Filament Emission vs. Heater Voltage

Last Update:

by: jac van de Walle

Introduction

The most important factors for tube lifetime, is the heater voltage, and not forcing full plate current from a tube, while warming up.

Warming up behavior inside a particular amplifier can be crucial for good life time, or can be the reason for early failures. However this is difficult to measure. Even more difficult with the Warming up behavior is, people tend to develop their own fixed ideas about this. To this pool belong things like driving voltage specified heaters with a current source, or also electronically controlled extremely slow heat up. At the same time shutting down the tube abruptly at power off, not realising this if far more destructive than very fast heat up. Funny designs ideas cause often tube defects, but this is not so much the topic here. My best advice is, do things the classic way, and use auto bias.

The heater voltage however, is a factor which is easy to measure, and we are always surprized why people do not bother at all to do this simple, but extremely important measurement. In addition to that, it becomes even harder to understand why users set up their own limits of allowed heater voltage deviation, ignoring manufacturer limits, are react surprized when the manufacturer claims the data sheet limits are real LIMITS and not just nice advice.

This is the situation

  1. The maximum tube life is achieved with the ZERO deviation of the heater voltage. This tube life is an unspecified number, but we have seen cases of +50.000 hours under ideal circumstances.
  2. The minimum (guaranteed) tube life is achieved with the maximum allowed deviation of the heater voltage.
  3. When going beyond the maximum allowed deviation, the tube may develop a defect. Some tubes tolerate it, some don't. The tubes that tolerate it are not "better". It is just the same coincidence why some smokers shorten their life with it, and for and some others that doesn't seem to be the case.

Here is a simple test with three different tubes

We used:

We measured how the tube plate current depends on the filament voltage. There are several reasons why we prefer this change to be as small al possible. Some of the larger Hickok tube testers, even use this method for the 'Lifetime Test' circuit. The Fullmusic tube 2A3 failed this lifetime test on a Hickok 750 tester.

Do not use an unregulated heater supply.

I am an old timer, and I can remember long ago, the mains voltage was so stabile, you could almost calibrate a volt meter with it. Today, with all the solar energy, and with energy often transported over 1000km or more, electricity providers have given up on a stabile mains voltage. You just can not transport energy if there is no voltage difference. With electricity, that is just the way it is. So if many l people in a street block are pumping energy into the grid, the voltage will have to rise.

In some amplifiers, the filament voltage is not electronically regulated, and so it is depending on the mains voltage variations. This means if the mains voltage changes 5%, the filament voltage will also change 5%, and this again leads to a change in plate current. How much that change is, you can see in the table. We marked any deviation in red, where the bias changed more than 10%, and also it must be said that it is very unlikely the filament voltage comes near 2.1 or 2.8 Volt. However, deviations of 5% up and 5% down of the mains voltage during the day is fully normal today, which represents a total change of 10%. For this reason, it is extremely important that the filament transformer does not add it's own deviation to this, due to production tolerance, or uncareful design. In practical situations it may be that a small percentage is added to this by the transformer.

Time of the day. When testing the mains voltage of your house, you need to do so at different moments, and different situations.

Emission Reserve

Another reason why we prefer this variation to be small is, that we at Emission Labs, pay much attention to achieving the highest possible Emission reserve. Such a tube will tend to be more independent of filament voltage variations. For good performance and long lifetime, it comes down to the filament being strong and healthy. If new, such a filament will can provide a lot more current than the tube actually needs. So, changing the filament temperature, will have a small effect only on the Anode current. On the other hand, if the filament has no high emission reserve, in that case lowering it's temperature will result in lower Anode current. This mechanism was already used in the laboratory testers of Hickok, such as 532 and higher - it is the 'Lifetime' switch. So the filament voltage was lowered, and the meter should stay in the green still.

It must be said that at the Emission Labs production, we use another (direct) method to find out the maximum emission, but observing how the tube reacts on filament voltage changes is still an indication for the same effect, and most of all this method can easily be practiced by DIY.

Anode CURRENT MEASUREMENT vs FILAMENT VOLTAGE
Filament
Voltage
Sylvania
NOS 2A3
Sylvania
Deviation
Fullmusic 2A3-Mesh
Fullmusic
Deviation
EML
2A3-Mesh
EML
Deviation
2.1V (*2)
47,9
-5,7
44,7
-12,8
54,2
-1,8
2.2V (*1)
50,8
-2,8
47,4
-10,1
54,7
-1,3
2.3V (*1)
51,8
-1,8
50,8
-6,7
55,5
-0,5
2.4V
52,8
-0,8
54,2
3,3
56
0
2.5V
53,6
0
57,5
0
56
0
2.6V
55,1
1,5
61,8
-5,3
56
0
2.7V (*1)
55,6
2
63,6
6,1
57,7
1,5
2.8V (*2)
56,7
3,1
66,0
8,5
58,1
2,1

All data measured with Amplitrex AT1000.

Green: Normal value for each tube, measured at Anode = 250V, -Grid =45V.
Red Values: This color indicates when the Plate current has changed more than 10%.
Blue Values (*1) : Heater voltage is more than 5% out of tolerance
Blue Values (*2) : Heater voltage is FAR out of tolerance

 

Same data as in the table above, but now graphically displayed

Results of this test:

All tubes passed, but it becomes interesting to see if we deviate the heater voltage more than 5%. That really makes a difference for some tubes. And yes, we find often, people just do this, by mistake or unaware.

The Fullmusic 2A3 works well as long as you stay within 5% heater tolerance. If, for whatever reason, the heater voltage is mote than 5% off, plate current becomes too low (loss of sound) or too high (over heating). It must be said to the advantage of Fullmusic, such failure is an application error, and not a tube error.

The Sylvania NOS 2A3 to work well, apart in 2.1V under heating, This is very much (20% off) and not realistic any more,

The Emissionlabs 2A3 passes all tests, even the (unrealistic) 20% deviation test.

Some remarks:
  1. Do not make a false conclusion, such as all tubes will work perfect for a very long time, as long as heater voltage is within 5%. These results are only INITIAL. Any deviation of heater voltage, any brand, also just 1% has a great effect on the life time. As you can see, the Fullmusic tube responds with a change of 9% plate current, when the heater voltage is changed only 1% up.
  2. The maximum life time is achieved at ZERO percent deviation.
  3. If a tube is left longer in the state of more than 5% deviation of heater voltage, defects may occur. Over heating damage is irreversible with any tube. We can say, under heating damage is reversible with EML tubes, but we can not speak for other brands here.