Crack Circuit questions

[rlyach]

I have ordered a crack and a pair of Bottlehead headphones. In preparation for building, I was going over the simple design and had a couple of questions. Using the HLMP LED's in the cathode circuit will give a forward voltage just above 1.5V according to the Avago datasheet. Doesn't this cause a problem if the input is at 0db attenuation at standard line levels? I understand that 0db attenuation will probably be too loud to listen to. Along the same thoughts, what would happen if you placed a cathode capacitor across the LEDs? I have done this with other LED's to provide instantaneous currents and and stabilize the voltage. Has anybody tried using cathode caps across the HLMP6000s?

[Doc B.]

Doesn't this cause a problem if the input is at 0db attenuation at standard line levels? I understand that 0db attenuation will probably be too loud to listen to.

Audio engineers love to have standards. That's why they have so many of them. If you are talking nominal level for consumer audio, the intended use of the amp, that standard level is -10dBV, which is about .894 Vp-p. The difference between consumer standard level and pro standard level is about 12 dB. This causes a lot of hassles if you are trying to design equipment to be used in both arenas, and it's why we have two different output levels available from our Tube Repro amp.

That cathode cap is a bit counter intuitive as we went to the LED in part to eliminate the sound signature of the bypass cap.  That said, it's easy to try.

[rlyach]

Doc,

Thanks for the explanation. I have been doing a lot of pro-audio stuff lately and had that in my head. I was calibrating line level compressors.

[Paul Birkeland]

The cathode cap would need to be lower impedance than the LED itself, which would be a very, very large cap. (Maybe 1F)

[rlyach]

Good point! According to the data sheet, the HLMP forward resistance is about 15 ohms. At 20Hz the equivalent Xc would require a capacitor of 531uF.
Xc = 1/(2*pi*f*C)

I am not a audio equipment designer, but to get an appreciable difference (1.5 ohm vs 15), assuming a capacitor and diode in parallel, you would need a 6000uf capacitor.

Probably not worth doing. Just for fun I plotted the impedance of the capacitor and diode in parallel over frequency using a 220uf mini electrolytic capacitor. You will only see an improvement at 500hz and above, but you will also introduce phase delays in the current. I probably will be ruled by Bottlehead wisdom and not even try this.