RC bias on reduction 1.1

[Yahhmuns]

I’ve seen many people switch the LED bias on the crack to RC bias and I was wondering if anyone had experience doing this to the reduction. I’m fairly new to amp design so I was curious to try it out but don’t know how to go about it.

Thanks!

[Paul Birkeland]

I’ve seen many people switch the LED bias on the crack to RC bias

You have?  I think maybe two or three people have done this over the last sixteen years.

and I was wondering if anyone had experience doing this to the reduction.

The thing about the LEDs is that they have very, very low dynamic impedance.  You'd end up needing a pair of really large caps for each 6922 to go along with the bias resistor, and there just isn't' room for that on the tube sockets.

(The LEDs also sound better)

On the Crack you can get away with just resistors and no capacitors because there's loads of extra gain and the output impedance of the first stage is relatively immaterial to the overall performance of the amp (I would add that this is true with the Speedball and a bit less true without it).  In the Reduction, a mod like this would change the RIAA EQ response in undesirable ways. 

[Yahhmuns]

A lot of my friends online swear by the RC swap on the crack but if the reductions RIAA response would deviate greatly maybe I won’t do it. But out of curiosity what ratings for the resistor and cap would be needed for the RC swap or how would I go about calculating this myself. Thanks a lot for the help!

[Paul Birkeland]

I can tell you that I've been messing with this stuff since the late 90s (before being employed by Bottlehead), and I've done the comparisons on my own and they agree with what Doc and PJ have also concluded independently.  I've seen a lot of people on audio forums swear by total nonsense, so I wouldn't put too much stock into that. 

[Yahhmuns]

How does RC biasing work? Do you just put in a resistor that would cause a similar voltage drop to the LED?

[Paul Birkeland]

The cathode bias resistor and plate loading resistor create intersecting lines on the tube curves to fix an operating point.  With a resistor used for plate loading and biasing, there's strong DC feedback and you get very reliable plate voltage coming out.  On the downside, you will lose gain across the cathode bypass resistor unless you use a capacitor across it, and that capacitor has to be chosen correctly so you don't get any response loss from the shelving filter formed.  The quality of this cap can also become important when the signal levels are very small.  If you substitute an LED for the biasing resistor, the circuit will operate as though there's a near-perfect cathode bypass cap with the downside that the plate voltage value will tend to wander a little more than it otherwise should.

It is important to keep in mind that not only do the LEDs we provide sound better than any RC combo we have tried, but they also cost a little more.

[Yahhmuns]

Ah I see, I think i’ll stay away from RC biasing for now, thank you very much for the insight!