Power Valves Seperate Cathode Resistors?

[guitarmike2107]

I am finally getting back to my 2 x KT66 amp project, its cathode biased. The data sheets indicate that the valves should have separate cathode bias resistors, but don’t say why. Anybody know why this should be the case?

Thanks

Michael

[Steve Conner]

The KT88 datasheet says the same thing. It makes the circuit less sensitive to mismatches between valves. Reading between the lines, it meant that there was quite a lot of variation between tubes, and GEC hoped to do away with the hassle of matched pairs.

If you get a matched pair, or don't care about the mismatch (some people even think it improves the tone) you can use a single cathode resistor.

If you're looking for KT66s I have one GEC looking for a home. It is used and tests OK, but I don't have a single-ended amp that can make use of it.

[guitarmike2107]

Makes sense, I will build it with two resistors since I have two (best to do what your told), just wondered if there was something inherent in these valves that required it to be seperate.

[pdf64]

I like seperate cathode resistors - each device can then naturally self regulate to find its own bias point, and they all tend to balance themselves.
1 downside - a major downside when parts were expensive - is that 2 bypass caps are needed, and they must be much larger than for a shared cathode arrangement (in which the signals naturally cancel out, so cap can be small). Pete.

[Gaz]

That's interesting about having to increase capacitance because Kevin O'Connor says in the AC30 chapter of TUT3 that the when splitting the cathodes the resistance needs to be doubled and the capacitance should be halved. It's axiomatic about the cathode resistors, but I just sort of took his word on the capacitance, although he does not qualify it. Your explanation does make sense to me.

[Steve Conner]

With a single cathode bypass cap, the signal cancels in the Class-A region of operation, but not when it moves into Class-AB, because one tube cuts off.

With two independent caps the signal never cancels.

[kg]

ahh, just use fixed bias.

[guitarmike2107]

ahh, just use fixed bias.

There's always one

I would but I am currently too lasy to work out the electronic power control for fixed bias.... so I am sticking with cathode bias until I can be arsed

[mooreamps]

There's always one

I would but I am currently too lasy to work out the electronic power control for fixed bias....

Mine works... Can't be that hard to do.....

-g