01A question

[ui2]

Well, these diodes usually drop about 1V. Multiplying by the current in question is the heat that needs to be considered. Example: (01a) 1V x 0.25A = 0.25W per Diode to be considered.

All the best,
Ulrich

[Rod Coleman]

With filament bias for 01A (250mA), the <250mW will not overheat the TO220 diodes, with no heatsink. TO220s Like these can manage 1W without heatsinking, in environments of 30-40°C inside an amplifier.

[andyjevans]

Thanks, Rod. That's just what I need. Heatsink calculations are rather dense to understand in simple terms. Do you have any views about using these SIC diodes as bias?

[Rod Coleman]

If the diode-bias sounds better than filament bias with a resistor, then we have an interesting data-point, but it is not immediately obvious why it should be. With filament bias, a resistor of about 30Ω would be a starting point.

30Ω in the cathode has almost zero effect on the gm of the 01A (800µA/V to about 780µA/V), so maybe we have something like a excess current-noise in the resistor - which is construction dependent. And as I know you have found sound quality differences among various resistors, this is quite likely.

The other possibility is that the little extra non-linear contribution to the bias, from the diode V vs. I - maybe it adds a little liveliness to the sound. You mentioned that a bit of fatigue was evident from longer sessions of listening (a known downside of such distortion), so maybe some measurement of the difference is worth the effort.

Grid bias should be better still, especially in terms of low distortion, but some don't like the grid capacitor. The Russian teflons sound fine to me in this position, but that's a matter of taste....

[jhstewart9]

We see many subjective observations here. Has anyone taken the time to look with a scope, how did a sine wave look after going thru? Better still, do the obvious, THD & IMD tests. Discover what is causing the changes in sound.

Good new & used test equipment is very affordable now. And a great investment for anyone serious about pursuing the mysteries of electronics.

In ancient times many of us built our own TE. From a kit. Or from the ground up, no kit at all! It worked.

[VT52com]

I asked earlier why fixed (grid) bias was not used so as to eliminate everything in the cathode circuit. You could use a TVC or input transformer and bias through the secondary, not needing a cap then and possible adding a little gain or the TVC functionality in one go.

Personally I would look into finding an optimal resistor for filament bias.

Has anyone tried to check if what you are hearing is not an issue caused by the filament current source, which is pumping current through the cathode resistor? In a regular cathode bias, the current is only a few mA vs the 0.25-1A that is going through there now with filament bias. So there may be two sources at play for this effect in sound that are compounding.

[andyjevans]

We have been auditioning many resistors. My favourite, displacing the otherwise good vitreous enamel ones, is a military wire wound one from Russia in a silver colour.

Filament supply is always an issue, as you say. I always use choke input for that reason - it's smoother.

[jhstewart9]

Seems like John Broskie's (Tube Cad Journal) comment regards audiophiles is not far off the mark. Refer to the attached!

[andyjevans]

That's simplistic and outdated. There's nothing at all wrong with measuring and nothing for 'audiophiles' to be afraid of. On the contrary, there's a lot of great modern measuring equipment about which can help in all sorts of ways. As long as it's used together with listening. But surely we all know this?

[jhstewart9]

Here is the info on a DH amp I built a while ago. Notice lots of performance data.
This amp has none of the warts I see others having problems with in this thread. It is not a rat's nest of wire in a wooden box, something that invites trouble in.
But the 1B5 is microphonic!