21-09-2026, 07:03 PM
(This post was last modified: 21-09-2026, 07:11 PM by Mike Watterson.)
Indirectly heated valves
All valves have a Cathode and Anode. The cathode at 0 to +20V and the anode to signal (Diode) or HT 45V to 500V. The anode drives the next stage either via a capacitor or transformer (RF, IF or Audio).
The input on a Triode (and usually Tetrode, Pentode) is g1. It must be more negative than the cathode. If the coupling capacitor is leaky to DC, the valve is hard on with excessive current that may destroy multiple things due to control grid (g1) being more positive than cathode. Like a J310 / 2N2819 depletion mode n-jfet.
A tetrode or pentode has a g2 (screen grid) which must be +25 to +200V depending on valve. It's decoupled to 0V/earth/chassis with a capacitor and if that is leaky the voltage on g2 is low and the gain is very low.
A pentode has g3, a suppressor grid, or beam forming plates (A Kinkless Tetrode) which stops secondary emission from the anode to g2. It's usually at 0V/Earth/Chassis or tied to cathode. On many valves (most audio) it's internally connected to cathode.
There are hexodes, heptodes, Octodes. Nonodes etc, with extra grids to mix a local oscillator with RF.
Pre-1960s models may use Silver Mica or Ceramic up to 1000pF. More than 2 uF is usually electrolytic. But the older sets may use paper dielectric (in a waxed card tube, porous plastic or metal with rubber bungs) for 1000pF (1nF) to 1uF. These need replaced with polyester (250V or 400V depending on location). There may be a 470pF to 2.7nF on final Audio anode, across speaker transformer or to 0V/Gnd/Chassis). I usually use a 1kV or 2kV ceramic part as the peak voltage can be HT x2! Often called Tone Correction, but it's a snubber. Damage to final valve can result if it's missing!
Except for highest quality hermetic military capacitors the paper dielectric has an 8 to 12 year life. They may test a bit high in value due to leakage and read even 20M+ on a DVM but they pass uA to mA at 40V + I use a modified flash gun (no tube & 300uF replaced with 1uF) and 2 x 1M Ohm in series. A neon & 400V 0.1uF plastic dielectric cap also in series.
Off = Good
Slow Flash = marginal
Fast flash or steady on = Useless. Damage will result.
I also test the Electrolytics with all valves out, no mains and current limited 2mA 30V PSU to see if they reform.
All valves have a Cathode and Anode. The cathode at 0 to +20V and the anode to signal (Diode) or HT 45V to 500V. The anode drives the next stage either via a capacitor or transformer (RF, IF or Audio).
The input on a Triode (and usually Tetrode, Pentode) is g1. It must be more negative than the cathode. If the coupling capacitor is leaky to DC, the valve is hard on with excessive current that may destroy multiple things due to control grid (g1) being more positive than cathode. Like a J310 / 2N2819 depletion mode n-jfet.
A tetrode or pentode has a g2 (screen grid) which must be +25 to +200V depending on valve. It's decoupled to 0V/earth/chassis with a capacitor and if that is leaky the voltage on g2 is low and the gain is very low.
A pentode has g3, a suppressor grid, or beam forming plates (A Kinkless Tetrode) which stops secondary emission from the anode to g2. It's usually at 0V/Earth/Chassis or tied to cathode. On many valves (most audio) it's internally connected to cathode.
There are hexodes, heptodes, Octodes. Nonodes etc, with extra grids to mix a local oscillator with RF.
Pre-1960s models may use Silver Mica or Ceramic up to 1000pF. More than 2 uF is usually electrolytic. But the older sets may use paper dielectric (in a waxed card tube, porous plastic or metal with rubber bungs) for 1000pF (1nF) to 1uF. These need replaced with polyester (250V or 400V depending on location). There may be a 470pF to 2.7nF on final Audio anode, across speaker transformer or to 0V/Gnd/Chassis). I usually use a 1kV or 2kV ceramic part as the peak voltage can be HT x2! Often called Tone Correction, but it's a snubber. Damage to final valve can result if it's missing!
Except for highest quality hermetic military capacitors the paper dielectric has an 8 to 12 year life. They may test a bit high in value due to leakage and read even 20M+ on a DVM but they pass uA to mA at 40V + I use a modified flash gun (no tube & 300uF replaced with 1uF) and 2 x 1M Ohm in series. A neon & 400V 0.1uF plastic dielectric cap also in series.
Off = Good
Slow Flash = marginal
Fast flash or steady on = Useless. Damage will result.
I also test the Electrolytics with all valves out, no mains and current limited 2mA 30V PSU to see if they reform.
