DC Filament Supply

[stokes]

Trying to rectify my 6.3 filament to dc.What is the best way to accomplish this.I have no ct on the tranny for the 6.3 tap.Can I just use 2-100ohm resistors for a ground reference?Anybody have experience with doing this?Thanks in advance for any suggestions.This is going to be used in a stand alone reverb which is kind of a cross between a Fender type and the Ampeg.

[Ray Ivers]

This circuit works great IME, as-is - "improve" it at your own risk: ;)

http://www.schematicheaven.com/marshallamps/jcm800_studio15_15w_4001.pdf

Ray

[Tom Phillips]

BTW, are you planning on using DC on just the pre-amp tubes? I wouldn’t bother for the power tubes and PI.

I'd use a FW bridge rectifier on the 6.3 Vac winding and filter with a big cap such as 3300 to 4700uF / 10 V. Problem is that you will end up with almost 9 Vdc which you will need to reduce before it hits the tube heaters.
To reduce the voltage you could:
1) Add series resistance to drop the excess voltage. PRO = easy & straight foreward, CON: voltage rises if you change the load by pulling a tube. This could damage the remaining tube(s).
2) Add series diodes to drop the excess voltage
3) Add a regulator

If you have an un-used 5V winding you’d be better off to use that. Your filtered DC voltage would be just over 7 V and could be dropped to safe range with the resistor method. Then power the non-critical tubes off the 6.3 Vac winding.

For the balance circuit you can use the two 100 Ohm resistors on the DC side or you could reference the DC heaters to ground by grounding the negative side to the chassis.
Note: If you use the series diode dropping technique then split the diodes equally in each leg (facing the correct way) to keep the circuit balanced.

Hope that helps. Food for thought anyway.
If you have TUT 1 there are a couple of pages on DC filament operation.

Regards,
Tom

[Tom Phillips]

Ray,
I see that you posted a neat circuit reference wail I was writing. I question how the designer figures that the DC voltage is only 2*3.25V = 6.5V. Seems like it would be much higher. Maybe no one checked and it didn't turn out to be a problem.
Tom

[Ray Ivers]

Tom,

I wondered the same thing the first time I saw it, but it turns out that insignificant-at-HV 1.2V diode drop in a FWB - combined with diode efficiency, PT regulation, etc. - all become quite an issue at these low voltages. I think I recall measuring about 6.4VDC in my 4001, with fairly substantial - but balanced - ripple, all of which apparently cancelled in the 12AX7 heater as it was really very quiet. I added the "at your own risk" thing in my original reply as this circuit just begs to be changed - artificial-CT elevation! more filtering! - both of which I tried, and both of which failed, at least for me.

Ray

[Tom Phillips]

Ray,
Ah yes...we are used to ignoring the diode efficiency when working with the HV supply.
Tom

[stokes]

Thanks for that schem,Ray.It worked fine.I got 7vdc with no load it came down to 6.5 w/load of 1-6CG7 and a 12AX7.Got rid of the hum from this rev.It's gonna make de-bugging this thing that much easier.BTW I am using 4x1n4007 diodes with a 1000uf cap.

[carl gigun]

In a reverb unit I question the need for a DC filament supply in the first place. High gain amps might be a different story, but in something like this it's easy to build an amp with no audible hum at all with normal AC heating.

[Ray Ivers]

OK, great! For such a simple supply, it really works pretty well - and as you pointed out, now there's one less thing to worry about hum-wise.

Ray

[stokes]

Carl,this thing was humming like crazy before the dc mod.I am using an extra reverb tank I had from an old Ampeg.So I am trying to make a stand alone reverb that has a circuit similar to the Ampeg-no rev.trans.I have done some mods to the circuit to make it usable in front of the amp.So far the input is similar to the Ampeg input and the out put is closer to the Fender stand alone unit,still trying to tame the out put some,so getting rid of the 60 cycle hum is a great aid-one less noise to worry about.For what it costs to make a dc fil.supply I think it would be beneficial in any effect unit you are going to plug into a tube amp.The slightest amount of hum in the effect is going to be amplified when it hits the amp,so it is a lot like making the first stage of a high gain pre-amp dc.

[carl gigun]

I get your point about the value of a quiet stage at the front of the amp, I just haven't had that problem. Did you try anything else before deciding on DC heaters?

[stokes]

I guess when I said hum like crazy it was a bit of an exageration,the thing has a lot of gain so I am tweaking it to get that down,having so much gain the hum seemed worse,with the dc heaters the hum is gone completely.This is my first tube effect build so it is mainly just an experiment of sorts.I've built a number of amps over the years but never got into effects till now,got no room for more amps so I figured I'd try some effects.For what it costs I think dc heaters are well worth it in an effect.

[yankeerob]

BTW, are you planning on using DC on just the pre-amp tubes? I wouldn’t bother for the power tubes and PI.

In my case I'm building an overdrive head with a 12AX7 first and second stage preamp tube and a 12AT7 that has the first triode being used as a gain stage and the second triode as a split load PI - so I'm kind of leaning towards wanting that to be DC heated... if not for any other reason but keeping that first triode quiet - this is good thinking, no?

I'd use a FW bridge rectifier on the 6.3 Vac winding and filter with a big cap such as 3300 to 4700uF / 10 V. Problem is that you will end up with almost 9 Vdc which you will need to reduce before it hits the tube heaters.
To reduce the voltage you could:
1) Add series resistance to drop the excess voltage. PRO = easy & straight foreward, CON: voltage rises if you change the load by pulling a tube. This could damage the remaining tube(s).
2) Add series diodes to drop the excess voltage
3) Add a regulator

Anybody know of a suitable regulator? I have found SMC packages to do the job but can't find anything else and don't want the hassle of building a PCB to mount it on so the safest available method is gonna be series diodes - is there a particular diode anyone would suggest?

If you have an un-used 5V winding you’d be better off to use that. Your filtered DC voltage would be just over 7 V and could be dropped to safe range with the resistor method. Then power the non-critical tubes off the 6.3 Vac winding.

I'm going with a 5U4 rectifier that's consuming 3 of an available 3.5A - I'm guessing that using the 5V supply for anything else but the rectifier tube is out of the question?

For the balance circuit you can use the two 100 Ohm resistors on the DC side or you could reference the DC heaters to ground by grounding the negative side to the chassis.
Note: If you use the series diode dropping technique then split the diodes equally in each leg (facing the correct way) to keep the circuit balanced.

I'm gonna have a couple KT66's consuming 2.6A of the 5 available on the 6.3V supply before rectification - I need to heat the aforementioned 12AX7 and 12AT7 and possibly another 12AT7 for an all tube FX loop as well - that's .9A for all three - any ideas of what I'm gonna need to balance this thing out?

Best regards in advance,

Rob

[Enzo]

If you have a 5U4, then the 5v winding will be floating at B+, so DO NOT use it for anything else. If you had used a SS rectifier, then the unused 5v winding would be freee for any purpose.

[yankeerob]

Thanks Enzo! I've decided to go with a Weber copper cap on the 375-0-375 (220mA) for the B+ (cheaper than glass in the long run and the right one will allow me to use EL34's, KT66's or 6L6's), 6.3V on the power tubes and a dedicated DC supply from the 5V (3.5A) for a 12AX7 and two 12AT7's... now all I have to do is figure out how to derive that - any suggestions are more than welcome - thanks for posting!!!

Best regards

Rob:D

[Bruce / Mission Amps]

Thanks Enzo! I've decided to go with a Weber copper cap on the 375-0-375 (220mA) for the B+ (cheaper than glass in the long run and the right one will allow me to use EL34's, KT66's or 6L6's), 6.3V on the power tubes and a dedicated DC supply from the 5V (3.5A) for a 12AX7 and two 12AT7's... now all I have to do is figure out how to derive that - any suggestions are more than welcome - thanks for posting!!!

Best regards

Rob:DIf you are going to forgo the 5v rectifier anyhow, why in the world would you blow good money on a copper cap rectifier (with SS diodes built in it) when you can just use $1 worth of 1N4007 diodes and a 100ohm 10w $1 resistor on a terminal strip or the free unused rectifier socket?
And for what it's worth, DC filaments in a guitar preamp is so unnecessary that it astounds me that people are still trying to figure out how to do it using the 5v winding with a FWB and cap filter.
IMHO... a tremendous waste of time and energy.

[MWJB]

"And for what it's worth, DC filaments in a guitar preamp is so unnecessary that it astounds me that people are still trying to figure out how to do it using the 5v winding with a FWB and cap filter.
IMHO... a tremendous waste of time and energy."...Amen to that.

[Steve Conner]

+1! You only need DC filaments if you're trying to make a 5150 gain monster. And even then, I've made gain monsters with AC filaments and the hum ended up below the noise floor with careful tweaking of a hum balance pot.

Also, if the first gain stage of an overdrive amp is sharing a bottle with anything other than the following gain stage, you're asking for oscillation trouble due to capacitive coupling between the plates.

And yes, the Copper Cap is just two diodes and a resistor in a purty looking copper can. Great for filling the rectifier socket of an existing amp with something that looks presentable, but doesn't make much sense in a scratch build.

[yankeerob]

Firstly let me thank you all for posting - I welcome all the comments so far - there are - however - some factors that need clarifying...

If you are going to forgo the 5v rectifier anyhow, why in the world would you blow good money on a copper cap rectifier (with SS diodes built in it) when you can just use $1 worth of 1N4007 diodes and a 100ohm 10w $1 resistor on a terminal strip or the free unused rectifier socket?

Probably because I wasn't aware that that was all a copper cap is - I'm learning as I go... I have a drawer full of 1N4007's and I've ordered some ally clad resistors to experiment with on another project - and there's no existing rectifier socket as it's a complete new build so if what I think you're saying is true - you've saved me some time and expense :D

And for what it's worth, DC filaments in a guitar preamp is so unnecessary that it astounds me that people are still trying to figure out how to do it using the 5v winding with a FWB and cap filter.
IMHO... a tremendous waste of time and energy.

In most cases I'd agree with you - I've built a 5F6A clone out of HRDv iron that's surprisingly quiet as I'm using a NOS Mullard CV2024 in V1 and a coupla used Brimar ECC83's (that after 30-40 yrs are showing no signs of giving up) in V2&3 - the noise floor is astoundingly low and everything's heated via the standard 6.3V winding from the HR PT - so I take your point...

But the amp in question (which I might add is still on the drawing board) is a higher gain Orange overdrive clone - and has zero negative feedback so that's why I'd prefer - by design - to use DC heaters on the pre-amp tubes as I do the occasional session and hate modellers with a passion - there's just something big and ballsy about a 4 x 12 and a good mic - especially if you have the dynamic range of an overdrive head with ZNF... and can if you desire - incorporate bias range switching so you can safely switch the sag resistor in or out.

As an aside - DC heaters are common practice in many Matamps and similar derivatives (including Orange) that have no feedback loop - I also have access to some info on how they do it/done it in the past from the 6.3V winding but it involves a regulator that I - so far - have found difficulty in sourcing - though the man with the aforementioned info reckons he can get them - that remains to be seen... but I digress..

My thinking was rather than go with a 5U4 (and the expense and reliability issues with another piece of glass in my amp which I plan to gig with too) - why not go for what seemed like a compromise between a FWBR and a tube - ie a copper cap - but if you're telling me that all a copper cap is is a FWBR with a 'sag resistor' then that's the route I shall take when designing the power supply. Perhaps now you get a better idea of why I - having the luxury on this one-off occasion of an unused 5V winding - was thinking of using it for the DC filament supply - as it would seem - from all I've read so far - that this would be an easier route than tapping off the 6.3V supply...

Finally - I really appreciate the posts - and understand why you've said what you've said! :D

[mooreamps]

I use a twin dc regulator for the pre-amp and power tube filaments. I do to ensure a consistant 6.3 volts regardless of how much the input power voltage varies. Even on my bench without regulation, I've seen the filament voltage swing as low as 5.5 volts up to over 7.0 volts. I do this to ensure long tube life and consistant tone.

-g

[Enzo]

Wow, with 120v mains, meaning a 20 to 1 voltage ratio on the 6v winding, that represents a 15-20v swing either way on your mains.

[mooreamps]

In my lab, all the power is off a 10 amp breaker, and it has other loads on it beside just my bench. This much of a voltage swing may be more common than what you would want to believe....

-g

[milosch]

I wonder, why not a zener diode after your bridge or half-wave? This assumes you need to do this at all.

[jrfrond]

After much experimentation in the past, I have deemed the DC heater scheme to be a waste of time, components and chassis real estate. A good, balanced and tightly-wound twisted pair is really all that is needed for hum reduction, and regulation is highly overrated. Most heaters are within tolerance at rated line voltage. Also, heaters just do not last as long on DC because they are always at peak voltage. AC heaters reach peak voltage 120x per second (at positive and negative swing) and get to rest at the zero crossing at the rate of 60Hz. It's just another high-tech approach for a low-tech "problem".

[mooreamps]

Also, heaters just do not last as long on DC because they are always at peak voltage. AC heaters reach peak voltage 120x per second (at positive and negative swing) and get to rest at the zero crossing at the rate of 60Hz.

The filaments are still "hot" regardless...

-g