DC filaments (ECC803) on a dedicated winding

[Gtr0]

I had a custom power transformer wound that has 1@ 5V/6A winding, 1@ 6.3V/7.5A winding, and 1@ 6.3V/3A winding.

Originally I had the 6.3V/3A winding added for switching relays, but I have decided to make this particular amp a single channel/high gain amp.

Since I have never done this before, I prefer to ask rather than assume...

Is it possible then to utilize the 6.3V/3A winding to power the preamp tubes with DC? I assume so, but I certainly do not understand everything with tube amps (though I study on a regular basis) - I would set it up as per Valve Wizard's page here --> Heater/Filament Supplies. See the last section at the bottom of the page.

Thanks!

[R.G.]

Possible? Sure. You'll run into several difficulties, all of which are relatively well known. There are solutions for them, though.

The 6.3V/3A winding can be full wave rectified into 6.3V*1.1414 - 1.4V = 8.9V - 1.4V = 7.5V. The actual voltage will be a little more or less, as "6.3V" varies a little in reality. The tricky part is that you don't have enough voltage from 7.5V to run a normal linear regulator to make a stable 6V with a simple linear regulator like a 7806. These things need 2V or so to work right, and the 7.5V estimate is uncomfortably on the low side.

You could simply make the 7.5(ish) and use a resistor in series with the 7.5V to the heaters, with 12A?7 tubes wired for 6V. The resistor eats up the excess voltage. The heaters are a fairly resistive load once they get hot, so this setup stabilizes. Heaters don't really need regulators as long as the supply is fairly well controlled.

There's a lot more to this if you want to dig in further.

[Gtr0]

Thank you, Kind Sir.

I built this amp as a prototype some months back, although I started tearing it apart today, it could have used a bit less noise (5 gain stages) BUT I did a rather messy wire job... no attention to lead dress at all since it was to more or less establish values etc.

So in other words, it isn't necessary to run dc on the preamp unless it is the best way to minimize possible noise issues... most other high gain amps seem to run at least pre tube 1 on dc... so I figured I best look into it before the next prototype.

As far as "digging in further"... do you have recommendations? Otherwise Google has some things, I would love more specific to tube guitar amps though.

Thanks!

[Gingertube]

Just for info:

If the 5V winding is not used or has spare capacity:
Bridge rectify the 5V winding using Schottky Diodes - 4 off something like 1N5821 (30V @ 3 Amps) into a 10,000uF electrolytic.

That will give you +5.95V @ 0.9 Amps (real results I got from powering 3 off 12AX7 heaters). That is close enough to 6.3V.

Cheers,
Ian

[R.G.]

The "has spare capacity" on a 5V winding might be a problem. If the 5V winding is powering a tube rectifier, nearly all tube rectifiers have the cathode connected to the heater. It floats on B+ in most cases. So you would not want to common up a preamp tube heater supply riding on B+, probably. If the 5V is unused, sure, works fine.

I fell into the "wants DC heaters" too. If DC heaters are not necessary, he can just hook them up to the isolated 6.3V, observing the fact that something about the heaters has to be referred to ground or it will hum.

But for DC, using the 6.3Vac winding, a full wave bridge, a BFC and a resistor or just more diodes to drop off the excess voltage ought to work.

[Gingertube]

Good point - I used an uncommited 5V winding, that is, ONLY my little supply was on it.

Cheers,
Ian

[Gtr0]

Thanks for the info!

Yes, I will be running two 5U4GBs on this transformer.

[Chuck H]

How many power tubes? preamp tubes?

[Gtr0]

4x 5881
1x 12ax7 PI
4x 12ax7 pre-amp
2x 5U4GB

[Chuck H]

My thinking was that maybe you could run the power tubes from the 3A tap and use the 7.5A tap for a voltage doubler that could then be regulated. Four big bottles @ .9A is going to be .6A over the winding "rating". I say rating because it may be able to handle the extra current (dependent on wire gauge) with only a small drop in filament voltage. That would allow you to dedicate the 7.5A winding to the entire preamp which only draws 3.75A. But then it occurred to me that you would lose too much capacity in the regulation for even the 7.5A winding!!!

I think the "less regulated" option suggested by R.G. is the best option. Unless you have reason to believe you'll experience great variance in AC mains voltage I don't think "absolute" regulation is critical in this application.

Note that subtracting a couple of preamp tubes from the 7.5A winding will mean it's demand is reduced. That will make for higher filament voltages on that winding. Probably not problematically high, but maybe something to keep in mind.

[Gtr0]

I'm wondering what would have the best overall impact (besides killer lead dress)... elevating the heaters? DC heaters? Humpot... none of the above? All of the above?

[R.G.]

I'm still a little confused by the question, but maybe I can add some useful data.

Is what you're thinking about an attempt to get lower hum? If so:
> You have to have some kind of reference tying the heater filaments to signal ground. If you leave them disconnected, it will hum.
> A grounded center tap on the heater supplies is a decent way to get low-ish hum. This is why many heater windings have center taps, so you can simply ground it and have low hum.
> Sometimes the centertap in the transformer is not really, really centered. That's where the idea of an artificial centertap constructed with two precision resistors came from. Two equal resistors in series makes a more accurate center tap than a winding.
> Close behind that, someone figured out the "humdinger", which is a pot across the heater winding, the wiper grounded. This not only let there be an accurate centertap, you could imbalance it in a way that could semi-cancel hum from other places. It can be better than a resistor center tap in some cases where it can cancel.
> Heater hum can come from electrostatic fields made by voltages on the heater wires. Centertapped can help cancel this, which is one reason the CT stuff works. But you can also sidestep any heater hum by using DC. DC heaters can eliminate that part of hum which comes from E-field emissions from the heaters. This leaves other heater/hum sources unaddressed. And the DC heater supply must be referenced to signal ground, or it will hum too.
> Heater hum can come from cathode coating materials getting stuck on the heaters. When/if this happens, the tube will be permanently noisy with AC heaters.
> Hot cathodes emit some hum all on their own, so DC can fix that part of it. Elevating the heaters, either AC or DC, puts the cathode at a positive voltage, so any electrons the heater emits will be sucked back into it by the positive voltage. That's the theory, anyway.

It's easy to get lost in this mess. The right way IMHO to do it is to
1. set up a resistor centertap and use it.
Still got hum?
2. Go debug all the other hum sources in the amp. Until you do this, no amount of laborious work on heaters will help. Amps with centertapped AC heaters can have quite low hum, so don't go zebra hunting when what you need is a horse.
3. Once the rest of the amp is as hum free as you can possibly make it, worry about DC heaters, elevated heaters, and magic spells.

[Chuck H]

Since the gain level was somewhat referenced analogous with "noise" I thought that it was just gain related noise being discussed. If it's hum or both hum and buzz/hiss then R.G. has the low down WRT AC filaments. But hum can come from other sources besides the filaments. So, if you have hum a good way to tell if it's from the filaments is to disconnect the transformer from the filaments and run them on a lantern battery. If the hum is significantly diminished then it makes sense to chase hum due to the filaments.

As to "noise", being buzz and hiss, there are limits to what can be done in a high gainer. Buzz could be 120 from the power supply. That should be managed with proper grounding and layout. Hiss should be minimized at the design level by considering desired circuit impedances and minimizing signal chain series resistance when it's practical to do so. IMHE lead dress is more about stability than noise, BUT, I suppose it can be noisy if positive feedback loops in high pass circuits are boosting above audio frequencies, but not quite enough to cause noticeable instability. That would create a hiss booster for sure.

[Gtr0]

I'm still a little confused by the question, but maybe I can add some useful data.

Is what you're thinking about an attempt to get lower hum? If so:
> You have to have some kind of reference tying the heater filaments to signal ground. If you leave them disconnected, it will hum.
> A grounded center tap on the heater supplies is a decent way to get low-ish hum. This is why many heater windings have center taps, so you can simply ground it and have low hum.
> Sometimes the centertap in the transformer is not really, really centered. That's where the idea of an artificial centertap constructed with two precision resistors came from. Two equal resistors in series makes a more accurate center tap than a winding.
> Close behind that, someone figured out the "humdinger", which is a pot across the heater winding, the wiper grounded. This not only let there be an accurate centertap, you could imbalance it in a way that could semi-cancel hum from other places. It can be better than a resistor center tap in some cases where it can cancel.
> Heater hum can come from electrostatic fields made by voltages on the heater wires. Centertapped can help cancel this, which is one reason the CT stuff works. But you can also sidestep any heater hum by using DC. DC heaters can eliminate that part of hum which comes from E-field emissions from the heaters. This leaves other heater/hum sources unaddressed. And the DC heater supply must be referenced to signal ground, or it will hum too.
> Heater hum can come from cathode coating materials getting stuck on the heaters. When/if this happens, the tube will be permanently noisy with AC heaters.
> Hot cathodes emit some hum all on their own, so DC can fix that part of it. Elevating the heaters, either AC or DC, puts the cathode at a positive voltage, so any electrons the heater emits will be sucked back into it by the positive voltage. That's the theory, anyway.

It's easy to get lost in this mess. The right way IMHO to do it is to
1. set up a resistor centertap and use it.
Still got hum?
2. Go debug all the other hum sources in the amp. Until you do this, no amount of laborious work on heaters will help. Amps with centertapped AC heaters can have quite low hum, so don't go zebra hunting when what you need is a horse.
3. Once the rest of the amp is as hum free as you can possibly make it, worry about DC heaters, elevated heaters, and magic spells.

Great. Thank you for that. So then i would add... what? 2x 100 ohm resistor to ground after the last filaments? Do I need to worry of any type of precision on those resistors, or even the value? The windings are center tapped, so I assume I would just tape them off? AND finally, if I get any custom would transformers in the future, shall I just eleminate the center taps in the specs? In as far as the filament supply anyway.

Since the gain level was somewhat referenced analogous with "noise" I thought that it was just gain related noise being discussed. If it's hum or both hum and buzz/hiss then R.G. has the low down WRT AC filaments. But hum can come from other sources besides the filaments. So, if you have hum a good way to tell if it's from the filaments is to disconnect the transformer from the filaments and run them on a lantern battery. If the hum is significantly diminished then it makes sense to chase hum due to the filaments.

As to "noise", being buzz and hiss, there are limits to what can be done in a high gainer. Buzz could be 120 from the power supply. That should be managed with proper grounding and layout. Hiss should be minimized at the design level by considering desired circuit impedances and minimizing signal chain series resistance when it's practical to do so. IMHE lead dress is more about stability than noise, BUT, I suppose it can be noisy if positive feedback loops in high pass circuits are boosting above audio frequencies, but not quite enough to cause noticeable instability. That would create a hiss booster for sure.

Well, there was a bit of both, but a low hum was the biggest concern. I havent heard the amp for some months now since I tore it down, but I seem to remember it not being 120Hz. If that means anything.

The first few "prototypes" were sort of layed out similar to a SLO, including the non-twisted heater supply (attached image sn_III below) - with my fourth prototype, I have reverted to twisting, like I always have done before... I was trying something new and this was where I meant I could do a major improvement in the lead dress area (see attached image sn_IV). And admittedly, they could even be twisted better... (solid core 18 guage)

[Chuck H]

I didn't take too much time analyzing everything about the build, but, I've never been a fan of the 'PT at one end OT on the other just across from the inputs' layout. It can certainly work if proximity and phase issues have been worked out in actual bench testing via trial and error as evidenced by the many amps that use this layout, but for a hand built amp with variable lead dress (not a PCB) and generally longer flying leads to the tubes and controls, not to mention whatever variations in specific spacing and proximities exist, I think it can be problematic. Consider that in something like a Soldano there may be a very strong induced NFB proximity affect due to layout that may be keeping the amp stable, or quiet, or something else in the face of what might otherwise be an unstable or noisy amp in the face of other induced feedback loops that are positive. Now change everything just a little and see if it all still works out. This is my problem with such layouts for home built amps. It can become an exercise in chasing proximity induced lead dress issues or worse. Once such a layout has been refined to work and can be replicated by manufacture for consistent layout and lead dress is predominantly replicated on a PCB it can work out just fine. But for a home brew that layout temps fate because it STARTS with an issue that needs to be managed.

JM2C on the popular Soldano layout. Rant over.

But it does lead me to my next observation. Which is the location of your relay board. I assume the relay board is switching things in the preamp signal path.?. So while it might seem like a convenient place to put that board it does move the signal path and it's leads in close proximity to later amplifier stages. I think the relay board and associated leads should be moved to a space over the preamp area of the board with appropriate care taken to fly the relay power supply leads up and away from sensitive signal leads.

Hopefully you have a good grounding scheme. Ground nodes close to where you need them can be very tempting in a hand built amp, but you cannot often daisy chain power amp or power supply and preamp 0V references without problems. Having the power amp so close to the preamp and desiring a tidy appearance invites potential errors in grounding scheme unless you steadfastly stick to a star ground mentality and do not daisy chain grounds.

EDIT: All this said I'll reiterate that you might want to try the lantern battery to power your filaments before assuming that any hum is due to the filaments. You wouldn't want to "fix" something that wasn't a problem just to find that you have to fix something different that is.