new Kaiju build and question

[Larpy]

I built my Kaiju at the end of January and have been enjoying it ever since.  After a couple of weeks listening to it in stock form, I replaced the coupling and cathode bypass caps with boutique caps I already had from other abandoned or cannibalized projects.  A little crowded but I like the sound.  Very warm and smooth without sacrificing resolution.

There is a bit of 120 cycle hum, even with the humpots dialed in.  It's low and barely detectable from my listening seat, and I can live with it.  But I wondered if I could lower it by wrapping the 300b heater wires in copper braid, heatshrinking them and connecting one end of the braid to ground.

Is this worth trying?

[Paul Birkeland]

I like that you used blue tack on those big parts, that's helpful.

The DC filament upgrade will lower the noisefloor.  Typically we expect about 5mV or so with an AC heated 300B.  I tested my Kaiju recently at about 120uV, so about a 32dB reduction in noise.

Wrapping the green twisted filament wires in copper braid will do absolutely nothing to lower this noise, and you'll be putting a big conductive piece of metal in the amp that can do weird things if it touches anything.  It would also make the DC filament board a little tougher to install. 

[Deke609]

Further to PB's response: Are you sure it's 120Hz and not 60Hz hum? Since you don't have the DC Filament Supply module, I suspect what you're hearing is 60Hz caused by having 6.3V AC (effectively a 6.3V 60Hz signal) on the filaments which then gets passed to the anode as signal. In other words, I suspect the hum you hear isn't radiating or getting coupled from the heater wiring (which I'd assume would be 60Hz not 120Hz in any event), but is instead intrinsic to the AC heating itself. If this is correct, as PB says shielding the heater wiring won't fix the problem (and this is coming from someone [me] who adds copper and/or mumetal-type shielding to just about everything  )

I'd get the DC Filament Supply module - it's purpose is to fix this problem.

I have the DC Filament Supply and the amp is silent feeding my relatively sensitive planar headphones.

[edit: I deleted my previous post that mistakenly identified the heater circuit as 6.3 VAC 120Hz instead of 60Hz]

cheers, Derek

[Paul Joppa]

The humpot cancels the 60Hz hum, leaving a residual 120Hz - that's the 5mV that PB measured.

[Larpy]

Thanks for the replies.  I'll listen to the amp as is for another couple of weeks and, if the hum still bothers me, I will order the DC filament upgrade.

[Deke609]

The humpot cancels the 60Hz hum, leaving a residual 120Hz - that's the 5mV that PB measured.

I hope I'm not jacking this thread and that my question is of interest to the OP as well: what is the source of the "residual" 120 Hz hum? Is it a harmonic of the 60Hz heater current (in which case the DCF upgrade will eliminate it)? Or B+ ripple? Or something else?

many thanks, Derek

[Larpy]

Jack away, Derek.  My question was answered, and yours is a natural follow-up.

[Paul Joppa]

It's the second harmonic of the 60Hz supply. The humpot cancels the fundamental by making the filament voltage constant (averaged across the filament length).

The harmonics are generated by the tube's slightly nonlinear response.

[Deke609]

Many thanks PJ.

[Larpy]

I have another follow-up question for PB:  I got the idea of shielding the heater wires to the 300b tubes from BH's longstanding practice of using shielded Belden wire for preamp tube heaters.  If shielding the 300b heater wires won't reduce hum in the Kaiju, is that because the hum isn't coming from the heater wires?

Another way of asking this question, I guess, is why BH uses shielded heater wire for preamp tubes but not the 300b tubes.

[Paul Birkeland]

The signal levels are far smaller in preamps so we provide some shielding for heater wiring (sometimes, not always!) in those circuits.  When you have 0.002V coming into something like a phono preamp, then every little bit helps.  On the other hand, the Crack uses unshielded heater wiring and is adequately quiet (though it is lower gain and accepting much higher signal levels).

The signal level on the plate of the 300B can be hundreds of volts, so what could possibly get picked up by the grid is very tiny in comparison, especially considering the noise contribution of the filament inside the 300B is not insignificant. 
 
As a test, you could float the shields on the heater wiring going to the 5670s.  That shielded twisted pair wiring is there because it's an appropriate size for the current demand of a 5670 and it's good practice to use shielded twisted pair wires for AC heaters, but I would bet you'd hear no difference in noise with the shield floating. 

[Rocketman248]

I had a bit of hum in my stock Kaiju.  It was barely noticeable from the listening position, and only in between tracks.  I installed the DC Filament upgrade a few months later and was blown away by the difference.  Absolutely silent from the listening position, and had a noticeable difference in the dynamics during music.  Highly recommended.