Schottky diode bias

[pointy]

due to a break of 30 years from valve amps i have not kept up with the modern uses of bits.
i have always liked the idea of Schottky diodes (fast and clean) and i have heard that now days they are used to bias sometimes. so i have 2 questions.

1. how?

2.which ones ( diode numbers)?

[analog_sa]

Any potential advantages Shottkys may have in switching are lost if used for biasing. As the forward voltage drop is less than other diodes you can probably get more accurate bias voltages by stacking them. In any case i have no first hand experience and see no obvious advantages. As for rectification, didn't like them much.

[pointy]

yes it seems that they haven't changed much .....shame
they did look like they could be the next big thing back then .
I just thought I'd ask as I had seen bits on the internet but not enough to see how they would work any better .
so its back to the search for the ultimate IR type or what ever . thanks!

[Wavebourn]

I use Shottky diodes to rectify AC for regulated DC filament powering. For bias LEDs are better.

[pointy]

so am i right in thinking that the types to use would still be the short red dome LEDs and the flat amber LEDs from the 1980s for K bias.
and for CCS the ultra fast recovery diode X series would be best.

[pointy]

is that a yes...................................?

[leadbelly]

Originally posted by pointy
is that a yes...................................?

Maybe nobody's replying because this has been discussed ad nauseam on this forum and it is also spelled out in Morgan Jones' book, so you're obviously not looking very hard if at all. The answer is cheap red LED's are best, both for cathode bias and CCS's, and if they don't get you the bias voltage you need, you can move to different colours or stack small signal diodes.

[ray_moth]

It seems to me that the actual voltage you get from an LED is a bit of a lottery. Maybe that's why I never see them used for bias in commercial amps?

[SY]

Within type, it's actually rather consistent. That's why I'll carry a battery, resistor, and voltmeter with me when rooting though surplus bins. If one LED is (say) 1.9V, all the others in the bin will be within 20mV or so of that value.

I suspect that the reason this is not done more commercially is pure inertia and marketing. The first I saw diode bias was in the late '70s. It never seemed to go anywhere. MJ made a very convincing case for me to try it again, and it's turned out to be such a good solution that I have trouble designing without it.

All the stuff about fast recovery and the like is totally irrelevant. Forward-biased AC impedance is the main parameter of concern. I've not seen any significant frequency dependence from LEDs below my measurement limit (100kHz).

[leadbelly]

Originally posted by ray_moth
Maybe that's why I never see them used for bias in commercial amps?

Does that include guitar amps? I remember reading a review of a boutique amp a while back that gave the impression that LED's were used in the circuit. I rarely see the innards of new commercial amps, either hifi or guitar, so I wouldn't know myself.