Elevated Heaters/DC heaters How?

[Ed_Chambley]

I was wondering if anyone would explain, or have a schematic showing DC heaters.  Also, John mentions he elevates heaters on octal pre-amp tubes.  I have heard of elevated heaters, but do know what it actually is.

I have made DC heaters, but I an not sure I do it correctly.   So there are 4 different heater supplies I know, but I do not know the differences, or if there is any difference to speak of.

Traditional ac center tapped 6.3 or artificial tap
dc heater
dc elevated (if different)
Is there an ac elevated and how is it done?

These are areas of interest I would like to play with as I am currently working on a design with 3 different preamps in one amp.  One is octal, one is 9 pin, and third will be AB 763.  I have not decided which actual pre the 2 will be, but it will be old tweed, second like a vox EF86 and third will be like a Twin Reverb.  This is the first where I will be using multiple relays and I figure it will take me some time, so I a simply gathering information.

PRR suggested 2, 6v6 running lower plates to achieve 10 watts in response as an alternative to 6AQ5 for the simple reason that Banjo World would have 6v6.  I assume he would not have mentioned this if there is not really much of a tonal difference and I do have 2 OT's designed for 2 6v6 so a nice savings for the build. I have done 2 in a stereo design, but have never used 2 in a mono design and do not know if you can in a traditional way like running a pair with each and a relay to switch out one of the OT's, but still have them combined at the speaker out.

Here is my dilemma.  I plan cathode bias, but would like to run 4 of them with a switch to remove 2.  If the OT for a pair 6v6 is usually on the range of 6k5-8k.  When running 4 it should be half, correct?  I am wanting to avoid, if possible, a mismatch when switching out 2 of the tubes.

Like when pulling 2 tubes from a Twin, it does decrease output, but it also changes the amp considerably. The common belief is it is like a half power, but how can it really be half power when you create a OT mismatch?

[Willabe]

Here's 2 drawings of a DC stand-off for heaters. 1 for if you have a PT heater winding with a CT and 1 for no PT heater winding CT.

Just woke up will write a little more later.


                 Brad      

[sluckey]

Traditional ac center tapped 6.3 or artificial tap

Most common. The center tap (real or artificial) is usually connected to ground. This means the filament circuit is REFERENCED to ground. We usually think of ground as zero volts.

dc heater

Simply build a 6 volt power supply and connect the filament string to it.

dc elevated (if different)

This has nothing to do with dc heaters. This applies to traditional ac heaters but with one difference. The center tap is connected to a positive voltage source rather than ground. No current flows through the center tap connection. But now the filament circuit is referenced to a positive voltage, rather than ground (zero volts). IOW, you could now say the filament circuit is ELEVATED to some positive voltage. There are two popular ways to get this positive dc reference voltage. For cathode biased amps, simply connect the filament center tap to the cathode of the output tubes. Or you can use two resistors and a filter cap to make a voltage divider and connect to the B+ rail.

Is there an ac elevated and how is it done?

no

If the OT for a pair 6v6 is usually on the range of 6k5-8k.  When running 4 it should be half, correct?  I am wanting to avoid, if possible, a mismatch when switching out 2 of the tubes.

That's correct. Said another way, when pulling two tubes the primary impedance doubles, so to avoid the mismatch you would double the speaker impedance. In the case of the TR, if you pull two tubes, disconnect one speaker to keep the impedance ratio the same.

[Ed_Chambley]

Here's 2 drawings of a DC stand-off for heaters. 1 for if you have a PT heater winding with a CT and 1 for no PT heater winding CT.

Just woke up will write a little more later.


                 Brad      

Thanks Brad and Steve, but I am going to need this painted with a bigger brush to break the ignorance I have with it.

I get the elevated.  No problem.

DC is alluding me.
I see in the schematic that the dc is taken from B+.  Are you still running the filament wires from the 6.3 PT taps?  If so, how does the once AC 6.3 tap become DC.  My thinking is the 6.3 taps are ac from the primary.

Are the filament string still twisted and wired in the same manner.

[John]

A couple things I have read: elevating the heaters is desirable when using a cathode follower, and DC filaments are most useful on the preamp tubes. The power tubes make enough noise to swamp any hum. (not the technical term, but it's how I remember it and I'm too lazy to go to Merlin's site  )

In my experience, the humdinger pot is worth the extra trouble if you're using the preamp octals. Really, once you do it once it's easy enough to just incorporate it in any build, IMO. I've made a little board with the pot and lugs for doing the voltage-dropping diodes (and resistor for LED) to keep everything more compact and neat. Although with my stuff, "neat" is relative.

[sluckey]

I see in the schematic that the dc is taken from B+.  Are you still running the filament wires from the 6.3 PT taps?  If so, how does the once AC 6.3 tap become DC.  My thinking is the 6.3 taps are ac from the primary.

The filaments are still heated only by the 6.3VAC from the PT. The AC 6.3 is not magically changed to DC. The DC has nothing to do with heating the filaments. The DC voltage is simply riding on the 6.3VAC waveform. Actually, the 6.3VAC waveform is riding on the DC voltage. There is NO DC CURRENT FLOW.

In the common filament circuit with CT connected to ground, the center of the 6.3VAC waveform is referenced to ground (zero volts). IOW, the sine wave starts at zero and rises to 3.15V positive peak, then falls back thru zero to a 3.15V negative peak, then rises back up to zero. This completes one AC cycle. The filament voltage swings between +3.15v and -3.15v 60 times per second. Total swing is 6.3V.

When you reference (elevate) the center tap to a positive voltage (use 50VDC as an example), the sine wave starts at +50V and rises to +53.15V peak, then falls back thru 50VDC to a +46.85V peak, then rises back up to +50V. This completes one AC cycle. The filament voltage swings between +53.15V and +46.185V 60 times per second. Total swing is still 6.3V.

In this example where we elevated the filaments to +50VDC, if you connect your DC volt meter between either filament pin and ground, you will measure +50VDC. And if you switch your meter to read AC volts, you will read 3.15VAC from either filament pin and ground.

Are the filament string still twisted and wired in the same manner.

Yes. The only difference is where you connect the center tap.

[Willabe]

All right, I'm awake now.      

Tubenit spent a lot of time on this and did a really great job, IMO, on this thread about dc PSU's for relay's.      

It's for a 5dcv relay but if you change the PT secondary voltage you can get whatever dcv you need. On a new build it's much easer because you can buy a PT with an extra secondary tap, say 10acv or you could add a second small PT with the secondary acv and current you need.

http://www.el34world.com/Forum/index.php?topic=14348.0


                      Brad      

[Willabe]

Kevin O'Connor's TUT books have several sections scattered in different volumes on dc stand-offs, dc heaters, regulated dc heaters and dc heaters with a dc stand-off.    

He's very clear in them that he finds a noise reduction (up to 30dB) with dc-elevated heaters that's well worth the few parts and space it takes to do this in a new build. Now he does have a lot of hi-gain amps in his books that might benefit from this more than a more standard gain amp.

This is an old idea that has been used for decades on different production amps and Hi-Fi stereos. As Sluckey pointed out 1 way to do this was they would tap a dcv from the power tube cathodes on cathodes bias amps. Look at old Ampeg amp schemos.

Kevin's books list 70 to 80 dcv for the stand-off, most guys here that I've seen use a dc stand-off are using 30 to 40 dcv and report very good results. (Kevin doesn't think you get the full noise reduction that you could get using ~75dcv.) On my 5E3, 5G9 and stand alone reverb I used ~75 and I think their very quite. (Now I also did use KOC's grounding too. It's a wired grounding scheme with all jacks isolated/insulated from the chassis and only 2 chassis ground connections, 1 for the circuit ground, the other is the power cord 3rd wire safety ground.)

As far as dc stand-off versus dc heaters, IIRC Kevin has said a number of times that a dc stand-off can be as effective or better than dc heaters. Merlin has said if dc heaters are not done correctly they will actually be noisier than ac heaters. He explains himself but I forget what he wrote.

I can't find it right now but I remember Kevin recommending going the distance by using dc heaters with a dc stand-off for things like mic pre's or any other very low signal circuits where any extra noise reduction would be desired.

Kevin and others (Power Scaling forum) have said many times they believe regulated dc heaters are unnecessary as they are overkill and offer no audible difference than just going with a dc heater with dc stand-off.  

Now if you had a proto bread board (like Pete, Richard, SG and others here and I think Pete and Richard have done some hands on tests.  ) along with a scope and maybe a few other pieces of tech test gear you could prove it on a micro level, but......    

To me, bottom line, best bang for the buck, considering what I've read, have heard here, have experienced in my few builds using it, cost of parts, chassis space and wiring time versus noise reduction in a guitar/musical instrument amp?

I'm sticking with a simple dc stand-off but at ~75dcv.  

What John said is true about saving a modern production 12 _ _ 7 tube from shorting out from being used as a CF because of high voltage between the K and filament. A number of guys have had that problem and posted about it here.

I also still want to try adding the hum-dinger pot that John brought up with the dc stand-off. It may well be worth the extra pot and hole in the chassis.



                         Brad            

[tubeswell]

Here's one I did recently showing the heater elevation (which doubles as the bleeder network for the filter caps)

[Willabe]

Last thing;

As I understand it, All the different dc heater circuits are to stop heater to cathode leakage because it can show up as noise being injected into the tubes grid and being amplified along with the ac audio signal.

HBP (jjasilli too?) has explained this several times technically in very good detail in the not so long ago past. I should have memorized it.    

Something about a diode effect between the heater and cathode? Adding the dc stand-off voltages reverses this effect?


                    Brad      

[Ed_Chambley]

All right, I'm awake now.      

Tubenit spent a lot of time on this and did a really great job, IMO, on this thread about dc PSU's for relay's.      

It's for a 5dcv relay but if you change the PT secondary voltage you can get whatever dcv you need. On a new build it's much easer because you can buy a PT with an extra secondary tap, say 10acv or you could add a second small PT with the secondary acv and current you need.

http://www.el34world.com/Forum/index.php?topic=14348.0


                      Brad      

Great thread, agreed!  I understand the concept now.  A little work for implementing and I will get it.  I never really get things until I do it wrong and have to figure it out.

[Tone Junkie]

Ed here is one that shows a 12 volt DC heater arrangement. Although I use a 9-0-9 transformer and a 4700uf 63volt cap I just thought 24volts to make 12 was over kill and and the 4700 uf was what I had laying around .  That LM7812 is good for 1 amp if I'm not mistaken, and in 12 volt each tube takes .165 instead of.33 like 6.3 volt
Bill

[HotBluePlates]

Kevin's books list 70 to 80 dcv for the stand-off, most guys here that I've seen use a dc stand-off are using 30 to 40 dcv ...

Depends on your reason for elevating.

Some 9-pin tubes are only rated for 90v heater-to-cathode. What do you do when you use them in a cathode follower circuit with the cathode more than 90vdc above ground? Or use the dual-triode in a cascode circuit where the 2nd triode's cathode is directly tied to the 1st triode's plate (very common in the 90v h-k 6DJ8)? You really oughta elevate the heaters for at least those tubes to avoid heater-to-cathode shorts.

I've seen the same in tube-regulated power supplies (either standalone or in tube oscilloscopes): a high positive voltage (sometimes the output voltage of the regulator) is applied to the heater center-tap for the tubes being used in at least the series-pass-tube positions, as these are often power output tube types and have their cathodes sitting at 250-500vdc (or more). These types can't take the strain without the heaters running at an elevated d.c. reference, because they are not like rectifiers which are either directly-heated (cathode is the filament) or have an internal  connection between the heater and cathode.

As I understand it, All the different dc heater circuits are to stop heater to cathode leakage because it can show up as noise being injected into the tubes grid and being amplified along with the ac audio signal.

Something about a diode effect between the heater and cathode? Adding the dc stand-off voltages reverses this effect?

Thermionic Emission.

Your heater heats the cathode, the heat liberates electrons, which then flow through a not-too-negative grid to a positive plate.

In some (generally rare) cases, the heater may emit electrons through compromised heater-cathode insulation towards the cathode. A 0v-referenced heater is less-positive than then cathode (which is likely cathode-biased and at 1-2 volts) during most of the a.c. heater swing. So when the heater voltage swings negative-enough of the cathode voltage, current passes from heater to the cathode as in a diode. This current gets summed with the intended audio signal and contributes a buzzy hum.

In this situation, the heater is the defacto cathode of our "diode" and the tube cathode is the "diode plate". For current to flow through this diode, the cathode has to be positive of the heater, just like a tube's plate has to be well positive of the cathode to pass current.

So when you apply a elevated reference voltage, you're making the heater (aka "diode cathode") too positive to pass current to the cathode (aka "diode plate"). The "diode" is reversed bias, and hum stops.

The technique is most useful on the first stage or two, as the hum current is generally small and amplification afterwards makes the hum more obvious (some old & new gear only elevated or used d.c. heaters for the first stage or two). CA cited a pretty wide range of effect voltages to use as a reference for heaters. O'Connor may be considering the use of cathode followers in an amp circuit, and elevating more to keep those tubes happy.

Too bad there is also a heater-to-cathode leakage mode that is more capacitive in nature and is not generally cured by elevated heaters (ask David Allen  ); this is also an individual-tube problem, with some leaking and some not. In these cases, the leakage gets swamped when you use a big-big cathode bypass cap, which is one of the real reasons (in my opinion) you see a 250uF cathode bypass in the first stage of the 5F6-A Bassman (low hum/noise is also why Fender used the 12AY7 in the input stages of tweed amps, where these problems would be most critical).

[LHPcope]

As I understand it, All the different dc heater circuits are to stop heater to cathode leakage because it can show up as noise being injected into the tubes grid and being amplified along with the ac audio signal.

This is true if a particular tube has a heater to cathode leak.  As a side note the defective tube may well change the amount of leakage with its temperature.  Think interment hum problems getting worse the longer the amp is on.  I have seen some power tubes with leakage so bad that the elevated DC would not cure the hum.

[marshallguy]

Here's one I did recently showing the heater elevation (which doubles as the bleeder network for the filter caps)

I have a question. If the filament center tap is normally referenced to ground or connected directly to ground, but now it has 22,00 ohms of resistance between it and ground because of the 22k being used as part of the voltage divider used to DC elevate it, how does that not interfere or reduce the amount of current now provided to the 6.3vac filaments? I would assume some current is lost or total current provided by the transformer specs is effected.

JC

[pdf64]

The reference just sets the DC voltage between the heater circuit and 0V common.
No load current flows via the reference.

[HotBluePlates]

... resistance between it and ground because of the 22k ... how does that not interfere or reduce the amount of current now provided to the 6.3vac filaments? ...

Restating pdf64's point more explicitly:

   Disconnect the center-tap; current still flows from the 6.3v winding to/through the tube heaters.

   The transformer winding, wiring, and tube heaters form a closed-loop.
   Current flows through/around that closed-loop.
   The Heater Circuit does not include Ground as part of its closed-loop current path.
   Other circuits do use Ground as part of their closed-loop, so disconnect-from-Ground stops current flow.

[acheld]

One (not always) obvious additional point is that since there is no significant current through the center tap ground, a resistor placed into that path does not affect the voltage potential of the center tap itself.   

Ohm's Law strikes again . . .     

[PRR]

> how does that not interfere or reduce the amount of current

Bird on a wire.

Electricity ALWAYS goes round in circles. There's no loop through the elevation link.

[marshallguy]

... resistance between it and ground because of the 22k ... how does that not interfere or reduce the amount of current now provided to the 6.3vac filaments? ...

Restating pdf64's point more explicitly:

   Disconnect the center-tap; current still flows from the 6.3v winding to/through the tube heaters.

   The transformer winding, wiring, and tube heaters form a closed-loop.
   Current flows through/around that closed-loop.
   The Heater Circuit does not include Ground as part of its closed-loop current path.
   Other circuits do use Ground as part of their closed-loop, so disconnect-from-Ground stops current flow.

Thanks everyone for clarifying that. A close loop ac current draws between the 2 sides of each tube.

What I did years ago was DC elevated off the main B+ using a 270k / 33k voltage divider with 100u / 100v cap and ran that to the center point of a humdinger 100 pot with 100 ohm on each side of pot to the filaments. I was able to ac balance and 50vdc elevate together. High gain modified  amp quiet as a mouse from efforts. I was redrawing the schematic and wondering if I fooked up even though it had worked for so long. I’ll post schematic when I get back home, wife and I in Jackson Hole WY on our way to Yellowstone and points in between.

Other circuits that do use the C.T. as part of closed loop, say for example a 12vac transformer with C.T. used as a full wave rectifier would have C.T. current being drawn to supply a circuit like a +5vdc regulator with ground as reference. That C.T. could be a place to insert a fuse to protect it all?  Same goes for a bridge rectifier?

One of the FNG’s I use to work with years ago slipped with 90 degree cable on a Localizer ILS and shorted the power supply causing a quick spark. He disappeared left me to troubleshoot.
What initially threw me off was that the fuses were good and indicator power light was lit but no RF power out. It would transmit on batteries but not AC.

After digging into the books and examining the schematic, I found I believe if memory serves me a .05 or 1 ohm 25w wire wound resistor between ground and the bridge rectifier being used (no C.T.). I located it buried in the chassis and measured high or about 50k instead. That now explained why the 28vdc power lamp would light on AC but not much else. It provided enough current still for some lights.

That resistor seemed to be to me nothing more then a protection resistor since the ohms were so low. What else could it be? Why not just fuse it instead and give it easy replacement access? Now I know how a wire wound resistor can ruin your day.

[marshallguy]

Update…schematic…..found it! On my iPad photos.

You can see my heater elevated and how I set it up. The .05 resistors are 25w to reduce filament voltage.

When I Return home I’m going to replace the B50k volume pot with an audio taper and lower the B+ 10k 5w after the screens to bump up the B+ another 20v.
     Why I have just a higher B+ tranny in the first place is another long sad story. I have enough leftover current on the 6.3vac that a I didn’t need the 5vac tap to run a few 5v relays and LDR’s.

Anyway on this amp I’m having still an issue.

It’s a combo (original Marshall Frankenstein version of the 4210) but even with the amp on the bench I’m getting a ringing behind notes played as I turn it up a little. Doesn’t need to be too loud to hear it. Both channels. Very annoying. I don’t recall hearing that in the past, had the amp since 1983.

Here’s what I’ve tried so far…..Replaced EL34’s with another set, no change.
Place 220p caps on the output tube grids to ground, no change.
Add a tone control circuit, (removed now) no change.
Moved the 20v zener from in series to screens (@467v on plates and @445v on screens in that set up) to series with 56v (as shown on circuit schematic with ref. Volt. points) no change.

Type speaker listed on schematic.  I’m running out of options. Next thing I’ll do is run the amp into the same 2x12 closed cabinet the other amp uses and eliminate the combo cab as possible cause. I even thought of placing a couple 1kv small caps in series across the plates and see if that helps but have stumbled across anyone else yet doing it. My logical mind says it’s the combo cabinet causing it but who knows could be the output tx.

Anyone come across note ringing? I don’t see anything on the scope on the output.

I apology I’m on my iPad and it’s difficult to copy and paste right at times. If I should have started a new trend again I apologize.

[pdf64]

If you can’t see on the scope what you’re hearing then you’re pretty much chasing a will ‘o the wisp.
Are you sure it’s not a bad speaker / something vibrating in the room / EMI?
eg try in in a different building through a different speaker cab.

[PRR]

I suspect a drawing error around here.

[marshallguy]

If you can’t see on the scope what you’re hearing then you’re pretty much chasing a will ‘o the wisp.
Are you sure it’s not a bad speaker / something vibrating in the room / EMI?
eg try in in a different building through a different speaker cab.

You’re probably right, when I get back from vacation I’ll do just as you suggest. I didn’t have time before will left to plug into my 2 x 12 closed back cabinet. I actually have an extension cabinet made specifically for the 4210 with same speaker to try too. Going to be at least til Monday following week before I can.

[marshallguy]

I suspect a drawing error around here.

Thanks, and yes, one of a few I keep finding, 270k definitely not located as drawn.

[sluckey]

What happened to my bias lines? 🤔

Same thing that happened to 95% of your schematic. The bias lines were clipped out because they were not relevant to the point PRR was making.

[marshallguy]

What happened to my bias lines? 🤔

Same thing that happened to 95% of your schematic. The bias lines were clipped out because they were not relevant to the point PRR was making.

Good point!
Having said that, I use Windows paint to draw my schematics. In order to remove the bias lines to make a point as you said, one would have to download the schematic, opened in paint or some other program, remove the lines, then save to repost it as it is. Seems like a lot of extra work when simpler alternatives are available to make the same point.

Maybe not….

Some guys don’t use pencils because they don’t make mistakes…I’m not one of them 😌

[AlNewman]

... resistance between it and ground because of the 22k ... how does that not interfere or reduce the amount of current now provided to the 6.3vac filaments? ...

Restating pdf64's point more explicitly:

   Disconnect the center-tap; current still flows from the 6.3v winding to/through the tube heaters.

   The transformer winding, wiring, and tube heaters form a closed-loop.
   Current flows through/around that closed-loop.
   The Heater Circuit does not include Ground as part of its closed-loop current path.
   Other circuits do use Ground as part of their closed-loop, so disconnect-from-Ground stops current flow.

I have an additional question...

What happens if one side of the ac winding has a different potential to ground than the other through the center tap?  Either short, or open, or somewhere in between?

[Jalmeida]

Not trying to hijack. But on-topic with DC heaters. I see a few amps using DC heaters just for the first preamp stage. Why not run all DC heaters? 

[acheld]

Why not run all DC heaters?

See  http://www.valvewizard.co.uk/heater.html

Basically, it costs efficiency, and puts more strain on your transformer which was designed to supply AC.  Usually, your input stage benefits most from clean DC.

[Jalmeida]

Why not run all DC heaters?

See  http://www.valvewizard.co.uk/heater.html

Basically, it costs efficiency, and puts more strain on your transformer which was designed to supply AC.  Usually, your input stage benefits most from clean DC.

It has been a while, but I have read that section. I assume by efficiency you mean added current of a rectifier circuit? Forgive my question which may be perfectly obvious to everyone else but me. But how much current increase(say percentage) roughly would one expect by adding a rectifier on the heater secondary? Seems like something that could be anticipated on the front end of a build by using a higher current transformer.

And also just spitballing conceptually, why couldn’t someone pull DC of the main power supply node like the DC ground reference standoff and drop the voltage using a voltage divider or something similar to power 1-2 preamp tube heaters? Or is that plugging one leak to create another so to speak?

[sluckey]

And also just spitballing conceptually, why couldn’t someone pull DC of the main power supply node like the DC ground reference standoff and drop the voltage using a voltage divider or something similar to power 1-2 preamp tube heaters? Or is that plugging one leak to create another so to speak?

One or two preamp heaters would require 6V at 300mA or 600mA. Can your PT high voltage winding supply that much current?

[pdf64]

I seem to recall an obscure cathode bias class A amp that used V1 heaters in series as the cathode resistor (with a bypass cap).

[Jalmeida]

And also just spitballing conceptually, why couldn’t someone pull DC of the main power supply node like the DC ground reference standoff and drop the voltage using a voltage divider or something similar to power 1-2 preamp tube heaters? Or is that plugging one leak to create another so to speak?

One or two preamp heaters would require 6V at 300mA or 600mA. Can your PT high voltage winding supply that much current?

Oh yeah!! That was the obvious answer I 100% missed. Most of these filament secondaries in amps as opposed mA. Thanks.

[PRR]

> Can your PT high voltage winding supply that much current?

Several quad power tube audio amps fed first heaters with power amp cathode current. Four 6V6 can pass 150mA; in two 12AX7 this is like 25V drop, close enough. I think I have seen it on two hot 6L6.

[passaloutre]

I have a ca. 1960 “hifi” amp that runs the V1 heater off the power tube cathodes. Seems to keep the phono pre quiet, but it has a looong warm time.

[kagliostro]

Ciao PRR

Several quad power tube audio amps fed first heaters with power amp cathode current

That is interesting but odd to me, never seen before

As the schematic linked by passaloutre seems apparently to show a (minus) -15v at the cathodes then 10.4v at the heaters, is possible to have a link to one other schematic that uses this architecture ?

Thanks

Franco

[sluckey]

@Franco... Here's one...

     https://el34world.com/charts/Schematics/files/Silvertone/Silvertone1474.pdf

[passaloutre]

It’s not negative 15, it’s an arrow pointing to the node.

[marshallguy]

Why not run all DC heaters?

See  http://www.valvewizard.co.uk/heater.html

Basically, it costs efficiency, and puts more strain on your transformer which was designed to supply AC.  Usually, your input stage benefits most from clean DC.

Agree, DC heaters too much trouble compared to just elevating and bal the AC filaments.

[marshallguy]

... resistance between it and ground because of the 22k ... how does that not interfere or reduce the amount of current now provided to the 6.3vac filaments? ...

Restating pdf64's point more explicitly:

   Disconnect the center-tap; current still flows from the 6.3v winding to/through the tube heaters.

   The transformer winding, wiring, and tube heaters form a closed-loop.
   Current flows through/around that closed-loop.
   The Heater Circuit does not include Ground as part of its closed-loop current path.
   Other circuits do use Ground as part of their closed-loop, so disconnect-from-Ground stops current flow.

I have an additional question...

What happens if one side of the ac winding has a different potential to ground than the other through the center tap?  Either short, or open, or somewhere in between?

Short or open = fault. Different potential to ground would introduce hum with all wire dressing being correct. A humdinger pot should correct that if not too far off.

[kagliostro]

@Franco... Here's one...

Many Thanks Steve

Franco

[HotBluePlates]

What happens if one side of the ac winding has a different potential to ground than the other through the center tap?  Either short, or open, or somewhere in between?

If the center-tap is actually in the center of the winding, potential-to-ground of both sides is the same.

But to your cases:
  - Short:  assume an entire half-winding is shorted.  Heater voltage to the tubes is halved, they might not get hot enough to pass sound.

  -  Open:  no current-path (because there is no other ground-connection to return current).  Tubes go dark & don't pass sound.

  - In-between:  there's no "between" for "open" so we set that case aside.  In-between for "shorted" seems to move the ground-reference away from "the middle" and closer to one of the ends of the winding.  You would think that would cause hum, but I've had amps with Hum Balance pots where you could actually move the ground reference to either side or anywhere in-between.  I found any setting of the pot except for all-the-way to an end was hum-free.  So I'm inclined to say this "partial-short" situation is no concern.

[Nathan]

Hey all,


I want to semi-hijack this thread because I am confused about the CF statement how the heater ''has to'' be elevated.


Per Valve Wizard: Some valve stages such as cathode followers require the heater supply to be elevated to avoid exceeding the valve's Vhk(max) rating.


Reason I'm asking is because I have 2 amps using a CF:
-Blues Jr (modified to add CF on unused triode) and
-Mesa Blue Angel (V1b)


I could be wrong, but I do not see anything about elevated heater in those cases?


Hope this isn't out of line?

[sluckey]

I never saw any heater elevation in the 5F6A Bassman or the famous Marshall amps we all love to copy. All had a CF driving the tone stack. Heater elevation became an issue about the same time Merlin began writing books. 

Not saying it's not a valid issue, especially with current production tubes, just sayin'...

[pdf64]

It’s not a new idea.
Tomer advised that biasing (ie elevating) the heater circuit was a good idea back in 1960.

[sluckey]

I wonder if anyone was listening.

[HotBluePlates]

Per Valve Wizard: Some valve stages such as cathode followers require the heater supply to be elevated to avoid exceeding the valve's Vhk(max) rating.

I never saw any heater elevation in the 5F6A Bassman or the famous Marshall amps we all love to copy. All had a CF driving the tone stack. ...

The max rating for plate voltage in a 12AX7 is 330v according to the data sheet.

In the 1950s, McIntosh placed 490v across the 12AX7 in their MC30 ("V4" in the schematic on the last page here).

It is best practice to stay within ratings and operate tubes conservatively.  But even in the old days, people didn't always stick to "best practice."

[pdf64]

I wonder if anyone was listening.

Gibson, Ampeg, Harmony

https://el34world.com/charts/Schematics/files/Ampeg/Ampeg_b15n_portaflex.pdf

https://el34world.com/charts/Schematics/files/Gibson/Gibson_ga20_12ay7_preamp.pdf

https://el34world.com/charts/Schematics/files/Harmony/Harmony_h306a.pdf

Unfortunately though, not Fender, so the copies / clones / inspired bys thereofs that make the vast majority of what’s around now don’t feature it either.

[Jonas]

If I pull my elevation from the same node as the phase inverter would I want to ground the voltage divider at the input jack or PT bolt?