembly.

3.4 Voltage multipliers for PDA tubes

The doubler in the OSC4.4 is the simplest case of a general trick: a ladder of diodes and capacitors — a Cockcroft–Walton multiplier — that turns an AC (or switched) input into a DC output several times its peak. You need one whenever the tube wants more than your transformer or boost stage can deliver directly, which in practice means PDA tubes: any CRT whose type code carries a PDA marking (Vol 6’s compatibility list flags them) uses a post-deflection accelerator electrode that wants +2 kV to +10 kV, far above the 300 V–1.2 kV a single rectifier stage produces. The TubeTime 1 kV board reflects this exactly: it reaches 800–1200 V on its own, and the documentation notes that an optional voltage-multiplier section must be installed to get the 3 kV+ a PDA tube needs.

The multiplier comes in two flavors:

 HALF-WAVE (Villard / Cockcroft-Walton ladder) — output adds one Vpk per stage
                                               
  AC in o──C1──┬──C2──┬──C3──┬─ ... ──► +N·Vpk DC out
              D1│    D3│    D5│
              ─┴─    ─┴─    ─┴─
  AC ret o──┬──D2──┬──D4──┬──D6── ... ──► (return rail)
            C1'    C2'    C3'

 FULL-WAVE — two half-wave ladders fed from a center-tapped or push-pull
 source on opposite phases; ripple is halved and output impedance lower
 for the same number of capacitors.

Each stage adds roughly one peak-input voltage to the output, minus the diode drops and the sag under load. The design rules that matter for a scope clock:

• Diodes must be fast (the switching multiplier runs at tens of kHz) and rated for the per-stage stress — the crt-driver power BOM uses RGP02-20E fast 2 kV recovery diodes for exactly this reason.
• Capacitors must carry the full output voltage at the top stage, not just the per-stage delta, and they must be HV-rated film parts: the same BOM lists 2200 pF / 3 kV, 0.010 µF / 2 kV, and 0.033 µF / 2 kV film caps in its HV string.
• Ripple in a multiplier grows with the number of stages and the load current and shrinks with frequency and capacitance. For an N-stage ladder driving current I at frequency f, the peak-to-peak ripple is roughly proportional to I·N(N+1)/(2·f·C). Two things save the scope-clock builder: the beam load is microamps, and the boost runs at tens of kHz rather than 50/60 Hz, so even a modest ladder is quiet enough not to wobble the screen. That is the whole reason the boost architecture is friendlier to PDA tubes than a 50 Hz line multiplier would be.

Keep the stage count to the minimum that reaches the tube’s rated PDA voltage. Every extra stage adds output impedance (the rail sags more when the beam brightens), adds ripple, and adds another pair of parts that sit at multiple kilovolts.

3.5 The TubeTime 1 kV boost board

The open-source crt-driver ScopePower board is the canonical low-voltage-derived supply and the best worked example of the boost architecture. Reading its bill of materials reveals the topology: it is a flyback/boost converter that takes 12 V in (connector J1, “+12 V”) and produces ~800–1200 V out (connector J4, “HV”), able to source several milliamps. The active parts tell the story:

Designator	Part	Role
Q1	FDB3502 N-channel MOSFET	The switch — chops the 12 V input
T1	CTX210605 CCFL transformer	The step-up element — a tiny high-turns-ratio transformer
L1	150 µH (DR127-151)	Energy-storage / boost inductor
U2	TS271 op-amp	Error amplifier in the feedback loop
U1	ZTL431A adjustable shunt reference	Sets the regulated output voltage
R1	10 kΩ trimmer (ADJ)	Output-voltage adjust
R3	10 MΩ HV resistor	Top of the feedback divider, sensing the output
R6, R11	1 MΩ	Feedback-divider / bleeder elements
C4, C5	1000 µF	Input bulk capacitance on the 12 V side
C3/C7/C8	2200 pF/3 kV, 0.01 µF/2 kV, 0.033 µF/2 kV	HV-rated multiplier / filter caps
D1–D4	RGP02-20E fast recovery	Output rectification / multiplier diodes

The control loop is a classic switching regulator: the TS271 op-amp compares a sample of the output (taken down through the 10 MΩ / 1 MΩ divider so the op-amp sees a few volts, not a thousand) against the ZTL431A’s precise internal reference, and modulates the MOSFET’s drive to hold the output where the 10 kΩ trimmer sets it. Because the loop regulates, the output voltage is stable against changes in beam current and 12 V input — something the OSC4.4’s unregulated doubler cannot claim (its rail sags or rises with mains voltage). The companion 60 V video-bias board in the same repo (12 V in, 60 V out) feeds the cathode video amplifier of Vol 4, and the README notes it can be re-tuned as a Nixie supply by changing R1 and C5 — a reminder that these little boost boards are general-purpose HV generators.

  +12V ─┬─[C4/C5 1000µF]─┬──[L1 150µH]──┬───► T1 primary ──┐
        │                │              │                  │  (CCFL step-up
       GND              GND          Q1 drain               │   transformer)
                                       │                    ▼
                              gate ◄─[ TS271 / ZTL431A      T1 secondary
                                       feedback loop ]          │
                                       ▲                   [D1-D4 + C3/C7/C8
                              R3 10M ──┤◄── HV sense          multiplier/filter]
                              R6/R11 1M┘                         │
                                                                 ▼
                                                        ~800-1200V (J4 "HV")
                                                        (+ optional multiplier
                                                         section → 3kV+ PDA)

The safety note from the README bears repeating verbatim because it is the design’s own warning: “The circuit can source several milliamps so be extremely cautious while using it to prevent nasty electric shocks. Voltages this high with even just a few milliamps can stop your heart!” A regulated, mains-isolated supply is safer to package but no safer to touch.

3.6 Regulation, ripple, filtering, and bleeders

The reason any of this matters to the picture is that the deflection electronics steer the beam, but the HV supply controls how brightly and how steadily it lands. Two supply defects show up directly on the screen.

Intensity ripple — if the A2/anode rail has AC ripple on it, the beam energy varies at the ripple frequency, and the trace gets visibly brighter and dimmer in step. On a line-derived supply that ripple is at 100/120 Hz (full-wave) or 50/60 Hz (half-wave), squarely in the band the eye perceives as flicker, which is the worst place for it to be. On a boost supply the ripple is at the switching frequency (tens of kHz), far above flicker and easy to filter, which is one of the boost architecture’s quiet advantages.

Positional waver — ripple on the rails that reference the deflection plates, or on the focus rail, makes the spot wander or breathe at the ripple frequency: a clock face that shimmers or whose lines won’t sit still. The cure is the same as for intensity ripple: adequate filter capacitance and, on a line supply, enough of it that the 100 Hz component is well below the visible threshold by the time it reaches the plates.

Filtering at HV is governed by the same V_ripple ≈ I_load / (f · C) relationship as any rectifier, with the crucial mitigations that (a) the load current is tiny (microamps of beam plus a milliamp or two of divider), so even small film/electrolytic caps suffice, and (b) raising f via the boost converter shrinks the required C dramatically. The OSC4.4 gets away with two 4.7 µF / 450 V caps precisely because the beam load is so light.

Bleeder resistors do double duty and are not optional. First, safety: a high-value resistor permanently across each HV cap drains the stored charge to a touch-safe level in a bounded time after power-off, so the supply does not lie in wait to shock you minutes later. The OSC4.4’s R37 10 MΩ and the TubeTime board’s R3 10 MΩ + R6/R11 1 MΩ string serve exactly this role. Second, stability and a defined discharge path: a bleeder gives the divider a fixed bottom reference so the rail voltages don’t drift with the (near-infinite) gun impedance, and it guarantees the multiplier caps discharge in order rather than trapping charge in a top stage. Size the bleeder so the RC discharge time is short enough to be safe (tens of seconds, not minutes) but high enough that it wastes negligible power: 10 MΩ across 300 V dissipates 9 mW and bleeds a few-µF cap in well under a minute.

FIGURE SLOT 3.6 — Oscilloscope/photo pair: a clock face shot with visible 100 Hz intensity ripple (under-filtered line supply) next to the same face with a clean rail. Suggested: build-bench photo, to be taken.

3.6.1 Current limiting and why “a few mA” is the whole point

It cannot be said too often, and the open-source design says it for us: a supply that can source only a few milliamps at 800–1200 V is still capable of stopping a heart. The human body’s danger threshold across the chest is on the order of tens of milliamps, but the margin between “tingle” and “fibrillation” is small, and HV defeats skin resistance. Series resistors like the OSC4.4’s R1 (10 Ω) limit inrush, not fault current; the high-value dividers limit the steady current the rails can deliver into the gun, but a direct touch across a charged filter cap bypasses all of it. The design intent is low available current — useful for not destroying the tube and for gentler fault behavior — but it is not a safety feature you can rely on with your fingers. The complete discipline (one-hand rule, shorting stick, isolation transformer, eye protection against implosion) is Vol 12; this volume’s only job is to make sure you respect the rails it just taught you to build.

3.7 The heater supply

The heater (filament) is the odd rail out: it is low voltage but comparatively high current, it is the only winding that does real work in watts, and it can be AC or DC. Most small electrostatic CRTs in this hobby want 6.3 V across the heater (the historical “6.3 V” being the standard receiving-tube heater voltage); a minority of older 3” types (906/3AP*, 908A, 905/907 5” tubes, etc.) want 2.5 V. The crt-driver compatibility list spells the per-tube value out in its FIL VOLT column, and getting it right matters: too low and the cathode under-emits (dim, sluggish trace); too high and you shorten the tube’s life or burn the heater out.

AC vs. DC. AC is simplest — a dedicated low-voltage transformer winding feeds the heater directly, and the OSC4.4 instructions tell you to measure 4–6 VAC across the filament (pins 1 and 12 of a DG7-32; the f–f pins on the board) as the bring-up check. AC is fine for the heater because the heater is a thermal mass that doesn’t care about polarity, and the gun’s geometry keeps heater hum out of the beam in most small tubes. DC heating (rectify and filter the winding) is used when even tiny heater-induced hum modulates the beam visibly — a refinement rather than a necessity for a clock.

The series filament resistor. Few transformer windings produce exactly the tube’s rated heater voltage, so the standard practice is a series resistor — the OSC4.4 calls it Rfilament, sits it next to the display on/off switch (S3), and ships 4.7 Ω / 3 W as the default that “works for many CRTs.” The procedure is empirical and worth quoting from the instructions: connect the tube, power on, measure the voltage across the filament, and if it is too high, raise Rfilament to bring it down. Some European types — the instructions name the DG7-6 — need 10 Ω / 3 W instead. The resistor must be 3 W because it carries the full heater current; a quarter-watt part would cook.

   heater winding  o───[ Rfilament ]───o  CRT pin f
   (e.g. ~6.3 VAC)                          │
                                         (filament)
   heater winding  o─────────────────────o  CRT pin f
                                          │
   Measure HERE across f–f: should read 4–6 VAC for a 6.3V tube.
   If high, increase Rfilament (start 4.7Ω 3W; 10Ω 3W for DG7-6-class).

One wiring hazard from the compatibility notes deserves repeating: some tubes are heater–cathode tied (HC), meaning one leg of the heater is internally bonded to the cathode. On the crt-driver deflection board you must connect that leg to the correct pin of the heater connector (the README warns: pin 5 of J2, not pin 4) or you short the filament transformer through the cathode. Check the tube’s base diagram (Vol 6) before wiring the heater on an HC tube.

3.8 Putting it together — a design checklist

To close the loop between the two architectures and the rails they feed, here is the order of operations a builder actually follows, whichever supply they chose:

1. Pick the tube and read its A2 ceiling and heater voltage (Vol 6). This sets whether you need a multiplier (PDA tube → yes) and whether the heater is 6.3 V or 2.5 V.
2. Choose the architecture. Line-derived if you are building the OSC4.4 or want the simplest-to-understand 300 V rail; boost/flyback if you want isolation, a small package, regulation, or a PDA-capable 1 kV+ rail.
3. Build and verify the main HV rail with no tube connected — the OSC4.4’s “~300 VDC at C6−/R8+” check, or the boost board’s trimmed output into a HV-rated bleeder load.
4. Confirm the gun-electrode rails come out at the right sign and magnitude (the negative cathode/grid, the modest-positive focus/anode of § 3.1).
5. Set the heater with Rfilament and measure 4–6 VAC (or the DC equivalent) across the filament before relying on it.
6. Verify the bleeders actually bleed — power off, wait, and meter the caps down to a safe level every time before you reach in.

The amplifiers that swing the deflection plates around these rails are Vol 4; the firmware that decides where the beam goes is Vol 5; and the safety discipline that keeps you alive while you do all of it is Vol 12, which you should have read before reaching this paragraph.

3.9 References (Vol 3)

• OSC4.4 — Assembly Instructions for OSC4.4, Part 1 “High Voltage supply” and the CRT pinout/voltage tables (DG7-32, 6Lo1i). Held in 02-inputs/OSC4_4 (I have this)/. Source of the D1/D3, R1, C1/C4, P1/P2/P3, R3, R37, neon, F1, jumper, transformer, B1 part values; the ~300 VDC test points; the 78L05/R102/D223/9 V battery-backup option; and the 4.7 Ω/ 10 Ω Rfilament and 4–6 VAC heater procedure.
• E. Schlaepfer (TubeTime), “Electrostatic CRT Driver Design,” tubetime.us/?p=183. Held in 02-inputs/A - Open Source/.
• TubeTime crt-driver open-hardware repository — README.md (1 kV PSU: 12 V → 800–1200 V, several mA; 60 V video-bias board) and ScopePower/ScopePower.csv BOM (FDB3502, CTX210605 CCFL transformer, 150 µH inductor, TS271, ZTL431A, RGP02-20E diodes, 3 kV/2 kV HV caps). Held in 02-inputs/A - Open Source/crt-driver-master/.
• TubeTime crt-driver — CompatibilityList.txt (per-tube FIL VOLT, MAX A2 VOLTAGE, PDA and HC flags) and BaseDiagrams.pdf / ScopeWiring.pdf (tube pinouts and wiring). Held in 02-inputs/A - Open Source/crt-driver-master/.
• ARRL, Winding Coils (inductor/transformer construction background for DIY boost magnetics). Held in 02-inputs/.