Long, long ago, some friends and I in school used to try to repair old valve (tube) TVs. Most commonly the repair was just by cannibalizing another old set; we did not have any budget for spares but we brought some sets back to life.

Many of these sets had no power transformer and just directly rectified the mains. This was the UK and, at the time, the nominal supply voltage was 240V RMS so a peak of about 340V which the circuits were designed to run on. Hence it was not safe to touch the chassis.

I am wondering at what potential the chassis was with respect to ground. Let's make a few simplifying assumptions.

Supply is 240V RMS.

Neutral is 0V with respect to ground.

No transformer.

Full wave rectification with perfect diodes.

Smoothing is just a capacitor. I forget how sophisticated the smoothing might be.

I think that the smoothing is irrelevant. For half the AC cycle, the chassis with be connected to the neutral line and hence at 0V. For the other half cycle, it will be connected to the live. So, I want the RMS voltage of the resulting waveform. A little bit of maths gives me an RMS of 120V.

Does that seem correct?

  • \$\begingroup\$ Some north-American TVs were transformerless too. They had a two-wire AC chord, with no earth pin. You could plug it into the outlet two ways: one way made the chassis "hot" with respect to ground (not nice). The other way was safer. Ask me how I know. \$\endgroup\$
    – glen_geek
    Feb 3, 2020 at 16:48
  • \$\begingroup\$ Reversible 2 pin plugs have never been common here but it would not have made these sets any safer. Did your sets use half wave rectification? If so then the smooth would be even harder. If full wave then I would not expect them to be safe either way around. \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 17:37
  • \$\begingroup\$ Such TVs always used half-wave rectification. The chassis was connected to neutral assuming the power plug was inserted and wired correctly. \$\endgroup\$ Feb 3, 2020 at 18:54
  • \$\begingroup\$ I didn't remember that but it was a very long time ago. I had presumed full wave rectification as the smoothing would be very challenging with half wave. Was a much bigger capacitor cheaper than 3 more diodes? I guess that is possible. I presume that these sets were built to be cheap rather than good. \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 19:43

3 Answers 3


The AC component will be 120 V, but with the positive peaks at Earth (Mains Neutral) potential so the peak voltage will be -338 V. This is worse than 120 V AC for electric shock current because the higher peak voltage punches through skin easier, but perhaps not as bad as 240 V AC because the shock frequency is 50 Hz rather than 100 Hz.

  • \$\begingroup\$ A good point. The RMS is lower due to half of the cycle being missing but the peaks will be the same. \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 19:41
  • \$\begingroup\$ @MichaelHarvey Please explain. \$\endgroup\$
    – badjohn
    Feb 13, 2020 at 11:24
  • \$\begingroup\$ @MichaelHarvey Perhaps you missed the "Full wave rectification" bit. Neither side of the mains is connected to the chassis. And I have worked on such TVs. The first time it was a bit of a shock (literally!) to discover that the mains transformer was actually just an autotransformer, not an isolating transformer. Never make assumptions about mains-powered equipment! \$\endgroup\$ Feb 13, 2020 at 13:03

The first valve/tube TV sets in Britain (TV started in 1936) generally followed the design philosophy of the mains valve radios of the era: a double-wound mains transformer with secondary windings for valve HT, plus heater voltage, often with a separate winding for the HT rectifier valve heater. The chassis would be at earth (ground) potential. However all was not safe because many of these got the EHT for the cathode-ray tube (CRT) from a separate (oil-filled) mains transformer, capable of supplying around 5-11 kV (depending on screen size) with a lethal current capability. The sets were heavy, expensive, and dangerous to service. They are very rare now, and if anyone obtained one, they should be very careful. if it doesn't electrocute you, it could set your house on fire.

After the war, when TV got going again, mass-production demanded simpler and cheaper designs. All the valve heaters in series and a half-wave rectifier for the DC HT, and flyback-derived EHT for the CRT. The chassis was connected to one side of the mains, and as long as that side was the neutral one, to answer your question, these chassis would have, effectively zero voltage difference from earth, but woe betide the service person or meddler if the plug was wrongly wired, or, if two-pin, reversed, or live and neutral reversed in the house wiring.

This carried on until the 1970s or so, when some color/colour sets had power supplies using a bridge rectifier, with the chassis always at half mains voltage, whichever way around the plug was wired.

Service people in the era, and hobbyists restoring them now, tend to use a double-wound isolation transformer between the AC supply and the set. People getting these sets generally think about replacing lots of capacitors (and they know which ones they are) before even applying power, and sometimes they use a variac between the AC and the isolation transformer primary and bring the voltage up gradually, watching for smoke or signs of life.

  • \$\begingroup\$ Thanks. This was the mid to late 70s so maybe some of the sets I encountered had full wave rectifiers, I remember being very cautious about touching the chassis. I think that some had solid state rectifiers even though they still used valves. (It is so long ago then I cannot be sure of any details.) We did have a 1:1 isolating transformer to slightly improve some safety. It is hard today to imagine a teacher leaving kids alone to play with such stuff. \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 21:22
  • \$\begingroup\$ I remember the series valve heaters. The valve code had a prefix indicating series or parallel and the current or voltage. If my memory is right E was 6V parallel heater and P was a series heater but I forget the current. \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 21:25
  • \$\begingroup\$ For valves intended for series heater chains, the U prefix (e.g. UL84) was the most common UK type, 100 mA, and P (e.g. PL84) was 300 mA. \$\endgroup\$ Feb 3, 2020 at 21:40
  • \$\begingroup\$ So, it seems that my P memory was correct but I don't remember U. I remember numbers such as PCL35 which was a combined triode / pentode. I also remember an ECC33 which was a 6V double tirode. (Numbers to be trusted even less than the letters.) \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 21:48
  • 1
    \$\begingroup\$ My first experience of valves was old 4 valve (octal) superhet radios which you could get for nothing, or next to, in the 60s. Standard layout was ECH35 triode-hexode (local oscillator + mixer), EF35 pentode (IF amplifier) EBC33 double-diode-triode (AGC rectifier + AF detector + audio amplifier, the triode grid connection was a top cap) EL33 (AF power output) and typically a GZ34 (full wave HT rectifier, not counted as an 'active' valve) The first three of these typically had a red coating of metallic paint for screening, which could flake off, causing instability if it was the EF35. \$\endgroup\$ Feb 3, 2020 at 22:41

The video circuits will not provide a good picture, if B+ has lots of ripple.

Assume 20 volts drop across rectifier vacuum diode: 5R3GT used in the bigger sets.

I'd assume 340 -20 == 320 volts for B+, which sounds about right.

If 10 triode functions at 10mA each, that's 100mA. Running 10mA and 10K ohm Rplate and Ctotal per node of 10pF gets you 100 nanosecond timeconstant (about right for video) and maximum plate delta_v of 100 volts (a good Z-axis swing to turn off the gun; and healthy pulses with which to synchronize the vertical and horizontal PLLs)

Using I = C * dV/dT and 120Hz, you need 8,333uF filtering for 1volt ripple.

At 10 volts and 0.1amps, you'd need a mere 80 uF total on your most sensitive B+.

  • \$\begingroup\$ Thanks. That detail is interesting and is the answer to a probable follow up question but here I am just asking about the potential of the chassis with respect to ground. In other words, how much it would hurt to touch it. \$\endgroup\$
    – badjohn
    Feb 3, 2020 at 17:39
  • \$\begingroup\$ The high-voltage supply was commonly about 200V. Even though it could have been up to 340V it was intentionally dropped to a lower value so that the set could be operated from a variety of different voltage supplies (including 200V DC). \$\endgroup\$ Feb 3, 2020 at 18:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.