There are several ways to heat a tube filament, each having it's own advantages. Here are a few methods. These are numbered. Explanations are sometime long, so read only what concerns to you! 1) Serialized filaments The filament current has some tolerance. So sometimes a 5Volt / 1Ampere filament can draw 1.1 Ampere, or sometimes 0.9 Ampere. Only at 5 Volt, the filament temperature is correct. A problem may come if you have an unstabilized DC heater circuit, since the voltage depends on the current load. So at 1.05 Amps the voltage drops slightly below 5V, and at 0.95 Amps it gets a little above 5V. Even larger problems can come if you serialize a 0.95 Amps tube and a 1.05 Amps tube of the same type! That is because a tube filament is a (very) unlinear resistor. One way or another, this circuit will draw of course 1Ampere. You may expect perhaps one tube has now 4.75V and the other has 5.25V but unfortunately the difference is a lot higher. That is because the one tube is operated at lower temperature, and the resistance of the filament will drop a lot making the effect stronger as expected. Tubes that are made for series connection, are selected to have low tolerance on the filament current, at the same time allowing larger deviations of the voltage that is needed for this. When tubes are series connected, they need to have extra immunity against higher start up current. The resistance of a cold filament is up to 8x lower then when hot. This is really very much, and this can cause large problems if the other tubes in the series circuit have another warm up time. (and they do, since warm up time is hardly specified). An interesting example is the 6SN7 tube, which was made specially for series connection in the GT Version. Whereas later GTB is for fastest warm up. All in all, series tubes suffer more, and will have filament breakage sometimes. Perhaps you remember the days of tube television, a broken filament was 1/3 of all problems. Whereas with parallel connected radios, a broken filament was very rare. Conclusion:Tubes may only be used in series connection, if they are specially made for it! In one sentence: EML Tubes are not intended for series connection. 2) Safe operated AC circuits In one sentence: 5% maximum tolerance, no matter how the 5% adds up. So if you have 5 mains variation already, you should have no additional variation due to amplifier transformers not being fully correct. 3) AC heating The grid voltage of a DHT (directly heated tube) when DC heated produces a DC electric field inside the tube plates. It is from the "left to the right" and indeed the tube is not symmetrically loaded by this. There is a misconception that AC Heating will make this effect disappear. The only way to deal with this is use tubes with best possible linearity and then it doesn't matter anyway. The EML tubes of the latest generation have a filament center tap, to ensure best possible symmetry inside the tube. Connecting this center tap to the outside world is not possible, since it requires an extra pin. However indeed in all (new generation) EML tubes this center tap is physically present inside the tube, enforcing in a mechanical way, that the filament electrical center and physical center are the same. (Even Western Electric 300B does not have this feature!). At EML we have drastically reduced AC hum of our tubes this way, and all output tubes can be used AC heated. For driver tubes or pre-amp tubes we do not recommend it. 4) Ultrasonic, or RF heating Heating a tube Ultrasonic or RF, is a very interesting method, since any noise resulting from this will be eliminated by the output transformer. Besides, it is not audible anyway. The electronics for this is more difficult as one expects, and if anything unexpected will go wrong with the oscillator, the result is a broken filament. So do not experiment with this, unless you are an expert. 5) Safe operated DC linear circuits This is what most people are using. HOWEVER.... there is no such a thing as a stable regulator IC, specially the low drop types are instable and tend to oscillate at the smallest mistake. Avoid low drop regulators if you are not an expert, and use normal ones. You can have an error with those unexpected and without explanation. Some modules are sold simply instable, and the very moment you make the slightest error, the output voltage can swing wildly, or add a few Volts RF signal to it. What you think is a silly simple regulator IC, is in fact a 10...30 Watt power amplifier with very high gain, very fast, and ultra high feedback. Any small error with the wiring and the feedback will become feed-forward due to inductance of the wire. This requires an RF frequency, for a piece of wire to become an inductor, but these IC's are incredible fast, and it happens before you know it. Then the manufacturer puts tricks inside to make it a low drop type. The normal (not low drop) types these cost only 30 cents in production in China, and for that price you can not expect something good. Here are some observed problems:
Best is always: to set up a normal schematic for AC and then connect DC to that. If that will hum only very little with AC heating, it will sure not hum with passive DC heating, even if residual ripple is on it. Hint: Try making a DC stabilized circuit with a suited power Zener Diode. You can't go wrong with this. |
Subscribe or unsubscribe our Mailing List
©2005. All Rights Reserved.
