Tube voltage ratings

[K O'Connor]

Hi Guys

Tube voltage ratings took a leap in the mid-1950s with a change from Design Center to Design Maximum ratings.

The old system took into account the variations of the tube manufacturer's process AND second-guessed the variations in the end user's product design, leading to very conservative ratings.

The new system only took account of the tube manufacturer's variations and suddenly the tubes sitting on your shelf or in your equipment were rated for much more voltage.

Unfortunately, you still see many data sheets for "new" tubes using old voltage values.

Have fun

[K O'Connor]

Here is an example of an application where the amplifier manufacturer "appears" to be abusing the tubes: Univox amps using 6L6GC.

The 'GC is rated at 500V for both plate and screen; Univox uses a voltage-doubler power supply to create about 700Va and 350Vs. The screen is obviously in no danger at all, but what about the plate? The plate load is about 6k6-aa, which requires 445V peak at 270mA peak for 120W peak output - all well-within the capability of a pair of 6L6GCs. The supply could sag down to 510V and the amp would make its output. That would represent 27% regulation from the supply, which is a bit excessive and does not happen in the real amp, and a bit more output than 60W is actually achieved. Were the supply to remain rigidly at 700V, the average waste heat in each tube would only be 25W, so even a 6L6GB would be okay in the real amp where there is some supply sag.

The fact the screen is operated at such a low voltage helps protect the tube. However, it means that the drive to the grid might have to be low-impedance inasmuch as the grid might have to be pulled positive into conduction to enforce high plate currents. In this specific case, there is no need for that, but the 6L6 applications show this in the data sheet to demonstrate how much extra output can be attained.

Most often we see the screen voltage applied to a tube being higher than what the data sheet suggests. For example, 6550 has a screen voltage rating of 400V and plate at 600V. That is the original version - the 'C' version is 440V and 660V, respectively. Yet, we see 6550s in Marshal amps with 500V on both elements, and in the large bass amps with about 700V on both elements. In both cases, 1k-5W screen stops are used, which protects the screens, as TUTs detailed and recommended from day one. Again, there is significant supply voltage drop at full output and the tubes are safe.

EL-84s are frequently represented to have very low voltage ratings. However, we see them with 450V on both plate and screen, used in pairs and quads. Was Seymour-Duncan inept? Certainly not; they simply know how to read the data sheet and how to protect this particular tube with high-value screen-stops.

[K O'Connor]

Hi Guys

Two important details regarding tube voltage ratings should be mentioned here.

The first is that with all power tubes that exhibit gain (triodes, tetrodes, pentodes) it is assumed that the plate load will be inductive, i.e., a transformer primary to couple the plate signal to the load. Normal transformer-action involves fly-back voltages that equal the amplitude of the plate drive voltage, and at full output this can be nearly twice the DC supply voltage. This high voltage is only present across the tube when it is not conducting, so the tube simply has to be able to sustain this voltage pressure without shorting out.

The voltage at which the plate shorts to the cathode is called the "arc voltage". This spec is not generally listed in the tube data sheet, but it is typically a bit more than twice the rated plate voltage. For example, the EL-34 is rated at 800Va and its arc voltage is 2kV. Mullard provided this information in articles presented through Wireless World magazine back when the tube was first introduced. 2kV is 25% higher than twice the rated 800V, but this cannot be taken as typical although we would hope that it is.

The lowly 6V6 has "secret" capabilities inasmuch as it can withstand 1500V pulses on its plate and therefore must have an arc rating reliably higher than that. This tells you that operating it with 500Va is no big deal and that the 6V6 can be used in nearly any amp Fender made that came equipped with 6L6s, just as we reported in TUT in 1995.

The second important detail regarding tube voltage ratings is that the tube never loses this capability to withstand high voltage. The cathode will lose emission capabilities with use, but the geometry of the tube does not change and thus neither does the voltage rating. The tube will last longer operating at high-voltage and low-current than at low-voltage and high-current.

[makinrose]

Great information! After reading the TUT volumes and seeing that info it really helped me understand some designs that seem like the voltages were too high and allow me to expand the tube types I use. The larger screen resistors are now standard in my shop and my customers experience less tube failures.

One thing I haven't tried is using El-84s with higher voltages. Kevin how much power is possible from a pair of El-84s?

[K O'Connor]

Hi Guys

A pair of EL-84s can produce 30Wrms. I've seen quads in 60W amps.

With fixed-bias, you can run the tubes at a modest temperature and as long as there is reasonable cooling they will last forever. Don't jar them.

Have fun

[makinrose]

Thanks for the reply. At some point I'd like to try to make a higher wattage El84 amp. Nearly all the vintage examples are boiling hot cathode bias so I'd be interested in the tone change as well. I'd also incorporate some of the tricks from the Soma project in TUT5....those strategies have served me well in make El-84 sound less brittle.