DC heater variations

[m1989jmp]

Hi,

I need some explanation in regards to DC supply for vacuum tube heater filaments. As I've read on Valve Wizard site, there is no need for referencing both the AC and DC side of the diode bridge to ground, yet I see every major amplifier company doing exactly that (artificial center taps on both sides of the rectifier).
Is it a bad practice that leads to heater supply failure or am I missing something here?





http://www.valvewizard.co.uk/heater.html

Every symmetric DC supply uses transformer center tap to reference negative and positive output voltages, why would that be a problem for heater supply?

[SoulFetish]

Hi,

I need some explanation in regards to DC supply for vacuum tube heater filaments. As I've read on Valve Wizard site, there is no need for referencing both the AC and DC side of the diode bridge to ground, yet I see every major amplifier company doing exactly that (artificial center taps on both sides of the rectifier).
Is it a bad practice that leads to heater supply failure or am I missing something here?

[ATTACH=CONFIG]56148[/ATTACH]

[ATTACH=CONFIG]56149[/ATTACH]

http://www.valvewizard.co.uk/heater.html

Every symmetric DC supply uses transformer center tap to reference negative and positive output voltages, why would that be a problem for heater supply?

Well, it's not a problem (in most cases). It's been a while since I've been over at the valvewizard's site, and I don't want to speak for Merlin, but he may be referring to the use of two 100Ω resistors to ground which is in common use.
Also, In the top example you posted, the same filament winding supplies AC to the output tubes and DC to the preamp. But in both examples, the heater supply is being used to provide a balanced + & - DC voltage. So, it is completely orthodox to ground the CT in these applications. Where one would have problems is if one were to use a bridge rectifier to generate a single positive DC rail, ground the centertap, and ground the negative side of the reservoir cap or any part of the DC "ground"

[trobbins]

I find it worthwhile to remember that each pair of bridge diodes only conduct for a short duration and then don't conduct for a relatively long period of time. As such, there are typically two ways to draw the circuit - one circuit when the diodes are conducting (effectively just show the conducting diodes as a link), and the other way when none of the diodes are conducting.

With no diodes conducting, there are two independent circuits. In the example schematic, all heater terminal voltages are symmetric wrt ground and each side has a ground reference. If you removed one of the ground links, then that 'side' would float for a large percentage of time. A floating heater circuit may have a ground referenced voltage that fluctuates, depending on parasitic capacitance and voltage.

With a pair of diodes conducting, the two heater circuits are linked. In that particular schematic's arrangement, there is no asymmetry as to where the grounds are on both sides at the time the diodes conduct, so the diode currents are the same. If there were differences between diodes or capacitance values to ground, then there would be some form of glitch as some current would flow between the two grounds.

Unless you take care to provide a symmetric grounding arrangement on either side of the rectifier (with the extra components that are required) then yes incorrect forms of grounding can cause problems, and confusion as to why. And rectifying a heater winding can sometimes cause noise to get in via other mechanisms, making it bewildering as to why an amp is still noisy/hummy.

[g1]

I wonder if those examples aren't using those DC heater supplies for other things as well, resulting in a need for the ground reference?

edit: I see Merlin does speak of shorting out the rectifier, so he is talking about grounding the low side of the bridge output like SoulFetish said.

[Gregg]

I wonder if those examples aren't using those DC heater supplies for other things as well, resulting in a need for the ground reference?

First supply is from Mesa and they usually using it for relays, switching and other things.
Depending on the design grounding at the bridge negative rail only (no CT) will provide hum free operation. The DC heater supply can also be DC lifted from ground if necessary but again that depends.
Below is another supply from Mesa that can be used for other things as well. In previous threads it was noted that those resistors should not be there. I've used this many times and it works very well.