AC filament supply?

[cminor9]

Just came across a schematic that has the heaters for the preamp tubes wired up using DC, and the power tubes using AC.

I can see why you'd want to do this, to keep that AC hum away from your preamp tubes. And for the power tubes, since the AC hum will affect these much less you just use AC.

I can see why you'd not want to do this: because of the extra circuitry needed to provide the correct DC voltage. Why bother when AC works just fine if you use a modicum of care with your build?

Outside of what I've mentioned, is there any compelling reason to use DC for the preamp heaters? And why would they go to the trouble of using AC for the power tube heaters? In other words, if you're gonna convert to DC why not just use it everywhere? Seems like an interesting way to design the amp. Any ideas what the benefit is?

[Enzo]

The heaters will need the same current more or less AC or DC. So your choice:

1. rectify the 6VAC from the transformer for all the tubes - maybe 7-8 amps of DC, so a large 15A or greater bridge is needed.

2. Send the 6VAC to the power tubes and non-sensitive smaller tubes, and rectify just the .6 amp or so for two 12AX7s - requiring a much smaller bridge.

Either way you need one bridge. The wires are the same, the transformer is the same. Only difference is which side of the power tubes is the bridge on. I vote for the smaller cheaper solution.

And as to the complexity of wiring the power tubes for AC, come on, they have to get wired either way. There is a bridge in there or there is not.

The point of DC heaters is not to keep the AC away from the tubes. The point is to keep the AC OUT of the tubes. Having the AC current flowing through the heater is more likely to insert some hum than having smooth DC in the heater. How you arrange the wires will not affect this.

[jrfrond]

In single-ended Class A designs, induced heater hum can be an issue, so DC is SOMETIMES ued for preamps, which are almost always single-ended Class A.

With push-pull power amps, heater hum in the power tubes tends to cancel in the power transformer, although many times, the heater string phase is flipped on one opposing set of power tubes during the design debugging stage to determine which phase has the most humbucking effect.

[loudthud]

DC heaters greatly reduce hum in preamps that are constructed on printed circuit boards. Because implementing a twisted pair or shielding would greatly increase the cost of the PCB, the only way AC heaters could be used would be to fly a twisted pair over the back side of the PCB which defeats some of the cost savings of the PCB. There are exceptions notalby Ampeg but those amps aren't really very high gain. DC heaters for preamp tubes typically use a series connection so that a 12V or 24V supply can be used to further simplfy PCB layout.

[hasserl]

DC heaters greatly reduce hum in preamps that are constructed on printed circuit boards. Because implementing a twisted pair or shielding would greatly increase the cost of the PCB, the only way AC heaters could be used would be to fly a twisted pair over the back side of the PCB which defeats some of the cost savings of the PCB. There are exceptions notalby Ampeg but those amps aren't really very high gain. DC heaters for preamp tubes typically use a series connection so that a 12V or 24V supply can be used to further simplfy PCB layout.

Peavey does this in the Classic series amps and in the older 5150 (don't know about the 5150 II)

[Enzo]

And Marshall does the straight rectification of 6V.

[kg]

i do it because i can then be lazy with heater lead dress, and i can share the ground buss.

[MWJB]

Maybe not relevant to the actual schem you saw, but old Dano/Airline/Wards sometimes used to use the 12AX7 heaters, in series, as a load for the power tube cathodes. Undoubtedly for reasons of cost, either to eliminate the cathode resistor, or to save on the 6.3VAC rating of the PT. Not really recommended for a scratch build.

[black_labb]

Maybe not relevant to the actual schem you saw, but old Dano/Airline/Wards sometimes used to use the 12AX7 heaters, in series, as a load for the power tube cathodes. Undoubtedly for reasons of cost, either to eliminate the cathode resistor, or to save on the 6.3VAC rating of the PT. Not really recommended for a scratch build.

thats an interesting idea. good use of electricity.

[d95err]

If you look around various DIY tube amp forums (e.g. AX84), you'll find lots of posts from people having trouble with their DC heaters. It is more difficult than it seems. It seems that DC heaters often creates more problems than it solves, at least for fairly inexperienced builders.

[Tubis]

I´m doing extremely high gain amps for heavey metal and I just run twisted cables of AC to the sockets on the pcb. I get no hum at all. It´s costly and takes up space on the pcb to do DC. You need big caps to smothe the ripple and a big rectifier.
In my experience the only reason to do DC is to put the filament on the pcb board.

[kg]

I´m doing extremely high gain amps for heavey metal and I just run twisted cables of AC to the sockets on the pcb. I get no hum at all. It´s costly and takes up space on the pcb to do DC. You need big caps to smothe the ripple and a big rectifier.
In my experience the only reason to do DC is to put the filament on the pcb board.

use a capacitance multiplier to save space. ie 2n3055 set up as series pass element with caps hanging off its base. effective capacitance is multiplied by beta.

my preamp is all dc heat and has less than 2mV of ripple fully loaded. there's about 35,000uF @ 16v total capacitance spread between base and post-regulator. one side benefit is a nice gradual turn on. another is having an existing LV rail to power switching circuits. a schemo can be seen here.

[jrfrond]

It seems that DC heaters often creates more problems than it solves, at least for fairly inexperienced builders.

This is ABSOLUTELY true! Amps can be and are dead quiet by using the tried-and-true twisted-pair AC heater string. Here's a hint: use two different colors of wire (I use green and yellow). Cut off a couple of feet of each, equal lengths. Clamp one end of the pair in a vise, and chuck the other end in an electric drill. Run the drill on the slower speed, and twist the pair up nice and tight. The tighter the wind, the better the humbucking effect of the pair. Just cut it off to length as needed. The two colors make it easy to see the phase of the wiring from socket to socket. Looks really neat and professional too.

[Enzo]

Oh, it also depends on the gain of the amp and the tubes. AC heater hum comes from more than just heater wires radiating. That is why they made the spiral filaments after all, to reduce him coming from right inside the tubes. And we also often elevate our AC heaters by some DC amount to reduce the flow of electrons between heater and cathode which also introduces hum.

Those two things are not an issue with DC heaters.

It seems that DC heaters often creates more problems than it solves, at least for fairly inexperienced builders.

And so do reversed phase OT leads, grounding of all sorts, improper bias circuits, bad layout, bad lead dress, backwards lytics, etc.

It is not DC heater operation's fault that some people don't know what they are doing when trying to use it. For the typical pair of 12AX7s on DC, like the front end of a Marshall, you need to rectify the AC with a bridge sufficient to handle the 6/10 amp with plenty of breathing room, filter it with caps of sufficient voltage, control the voltage usually with a dropping resistance (the 12AX7s load should remain constant once lit), and you need to put those parts where they won't bake. I find that a whole lot more likely to work out than for example turning on a new build and not getting parasitics.

The problem I see in those Marshalls more than any other is that the solder never took to the legs of the bridges they use. That is not the fault of DC heating, it is the fault of Marshall for not properly preparing parts for installation on the circuit board. I find the legs are oxidized or filmed somehow. Clean them with abrasive, the solder wets, the solder sticks, the part becomes reliable. I don't know if the parts come dirty, or if the wave solder doesn't have enough chance to heat those heavy legs sufficiently for solder to wet them and that results in the oxidation. Either way that is Marshall's fault, not DC's.

[loudthud]

Another point in the series connection's favor, less current means it's easier to filter. It's a lot easier to build a 24V .15A supply than a 6V .6A supply. But if all you have is a 6.3VAC winding, what do you do? The original Sunn amps copied Dynaco and used a voltage doupler made with two 1000uF caps and two 1amp diodes. The two tubes used the 12V connection. Look on the Sunn site for an M60 schematic.