300B DC heaters confusion
 

I have read and researched about DC heating my 300B SE amp.

What is the best practice to ground the cathode?
1. Same as the AC setup with each leg to a humpot and then to the 880 ohm resistor?
2. No humpot and one leg grounded through the cap and the other through the 880 ohm resistor?
3. Simply one leg to ground?
http://www.tentlabs.com/Components/T...s/page32_3.gif

Others?

Also, I am running a regulated 5V supply - Broskies H-PS1 using LD1085. Indvidual 5VAC 6A xformers form a voltage doubler so the LD1085 is fed enough to get the 5V.

Hi!

If your DC supply is really ripple free,you don't need no ugly hum pot. Just connect your cathode resistor and bypass cap to one leg of the filament and the filament supply to the other.

Your option 3 would require a separate bias supply

Best regards

Thomas

Rod Coleman is selling a great little regulated supply board and components for very little. I am using them on the GM70 which can get VERY noisey even with hum pots. With the pure DC supply, it should be silent.

Here is a good thread about it. Go to the later pages.

http://www.diyaudio.com/forums/tubes...ht-heater.html

OK, another question...do I now ground the center tap on the filament xformers?

A single connexion for a cathode resistor is best. usually the Filament + terminal is a better choice than Filament -, but it is worth checking both, on a new design of 300B-SE.

When comparing dc heating solutions, there are a few things to watch out for:

- the filament voltage affects the cathode-to-anode voltage along the filament's length. This means that there is a potential gradient of signal (MUSIC!) voltage across the filament. Therefore, any heater solution that requires a capacitor across the filament (ie any dc voltage regulator) will force the signal through this capacitor (usually electrolytic). This will give distortion of its own. And, if the capacitor is less than about 47000 uF, there will be a LF-rolloff in this diverted current, too.

- The heater should buffer the anode current from the dc heater's raw dc supply, because of the rectifier noise, AND because the stray capacitance path causes anode-current leakage to safety-ground - exposing it to poor dielectrics, like PVC transformer leads, etc. and other noise sources.

- The feedback path of the heater-current should not be exposed to the anode-current. If this happens, the anode current transients may enter the feedback loop, and corrupt the dc heater current setting.

- the 3-terminal regulators are noisy, and perform badly in current-source mode. Using these will give a bad impression of what dc heating can do!

Best choices are the Tentlabs module for a ready-made solution, or very large chokes, or if you want a low-cost DIY kit, I have these available still.


http://www.diyaudio.com/forums/tubes...ml#post2425121


All of these can operate without capacitors across the filament, and for LCL solutions, and my kits, and maybe Tentlabs too - the feedback loop is either absent or out of reach of the anode current.

Never mind just answered my own question....

Here is what I did...I took vinylsavior's advice and simply connected H+ and H- to the tube socket. I took one of the legs to the bypassed cathode resistor. I am getting exacly 70V across both channels for a bias of 79.5mA.

The amp is much quieter now but there is still a little hum maybe 120hz...it used to be audible when in the room near the speakers. Now you have to get right up to them.

They are some homemade speakers based upon Zaph SR-71s ~91dB sensitivity.

Thanks for the guidance!

I guess that's the best you can do with 91dB speakers.

djn, Rod, what supply voltage does your board need for push-push GM-70?

Kenneth

Hi!

I disagree. 91dB is not a very efficient speaker at all. With DC filaments it is possible to make a 300B amp absolutely quiet on a 91dB speaker, even with your ear right at the speaker.

Even on a 100dB speaker, a DHT amp can be made absolutely quiet.

I'd check the cause of the remaining buzz. It can be from the filament supply, or from insuffiecient filtering on the B+. Other causes are possible too, like coupling from the PSU, poor layout, especially ground wiring.

Best regards

Thomas

Kenneth, can run the boards from 24.5V (min) up to about 29V with well chosen heatsink.