Any advantage using DC for filament?

[Mick Bailey]

The main problem with heater hum is NOT coupling of 50 Hz or 60 Hz filament hum into the audio, but the capacitive coupling of rectifier switching artefacts from the HT.

Switching artefacts capacitively couple from the HT winding to the LT winding in the PT, and thence couple from the LT winding to the valve cathode.

Switching noise is 2x line frequency and easily distinguishable on the scope from the main heater waveform. It has a harder sound too - what I would describe as buzz rather than hum due to the abrupt frequency bursts generated from switching. It wouldn't apply to tube rectifiers either. Or SS amps with effective snubber caps.

[trobbins]

HT rectification noise is certainly 2f fundamental, but is a transient, so contains higher frequencies that more easily couple. One noise path is via the HT and heater windings in the PT, so external snubbering is not per se going to change that coupling. A tuned humdinger can be effective in suppressing that form of noise being amplified by the first stage if it is an issue.

[jpfamps]

I have to correct my posts, I have no issue with using rectifier and a cap as it's easy. As I said, I started this thread because of the other thread that has so much problem using a regulator. Then I started to raise the question of the validity of DC supply.

It is all about practicality here. If people have direct experience, I am all ears.

I've used regulated filament supplies with no problems at all; I don't really see why there should be.

I do recommend using a 12VDC supply off a dedicated winding though.

Soft start (which is good practice) is cheap and easy to implement and will prevent any of the non-start up issues.

Mick Bailey has already provided a circuit which is in the regulator application notes.

[kg]

So there is no advantage of DC filament supply?

I am designing an amp, I'm not going to even entertaining the idea of DC filament unless there's a compelling reason for it.

i had/have two reasons:

1. i want to be lazy about my heater lead dress
2. i use the 6vdc rail to power electromechanical relays

schematic is here.

[Alan0354]

I've used regulated filament supplies with no problems at all; I don't really see why there should be.

I do recommend using a 12VDC supply off a dedicated winding though.

Soft start (which is good practice) is cheap and easy to implement and will prevent any of the non-start up issues.

Mick Bailey has already provided a circuit which is in the regulator application notes.

You have the winding on the PT, but most of the ready made PT for tubes don't. You can't use 6.3VAC and put regulator as you need head room. Using a step up switching regulator introduce different set of problem you don't really want to deal with.

I am not familiar with your particular regulator, but I use plenty of those 78xx in circuit boards through out the years. You better read into how to protect the regulator using diodes as they can burn if input collapse before the output. You need heat sink for the regulator. Some of the 78xx heat up even you don't pull a lot of current and apparently it's normal. It is not as straight forward as people think using those simple regulators.

That's the reason I start this thread to ask the validity of the regulating DC and I don't want to interrupt your post. I won't think much using a bridge rectifier DC as it's simple, cheap, trouble free. But I question the validity of a regulated DC. For pcb, at least it's not a lot of work, but for people that hand wire amp, this circuit is quite a few components. To me, you have to have a very compelling reason for a regulated DC.

[jpfamps]

You have the winding on the PT, but most of the ready made PT for tubes don't. You can't use 6.3VAC and put regulator as you need head room. Using a step up switching regulator introduce different set of problem you don't really want to deal with.

I am not familiar with your particular regulator, but I use plenty of those 78xx in circuit boards through out the years. You better read into how to protect the regulator using diodes as they can burn if input collapse before the output. You need heat sink for the regulator. Some of the 78xx heat up even you don't pull a lot of current and apparently it's normal. It is not as straight forward as people think using those simple regulators.

That's the reason I start this thread to ask the validity of the regulating DC and I don't want to interrupt your post. I won't think much using a bridge rectifier DC as it's simple, cheap, trouble free. But I question the validity of a regulated DC. For pcb, at least it's not a lot of work, but for people that hand wire amp, this circuit is quite a few components. To me, you have to have a very compelling reason for a regulated DC.

You can use an auxilliary transformer for the 12V supply if you don't have the correct windings (I've also combined unused 5v and 6.3 V windings to derive a 12V supply).

In my opinion you definitely don't want to derive any DC supply from the LT winding for the power valves, as a valve short can potentially result with the HT on the filament supply, and the components in the DC supply are unlikely to survive.

Thanks for the advice about how to use regulators, but as have been employing them for sometime now with no failures I suspect I know how to use them......

For what it's worth I don't bother with DC filaments in guitar amps.

[g1]

Think again about the context. The other thread was about a mic preamp, probably for studio use. In this application, every trace of noise is worth trying to eliminate. Another place you may see regulated heaters is in tube mic power supplies.
For guitar amps, I think DC heaters will be sufficient enough when required, regulation is probably overkill.

[silverfox]

So there is no advantage of DC filament supply?

I am designing an amp, I'm not going to even entertaining the idea of DC filament unless there's a compelling reason for it.

No, I don't think there is after reading the posts. The few amps I've built have used ground lifting and twisted heater wires. Perhaps you could make room for an alteration after the fact should you have a problem using the no DC method.

Silverfox.

[Chuck H]

JM2C about this dead horse...

Alan wants to know if there is a compelling reason to use DC filaments in a guitar amp. I don't think we've poo poo'd DC so much as to mention that most long standing alternatives work fine without issues if implemented properly. I never have and may not ever make an amp with DC filaments. BUT, there are two scenarios "I" can think of where I wouldn't hesitate. And I think it's these same two reasons that DC filaments seem to be getting more popular with manufacturers and getting more press on forums like this one.

1) Uber gain guitar amps. Some of these amps multiply the sensitivity to such extremes that smart design means every single noise reduction method possible should be employed.

2) PCB construction. With modern PCB construction the filament circuit is usually on the board. It's much easier to layout the board if the filaments are DC because their actual trace path can almost be ignored in lieu of other, more critical considerations.

Since many modern manufactured amps are both very high gain AND of PCB construction with tube sockets on the board I think DC filaments make a lot of good sense. Even in these scenarios I don't think regulation would be necessary.

A typical vintage design Marshall or Fender type amp that is hand wired or uses a PCB without board mounted sockets probably can't be significantly improved by using DC filaments and I would never bother.

[soundguruman]

JM2C about this dead horse...

Alan wants to know if there is a compelling reason to use DC filaments in a guitar amp. I don't think we've poo poo'd DC so much as to mention that most long standing alternatives work fine without issues if implemented properly. I never have and may not ever make an amp with DC filaments. BUT, there are two scenarios "I" can think of where I wouldn't hesitate. And I think it's these same two reasons that DC filaments seem to be getting more popular with manufacturers and getting more press on forums like this one.

1) Uber gain guitar amps. Some of these amps multiply the sensitivity to such extremes that smart design means every single noise reduction method possible should be employed.

2) PCB construction. With modern PCB construction the filament circuit is usually on the board. It's much easier to layout the board if the filaments are DC because their actual trace path can almost be ignored in lieu of other, more critical considerations.

Since many modern manufactured amps are both very high gain AND of PCB construction with tube sockets on the board I think DC filaments make a lot of good sense. Even in these scenarios I don't think regulation would be necessary.

A typical vintage design Marshall or Fender type amp that is hand wired or uses a PCB without board mounted sockets probably can't be significantly improved by using DC filaments and I would never bother.

Voltage reg on filament actually does have an advantage.
You can use less filter capacitors. More ripple rejection...
Typical DC filament, non regulated may require 20,000 uF to get the noise out to an acceptable degree.
Boogie is using 27,000...McIntosh using 20,000...
With regulator, maybe 4700 uF if you set this up ideally..

That's correct. Lower gain amp with good layout will not need DC.
So, there really is a plus side to it.

[Alan0354]

Yep, I am not trashing regulated DC, I just look at the real life situation for mortals like us that don't have a big Fender or Marshall standing behind us. If I do a hand wired in an all tube amp, I have to go out of my way to put a regulated DC as reasons stated many times, then I want to really justify the effort.

But for multi channel switching with opamp and digital controls, by default, you have the LV AC source, and you are on a pcb, putting in an extra regulator is no big deal and I won't think too much but to just put it in.

It is good for people to know by reading the responses here and be the judge themselves.

As for mic amp, are they supposed to be much higher gain than guitar amp, like the cascade gain amp? I read the Fender PA100 schematic, the gain is in the ball park of a Twin which is low gain. So is there a compelling reason for a regulated DC filament? Example like Leo Gnardo gave from his experience is what I am asking about.

I am just asking as people here have a lot more experience in audio tube electronics than me.

[big_teee]

So there is no advantage of DC filament supply?

I am designing an amp, I'm not going to even entertaining the idea of DC filament unless there's a compelling reason for it.

Maybe you said already?
How many preamp tube stages?
How much gain?
I have 2 2204 Amps, with cascade preamps, and I have a good amount of gain.
No excessive hum with AC heaters.
T

[Alan0354]

Voltage reg on filament actually does have an advantage.
You can use less filter capacitors. More ripple rejection...
Typical DC filament, non regulated may require 20,000 uF to get the noise out to an acceptable degree.
Boogie is using 27,000...McIntosh using 20,000...
With regulator, maybe 4700 uF if you set this up ideally..

That's correct. Lower gain amp with good layout will not need DC.
So, there really is a plus side to it.

You have to have at least two caps for the regulator. They are closed loop feedback device, they can oscillate if you don't put filter caps or too small. They designed with a minimum cap value that form the dominant pole for compensation.

A big cap of 10VDC is not going to cost an arm or a leg. Cost is a big disadvantage using regulating DC. YOu definitely need heat sink or design mechanically so you can screw the regulator down onto the chassis. That cost time for mechanical design.

[kg]

Chuck, gtfo with your common sense!

[Chuck H]

That's fine SGM. But that would require a transformer with an appropriate winding (that 99% of guitar amps or readily available transformers do not have) or an auxiliary transformer. Both of which get us even further into the "more trouble than it's worth" category. If you're a hobbyist and tweaking out that sort of thing for little or no advantage is what gets your nut, go ahead. If you're a designer for a guitar amp manufacturer that supplies you a budget and they want you to design and spec parts for their new high gain model, go ahead. Neither of these things apply to me.