Klausmobile Tube Tester Files
How Long It Takes to Warm Up ? To Cool Down ?
Recently, DrM at Audio Asylum shared his observations of very long, up to 4 minutes, tube warm-up time. His tube, indeed, was a big one, and, being fed with a current source, it certainly needed more warmup time than with a constant voltage supply.
Smaller tubes need shorter warmup times, but how much, exactly? This is important in designs with SS rectifiers using an explicit time delay in plate supply network. Using a tube curve tracer these measuremnents are quite straightforward, although they can be done with a pair of amperemeters (and appropriate power/bias supplies, of course). The graphs were drawn into Excel to save programming time - these are true raw measurements, not a sim.
These tests - to be meaningful - require hot-starting the tube, with plate voltage applied just before swithing on heater supply. Quite a hazard for the tube (that's why you won't see any exotic NOS under test, just cheap disposable ones). Just watching heater current cannot give a full picture. Heater current saturates when filament temperature stabilizes, but at this time the cathode hasn't warmed up yet! In fact, it takes at least just as much extra time to heat up cathode and reach steady-state emission levels. Plate heating up with current also affects tube's thermal steady state point, but this effect is of far less importance, at least in small tubes running at modest plate wattage. Note: in my tester, the tube is mounted horizontally and a fan blowing air through the chassis exhausts through the hollows in tube sockets, so a tube runs cooler than it would in normal application and same wattage.
Test 1. 6N1P, one section fired up to reach 75% rated wattage (200V, 8.6mA). Standard heater voltage (6.30V DC, 0.60A steady-state current) powered by a regulated voltage source (30mOhm DC impedance). Emission begins when heater current is down to 105% of steady-state value (t=8s), reaching 95% of steady state current at t=32s. Plate current transient saturate at t=60s.
Legend: Red: Heater current; Green: Plate current, relative to steady state values. X-axis in seconds.
Test 2. 6N1P, same but heater supply configured as 7V voltage source with 1.5 Ohm inline resistance, yielding 6.1V / 0.55A steady-state filament supply. This is close to how an unregulated voltage source will react to load. Watch the delay in plate current transient, which is apparently longer than the delay in heater current saturation.
Test 3. Does plate current / wattage affect warmup ? Maybe... Here the two superimposed plate current curves refer to 7.7mA (Green) and 1.9mA (Blue) currents. It seems that higher plate current causes faster warmup, but in fact it's only 2-3 seconds delay in 10-30s interval, and whatever happens after that is within the hardware error margin (note the rough shape of blue curve). This is indeed a second-order effect. Since plate warms up only AFTER cathode emission reaches sufficient level, real difference shows only at later warmup stages. I didn't plot the second heater current curve - it's identical.
Test 4. 6N1P - Turn-off test. Plate current decay at sharp heater turn-off. First 1 second - no change in plate current. Then it takes 25 seconds to drop below 10% of regular current.
Test 5. EL84 (Saratov, 05.2000) - 6.3V constant voltage heater supply, grid-1 at -3.0V, plate and grid-2 at 150V yielding 29mA steady state current. This came as a surprise - plate current reaches 32 mA at exactly 1 minute on poweron, than it decays to 29mA. What's up?
English Home - Russian Home - English Tube List - Russian Tube List - Mail (c) klausmobile 2001