Cathode Bias Rk Recommended Values ?

[makinrose]

After reading TUT and seeing the higher Rk values in the TUT5 projects I've begun questioning the lower more traditional values seen in Fender, Vox, Valco, and other vintage amps.  For example, I've measured the dissipation in a Fender 5E3 style deluxe to be about 45ma per 6V6 with 250 ohm resistor.  The plate minus cathode voltage is about 390 VDC so it's running the tubes much higher that 100% dissipation.  The tubes are really cooking or I am missing something?  Why did so many vintage amp have such low values for Rk? 

I've experimented with using a 470 ohm shared resistor which makes the tube set had run at about 33ma per tube. 

I'm just wonder what other people's experiences how they prefer to select the Rk value?  What values are good starting points for common power tubes?

Thanks!

[K O'Connor]

Hi Guys

Wow! How did everyone not see this post?

Vintage amps were based on hifi circuits from the various tube manufacturers, and the guitar use was considered more to be a portable PA than a dedicated guitar amp. Many units had inputs for microphone, accordion, bass, guitar and even phono - that would have to be a ceramic pickup.

In a cathode-biased amp designers then and now seem intent on getting maximum power out of the tube and have to run the tube flat out at idle. Some of those old amps will have low-enough B+ and high-enough OT primary impedances to keep the tubes in class-A, but those are few and far between. Most self-biased amps are just very high bias class-AB.

To stretch the idle power, designers will add the plate and screen power ratings together to arrive at the idle current, and eventually to the Rk value. Considering that most such amps go out of
class-A to reach full output AND share Rk among many tubes, there is no point risking tube failures by having idle currents be so high. There are many ways to improve the hum issues and tube life problems for these amps, as TUTs discuss.

In those early days of electronics, mains voltages were generally lower than they are today. On today's mains, old amps will have higher voltages all around because the primary on the PT might be rated for 115Vac instead of 117Vac or 120Vac - double these for everywhere other than Japan and North America. An easy way to correct the mains expectation of the PT is to buck the voltage down. This will help preserve the amp and tubes.

As I have said here and elsewhere, I am not a fan of self-biased power stages and only use them as a switchable option for a different tone. In that mode I set Rk high to keep tube dissipation modest. I also have a very clean supply so hum is not an issue if there is an option to use a shared Rk for a push-pull tube pair.

One thing you might want to try with self-biased output tubes is to use a much higher value for Ck. 22-100uF is really inadequate to keep Vk constant. You have to get into the mF range for it to work more ideally.

[K O'Connor]

Hi Guys

If you are lucky enough to have a schematic of the vintage amp with voltages marked then you could make an educated guess regarding how high to bump the Rk value.

Suppose the schematic indicates +25V at the tube cathode and 300V for B+ with 115Vac mains.

With the schematic voltages and Rk value, say 200R, we can calculate the total current through the tubes to be 25V / 200R = 125mA. Let's assume that on this schematic this is shared by two tubes, so 62.5mA per tube. The schematic indicates B+ as 300V, then 300V - 25V = 275V across each tube. The heat per tube is 17W.

We would expect that B+ would track the mains increase. The change from 115V to 120V is 4.4%. Is this enough to over-dissipate the tubes? Considering how self-biased amps are usually designed, the answer would be absolutely.

The increased mains should produce a B+ of 313V. If Vk stays in proportion, it only rises by 1V This seems like a small increase, but it is almost twice the increase in voltage.and total Ik is 130mA. There is now 287V across the tubes for a net dissipation of 37.3W, 18.7W each.

The take-away is that a small increase in mains voltage is enough to make a not so small increase in tube heat when considering cathode-biased amplifiers. The voltage increases. Current increases. The increase of heat is the product of those smaller increases.

One other "little detail" is that heater voltage has also risen by 4.4%. Does this put the new heater voltage out of spec for the tube?

Tube heater voltage is rated as +/-10%, meaning that a nominal 6.3V can be as low as 5.67V and as high as 6.93V. much depends on how the power transformer was specified and upon how well the PT regulates its voltages. A massive transformer has very good regulation where a small one does not. Copper losses can help to compensate against rising winding output voltages, but only so far.

Say we have a 6L6 with an ideal heater that is 900mA at 6.3V. The hot resistance is then 7R and the heat is 5.67W, which drops to 4.6W at the lowest allowed voltage, and rises to 6.9W at the highest. Again we see that the change in power is twice the percentage as the change in voltage. The heater resistance should go down a little at the low voltage and rise with the higher voltage, but we cannot know what is happening without monitoring the heater current.

[makinrose]

Thanks for the reply. I learned a lot.

One quick point I wasn't too clear about this in the post, but I was referring mostly to modern clones of the Fender 5E3 Deluxe. Fender kept the same 250 ohm shared resistor for the cathode from 1948 to 1960 on their various Deluxe iterations. However, Va was increased in the later models without changing the resistor value (although other things in the circuit changed). Consequently, many modern work-a-likes with power transformers for modern wall voltages are biased too hot.

What you wrote about vintage amps is very true; used with modern wall voltages everything runs hotter. I'm often dealing with that with amps that I repair.

[K O'Connor]

Hi Guys

The fact the circuit values have not changed in the re-issue amp suggests that the new designers are not paying attention.

Obviously a higher Rk value is needed. If stock is 250R, a 270R may or may not restrain things properly, and you might want to go to 330R directly.

Cathode-biased amps are in dire need of screen protection, so a change to 1k-5W per tube is important.

Have fun
Kevin