Let's build a tube preamp

[beans-on-toast]

A bit late, but better late than never! Please excuse my horrible handwriting and I encourage anyone to give me feedback and/or criticism.

With a positive or negative DC offset, you can shift the ground reference of the heater supply. But AC supplies are noisy, that's why DC heater supplies are better.

 

[rufus.K]

With a positive or negative DC offset, you can shift the ground reference of the heater supply. But AC supplies are noisy, that's why DC heater supplies are better.

Will you type a bit about what, why, and how shifting ground reference is done, for the class, please

 

[cfsporn]

View attachment 802636
What is output voltage then?
Do DV heaters need to be 12v, or can they also be 6v?

Before rectification, output voltage is 4.481V, after rectification it is 6.318V.

The 12A*7s have center tapped heaters. This means they can run at either 6.3V or 12.6V.

 

[rufus.K]

Before rectification, output voltage is 4.481V, after rectification it is 6.318V.

The 12A*7s have center tapped heaters. This means they can run at either 6.3V or 12.6V.

Excellent

 

[Old Garage-Bander]

hmm never seen that

The idea is that with DC heaters, one side is always a bit hotter (has more voltage) than the other due to resistance through the heater. The hotter side eventually "wears out" faster and the heater burns open. With a polarity switch, you can minimize the effect by reversing the DC voltage to the heater. It is sort of a manual method. One time the polarity is one way, the next time, or whenever you remember you flip the switch and reverse the polarity. In theory, the heater filament lasts longer. Obviously, with AC heaters the reversal happens automatically at power line AC frequency rates.

 

[rufus.K]

We're almost ready to talk about triodes. Somebody recently mentioned a first stage scheme using both triodes of a 12ax7, cant wait to hear more and see a sketch.

The idea is that with DC heaters, one side is always a bit hotter (has more voltage) than the other due to resistance through the heater. The hotter side eventually "wears out" faster and the heater burns open. With a polarity switch, you can minimize the effect by reversing the DC voltage to the heater. It is sort of a manual method. One time the polarity is one way, the next time, or whenever you remember you flip the switch and reverse the polarity. In theory, the heater filament lasts longer. Obviously, with AC heaters the reversal happens automatically at power line AC frequency rates.

Great Explaination.

By " one side is always hotter", he mean one triode's heater versus the other one inside a 12__7 tube. A 12__7 tube has , in effect (and very generally speaking) two independent amplification stage inside it.

 

[rufus.K]

The power supply conversation will have to be finished/settled before we can actually make a design and play with numbers. I will try to start a recap, filling in new info, and getting final input. I guess its best to start a fresh redraw of our proposed power supply with DC heaters, DC elevation, and regulation. Im sure the knowledgeable cats will still need to help us a little... and I'll include power switch & fuse.
Watch this for a simple explanation before it becomes a really fun discussion here...

 

[Old Garage-Bander]

We're almost ready to talk about triodes. Somebody recently mentioned a first stage scheme using both triodes of a 12ax7, cant wait to hear more and see a sketch.


Great Explaination.

By " one side is always hotter", he mean one triode's heater versus the other one inside a 12__7 tube. A 12__7 tube has , in effect (and very generally speaking) two independent amplification stage inside it.

Yes, it can be a bit trickier with a dual heater w/center tap. I was thinking in terms of a single heater. Think light bulb.
As I look at the schematic for a 12ax7 for instance, if you were using twelve volts with series heaters it would just be a matter of reversing pins 4 & 5.
With six volt parallel heaters 4 & 5 are common to one side of the heater supply and pin 9 is on the other side. So you'd have to have to reverse 4 & 5 (together) with 9. I am not sure how practical it would be to have a reversing switch, but it is something I remember seeing in old tube gear so it must have had some value at some point in time. Maybe with military gear where tube replacement/availability in the trenches might be in question and this helped extend the tube life a bit.

 

[rufus.K]

Yes, it can be a bit trickier with a dual heater w/center tap. I was thinking in terms of a single heater. Think light bulb.
As I look at the schematic for a 12ax7 for instance, if you were using twelve volts with series heaters it would just be a matter of reversing pins 4 & 5.
With six volt parallel heaters 4 & 5 are common to one side of the heater supply and pin 9 is on the other side. So you'd have to have to reverse 4 & 5 (together) with 9. I am not sure how practical it would be to have a reversing switch, but it is something I remember seeing in old tube gear so it must have had some value at some point in time. Maybe with military gear where tube replacement/availability in the trenches might be in question and this helped extend the tube life a bit.

I like the idea. I'm not sure it's needed in the core of this project, however, when the PSU is settled it would be nice if you drew it in as an option.

 

[Passinwind]

The idea is that with DC heaters, one side is always a bit hotter (has more voltage) than the other due to resistance through the heater. The hotter side eventually "wears out" faster and the heater burns open. With a polarity switch, you can minimize the effect by reversing the DC voltage to the heater. It is sort of a manual method. One time the polarity is one way, the next time, or whenever you remember you flip the switch and reverse the polarity. In theory, the heater filament lasts longer. Obviously, with AC heaters the reversal happens automatically at power line AC frequency rates.

There's some discussion of this here: Valve Amps: Power-supply

Personally, I wouldn't bother, but the more stuff this thread brings to the table the better.

 

[beans-on-toast]

Will you type a bit about what, why, and how shifting ground reference is done, for the class, please

With an unregulated AC heater, what vintage amps use, in some cases you can apply a positive or negative reference voltage to the heater's ground. This shifts the heater ground reference to a potential that is above or below that of the signal ground. It can help reduce the AC hum. An amp can have more than one ground plane.

Another thing that some preamps have to lower the noise floor is to offset the signal ground. One way to do this is to use a chassis-to-ground filter, this is an old technique. Line power ground is often noisy, you can have hundreds of millivolts. With an AC volt meter, read your neutral (wide spade) to safety ground (third prong) voltage. It should read zero volts. Anything above that is noise. That can get into your audio signal. A line filter is good to have in any amp. A second "line" of defense is shown below.

This board holds a parallel resistor, capacitor, and diode bridge (some use antiparallel diodes). For this to work, all jacks have to be isolated from the chassis. The cap (0.01uF, the value isn't critical) shunts high frequency noise to the chassis ground, the diodes conduct only if the voltage difference exceeds around +/- 1.4V, the resistor (10R, 5W ) holds a DC voltage. DIY kits are available, here is one source Link Removed, a schematic is on this page.

 

[BassmanPaul]

View attachment 802636
What is output voltage then?
Do DV heaters need to be 12v, or can they also be 6v?

With a bridge on the 6.3V secondary and filtered you'd get about 8.8V theoretically. You'd need to drop around 2.5V to return to 6.3V. As to a 12V heater supply it would depend on what tube were selected. Using double triodes only then a 12.6V heater supply could be used. That could be derived from the existing heater winding using a voltage doubler. That would give about 16V DC which could be regulated down to the 12.6V the tubes run on. Again these values are theoretical.

 

[Old Garage-Bander]

There's some discussion of this here: Valve Amps: Power-supply

Personally, I wouldn't bother, but the more stuff this thread brings to the table the better.

I agree. 12ax7's and their relations are easy to come by and its not like you're gigging in the trenches in Europe.
Thanks for the link.

DC heaters. The advantage of DC heaters is that the small induced magnetic humm leakage into the Cathodes is eliminated enabling the valves to be quieter than they would normally be.

There have been never-ending arguments for valve heater filaments (including output valves) to be supplied with DC rather than AC. High current DC has a disadvantage similar to the Electrolytic corrosion caused by vehicle batteries. Electrolytic corrosion effects cable, connectors and valve filaments resulting in reduced filament life. However, Cathode emission life is less than the reduced life caused by DC corrosion of the filament. Therefore, DC for valve filaments is not a real problem. A simple solution to minimise DC electrolytic corrosion is to switch (reverse) the polarity of the DC to the filaments approx each month. The heaters should be wired in series 12.6V so current is 1/2.

 

[winegamd]

With a bridge on the 6.3V secondary and filtered you'd get about 8.8V theoretically. You'd need to drop around 2.5V to return to 6.3V. As to a 12V heater supply it would depend on what tube were selected. Using double triodes only then a 12.6V heater supply could be used. That could be derived from the existing heater winding using a voltage doubler. That would give about 16V DC which could be regulated down to the 12.6V the tubes run on. Again these values are theoretical.

That is fine, because once you pass through an RC filter or two, you will drop those volts.

 

[rufus.K]

How about the triad transformer I posted?
I have it, readily available, it's $20 new, and we could start figuring things out.
Can it be used for the heaters, anyone opposed?
triad f212z
http://catalog.triadmagnetics.com/Asset/F-212Z.pdf

Also, let's assume we're going to use 3 or 4 12ax7 from here on out.

 

[rufus.K]

For the folks folllowing the thread who this is all new to:
What would you like more detail on?
What doesn't make sense?

 

[beans-on-toast]

How about the triad transformer I posted?
I have it, readily available, it's $20 new, and we could start figuring things out.
Can it be used for the heaters, anyone opposed?
triad f212z
http://catalog.triadmagnetics.com/Asset/F-212Z.pdf

Also, let's assume we're going to use 3 or 4 12ax7 from here on out.

If you are going to use a separate heater transformer, consider the Hammond 229 series. The design helps it to be quiet. Sometimes in a U2 space, this is the best way to go unless you want to mount on top. Have you decided if you want the tubes inside the chassis? This makes a big difference.

A single unit that has high voltage and heater windings can be used. A toroidal design would fit the bill.

 

[BassmanPaul]

If A single unit that has high voltage and heater windings can be used. A toroidal design would fit the bill.

I have a couple of suitable toroids that fit into a 1U case. IIRC the case and the toroid came from the same vendor..

 

[Passinwind]

If you are going to use a separate heater transformer, consider the Hammond 229 series. The design helps it to be quiet.

The 166 series that Mark and I mentioned has a split bobbin format too, and fits just fine in a 2RU case.

 

[beans-on-toast]

The 166 series that Mark and I mentioned has a split bobbin format too, and fits just fine in a 2RU case.

I've used those as well. Nice product. The hammond 229 is low profile but requires a PCB. The Triad that rufus asked about is a single bobbin, nice product but there are better choices.