FAQ
TL;DR: Restoring a Rochar A.1335 (max display 5998) means taming a 200 kHz pulse‑timing DVM; “equipment that is over 60 years old measures correctly.” [Elektroda, sq2aki, post #21759194]
Why it matters: This FAQ helps vintage‑test‑gear fans diagnose, calibrate, and sympathetically upgrade early European digital voltmeters without erasing their history.
Quick Facts
- European early DVM: concept 1963, production 1964; prototype built 1962 by Roger Charbonnier. [Elektroda, sq2aki, post #21759194]
- Measurement engine: pulse‑timing with a 200 kHz quartz generator; no integrator stage. [Elektroda, sq2aki, post #21759194]
- Display tech: NIXIE tubes; original Burroughs B5031, later CSF F9057; LC513 used as drop‑in replacements. [Elektroda, sq2aki, post #21759194]
- AC behavior: simple rectifier front‑end; works reasonably to ~50 kHz; not True RMS. [Elektroda, sq2aki, post #21764145]
- Auto‑range present; faults often trace to reed relays and oxidized rotary contacts. [Elektroda, sq2aki, post #21759194]
What makes the Rochar A.1335 historically interesting?
It is among Europe’s earliest digital voltmeters. A prototype existed in 1962; the catalog model followed in 1964. It uses a pulse‑timing method and early NIXIE display drivers built from germanium transistors. The restored unit was a prototype without a serial number. [Elektroda, sq2aki, post #21759194]
How does the A.1335 actually measure voltage?
It counts 200 kHz clock pulses while a linear ramp climbs until it exceeds the input voltage. A comparator stops the gate, and the counters show the result. There is no integrator or latch; displays flash during counting. "Seemingly simple…a masterpiece" without ICs. [Elektroda, sq2aki, post #21759194]
Why does the display top out at 5998 and why are last digits even?
The chain is /5, /10, /10, /6. That makes the top reading 5998, and the final division by two yields only even least‑significant digits. Charbonnier optimized parts count when high‑voltage transistors were expensive. [Elektroda, sq2aki, post #21759194]
Is the A.1335 a True RMS voltmeter?
No. With a 2 Vpp setting, it reads correctly only for sine waves. Its AC front‑end is a simple converter. Measured performance is reasonable to ~50 kHz but shape‑dependent, so non‑sine waveforms read off. [Elektroda, sq2aki, post #21764145]
What are common power and NIXIE supply pitfalls?
Electrolytics dry out and wiring insulation cracks. Add a resistor and Zener to hold NIXIE supply near 180 V, since the 2N1990 drivers are rated Uce 60 V. Overvoltage can leave inactive digits faintly lit. [Elektroda, sq2aki, post #21759194]
How should I power up a vintage DVM safely?
Never plug straight into mains. First isolate and test the transformer for shorts and insulation. Use an autotransformer to reduce 240 V to ~220 V. High mains can overheat series supplies or saturate the transformer and blow fuses. [Elektroda, sq2aki, post #21759194]
The display shows random digits. Where do I start?
Check the first divider (quinada). A single open diode on the S1 collector blocked reset. In the next decade module, a shorted diode upset division and display. Replacing the diodes restored stable counting. [Elektroda, sq2aki, post #21759194]
How does auto range switching work, and what fails?
Range logic drives reed relays and rotary contacts between switch and automation boards. A broken reed capsule and oxidized rotary wafers prevented ranges from engaging. Replacing the reed and cleaning contacts restored all ranges. [Elektroda, sq2aki, post #21759194]
What is “quinada” in this context?
It is Charbonnier’s coined term for a divide‑by‑five stage built from three flip‑flops and a transistor driver. In the A.1335 the first stage is a quinada feeding downstream decades. [Elektroda, sq2aki, post #21759194]
Can I rejuvenate NIXIE tubes with missing segments?
Yes, if the issue is cathode poisoning. Raise voltage so dark portions just light, and limit current to about three times rated. "Chew" each digit for hours until emission returns, repeating across digits. This saved LC‑531 and IN‑12 tubes. [Elektroda, predom, post #21759928]
What is a chopper amplifier in older meters?
A chopper periodically switches tiny DC signals, converting them for stable amplification while minimizing offset and drift. Specialized contact materials and shielding matter more than current rating. It differs from car vibrators. [Elektroda, OTLamp, post #21761393]
My Marconi mechanical chopper is worn out. Is there a solid‑state retrofit?
Yes. Marconi’s 1984 service addendum shows a J‑FET chopper driven by a CD4011 module that substitutes the mechanical syncroverter. Matched J‑FETs like J112 are recommended for low‑level ranges. [Elektroda, sq2aki, post #21761636]
Did early DVMs auto‑range, and did Polish meters have it?
Yes. Auto‑range was common in first‑generation DVMs. Polish examples include CEMI’s AMS3 and Meratronik models V542, V545, V553, and V563. [Elektroda, OTLamp, post #21761267]
Were Weston cells used as the reference inside early digital voltmeters?
No. Weston cells served calibration and adjustment, not as internal references. Polish Elpo V523/V524/V527/V529 compensating meters and the V552 had cells for that purpose. [Elektroda, OTLamp, post #21761267]
How accurate is AC at high frequency on the A.1335?
Catalog AC specs are optimistic at the top end. At 200 kHz, measured error exceeded 2% versus a 1% catalog claim. The simple converter limits true high‑frequency accuracy. [Elektroda, sq2aki, post #21764145]
How do I perform a safe first power‑on and basic stabilization tweak?
- Test transformer off‑circuit, then recap and rewire per the 1960s schematic.
- Bring up on an autotransformer, monitoring rails and NIXIE supply.
- Add a series resistor and Zener to hold NIXIE supply near 180 V, then calibrate. [Elektroda, sq2aki, post #21759194]
