Elevated heaters/power supply technical info

[frusciante89]

Hi guys, I was having a discussion with a friend on elevated heaters and we seemed to disagree on the correct definition of it. I felt I had to go back to some power supply schematics to make this clear once for all..

I was looking at some friedmans schematics and I was wondering: is the 0V in the 12V rectified section connected to GND? If not, since there's no center tap on the 12VAC secondary, where is the 0 referenced to? Is it floating?



Can someone point me to a schematic of how elevation is normally done?
Also, on another note, what are the pro/cons of the following strategies in power supply design (full wave/bridge, etc.)? Is there any strong arguments for choosing one over the other?


.

Why is tap 5 not used here? Are the diode couples used for redundancy?
.

why is tap 5 grounded here?

[Enzo]

Some of the heaters are running on 12vDC. The 12vDC supply has o and +!@ terminals/ Neither is ground, so yes, it floats.

Elevated heaters is simple. Connect it to a DC voltage. Typically we use the center tap. COnnect it to some DC other than ground. But it works as well by just connecting one side of the heater string to the DC offset.

As to full wave style etc, they all work. Choice is based upon convenience, cost.

Transformer center tap 5 should be grounded. As drawn, the high voltage supply doesn't work. The drawing is in error. The drawing wwith 5 grounded is correct.

Doubling diodes is done to increase the reverse voltage rating of the circuit.

[Dave H]

I think the 0V of the 12V supply should be grounded.

Elevation is normally done by connecting the centre tap of the 6.3V winding to an elevated voltage. This could be the cathodes of the power tubes in a cathode biassed amp or a potential divider from the B+ in a fixed bias amp.

Two diode full wave is normally used with a tube rectifier. Bridge (still full wave) is used with solid state diodes. The bridge only requires half the turns on the B+ winding.

Tap 5 should be grounded (or connected to ground by the Standby switch).
The diode couples are used because the inverse voltage is close the the max rating of a single diode.

Tap 5 is grounded because it won't work otherwise. If it's open it's as if it's always in “Standby”

[Helmholtz]

I was looking at some friedmans schematics and I was wondering: is the 0V in the 12V rectified section connected to GND? If not, since there's no center tap on the 12VAC secondary, where is the 0 referenced to? Is it floating?

The "red" schematic shows DC preamps heaters, probably floating and not elevated - but depends on rest of circuit.

The "black" schematic shows elevated AC heaters.

Both methods help to minimize heater hum.

[frusciante89]

Is there a reason why someone would not ground the 0V and leave it floating? Will the floating ground generate issues?

[pdf64]

Valves have a limit on the max voltage between heater and cathode. If the heater circuit is floating, no dc reference to the cathode, then we can’t be sure that the voltage limit won’t be exceeded (possibly damaging the heater to cathode insulation).
Hence the need for the dc reference.

[Enzo]

With 0v floating, then the DC supply to those tubes is like a battery.

Unless we specifically have a situation like a cathode follower, elevating DC heaters doesn't do much. The whole point of elevating heaters is to get them electrically distant from the cathode with their (normally) AC influence. With DC heaters, there is no hum influence to abate.

[Helmholtz]

I think that even with DC heaters it is advisable to ground some point of the heater supply. Reason is that while there is no ACV across the heater filaments, the whole heater supply will have some "common mode" AC potential wrt ground caused by capacitive coupling between PT windings, think some 100pF between HV and heater windings (capacitive coupling favors higher mains harmonics, so also AC grounding via a sufficiently large capacitor should work).
Grounding will short-out this source of interference.

[Gregg]

Black schematic number one is one of the correct ways to elevate DC heaters. For higher elevation voltage like 70-85 or more you can take the reference voltage right after the choke.
I usually do this type of elevation in high gain amps (new builds) and especially after the CF failure issue of some tube brands that has become known years ago and thoroughly discussed here.
The other two ways are creating an artificial center tap or connect pin 9 to the elevation point. No need to ground any part of that. Time and practice proof.

[Helmholtz]

Black schematic number one is one of the correct ways to elevate DC heaters.

Black schematic shows (DC) elevated AC heater supply. Of course DC heaters could be elevated as well if required for the reasons described by nickb (cathode followers etc.).

[g1]

Of course DC heaters could be elevated as well if required for the reasons described by nickb (cathode followers etc.).

I had thought so, and vaguely recall seeing an example sometime.
So in post #8, where you said 'ground some point' it would be correct to say 'ground or reference to some other potential' ?

[Helmholtz]

it would be correct to say 'ground or reference to some other potential' ?

Sure, if that "other potential" has very low AC impedance to ground, meaning that it actually IS an AC ground.

[trobbins]

A couple more comments on elevating heaters.

There is usually a very high resistance between heater and cathode inside each tube. That resistance is not 'constant' and has a value that varies with the voltage difference between the heater and cathode. Typically, the larger the voltage separation, the higher the resistance gets, and the lower any related hum becomes - its a bit quirky, and different valves can show markedly different resistance values, and how that value changes with voltage difference. As a valve ages, the heater cathode resistance of some tubes can get quite low, and that allows the AC voltage between heater and cathode to pass through that resistance and that can cause a hum signal to 'appear' over time between grid and cathode, and hence get amplified. Not all circuits are prone to this form of hum (especially when the cathode is grounded, or the cathode bias resistor has a decoupling capacitor). For hi-fi, or hi-gain circuits this mechanism for introducing hum may be noticeable, so designers sometimes make the effort to alleviate it by 'elevating' the heater voltage with respect to the cathode voltage of 'sensitive' tubes (eg. the input stage). If you are chasing a hum in an amp, and it was due to this effect, then a simple work-around is to tube swap and choose a tube that shows lower hum because it inherently has higher heater-cathode resistance.

What can get tricky to work out is what happens when DC is supplied to the heaters. If the DC level is pinned to circuitry ground, then the heaters don't see any hum related AC voltage relative to cathode, and hence that form of hum ingress is 'designed out'. If the DC supply was floating, then as discussed there may be an AC voltage between it and cathode - driven by parasitic capacitance in the power transformer winding - so not a good idea imho. If the DC supply is derived from the transformer by aggressive rectification and filtering, then the rectifier current spikes may somehow transfer over to sensitive circuitry and show up as hum (eg. transfer could be between power transformer secondary windings, or cable layouts) - so it could end up being 1 step forward by using a DC heater, but 2 steps back by rectifier noise ingress appearing as another form of hum - this is not uncommon for DIYers jumping on the 'tweaker' bandwagon to add in this type of modification.

[Gregg]

Black schematic shows (DC) elevated AC heater supply. Of course DC heaters could be elevated as well if required for the reasons described by nickb (cathode followers etc.).

You're right. I just saw the "0V" references and didn't follow further.

If the DC supply was floating, then as discussed there may be an AC voltage between it and cathode - driven by parasitic capacitance in the power transformer winding - so not a good idea imho. If the DC supply is derived from the transformer by aggressive rectification and filtering, then the rectifier current spikes may somehow transfer over to sensitive circuitry and show up as hum...

In my practice I never observed hum in a DC floating supply which doesn't mean that theory is wrong. My DC supplies are pretty simple - Schottky diodes, big caps like 3300-4700uF on both sides, 7812 or LM2940 depending on the secondary. I usually take the reference voltage right after the choke, same values as on the black schematic.

[Helmholtz]

In my practice I never observed hum in a DC floating supply

I usually take the reference voltage right after the choke, same values as on the black schematic.

If you elevate a (DC or AC) heater supply it's no longer floating. Moreover C38 in the black schematic provides a good, low impedance AC ground.