Question about centre tapped filament supplies for vacuum tubes/valves. / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Question about centre tapped filament supplies for vacuum tubes/valves.

1 2 next

Move Thread

LAN_403

Ash Small

Thu Mar 24 2016, 01:20PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I have a power transformer I'm using for a valve amp which has both centre tapped and non centre tapped filament windings.

My question is when should I use the centre tapped windings, and when should I use the non centre tapped windings?

Thanks in advance

Sulaiman

Thu Mar 24 2016, 01:43PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

From vague memory, centre-tapped filament supplies are used for low hummmmm
so probably the early stages of amplification / input stages.

Proud Mary

Thu Mar 24 2016, 02:07PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

AC heater lines in audio equipment (and other sensitive applications) need to be balanced to diminish their EM field or disagreeable hum will result, especially in the first audio amp stage. Balanced heater lines not only halve the voltage on each wire i.e. 3.15V-0V-3.15V instead of 6.3V-0V but the contrary fields of a carefully twisted pair will tend to couple equal-but-opposite hum signals into the audio circuits, which will go some way to mutual cancelation.

While you're at it, the heater winding centre-tap to Earth is a good and popular place to locate the heater fuse.

There are, of course, plenty of circuits that couldn't care less about hum, thermionic diode rectifiers for example, and the 6.3V-0V winding can be used to supply these.

Ash Small

Thu Mar 24 2016, 02:58PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

So maybe it would make sense to add a bridge rectifier and smoothing cap to the heater supply?

I don't need to worry about the 5V floating supply for the GZ37 rectifier tube as the filament is tied to the cathode internally anyway?

I just need to worry about the 6.3V 300mA supply to the ECC82 and the 6.3V 900mA supply for the 807.

Proud Mary

Thu Mar 24 2016, 04:20PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Ash Small wrote ...

So maybe it would make sense to add a bridge rectifier and smoothing cap to the heater supply?

You have to consider the forward voltage drop across the bridge elements which will rise with increasing current. This will likely be 0.6V - 1.1V per diode, but you will have to consult the data sheet. As you will have a diode in the forward and return legs of your current flow you will drop approx 0.6V - 1.1V across each one, giving a Loss Grand Total in the range 1.2.V - 2.2V. Trust me when I tell you that the valves won't like this sadly diminished heater voltage at all, and will start grumbling about electron starvation and even cathode stripping if they are especially put out. But there is a chink of light in this desperate picture: if you put a stout-hearted capacitor after the bridge your DC may turn out higher than your AC RMS transformer voltage (i.e. DC~1.4xVunregRMS). What you loses on the diode forward voltage drops you gains on the smoothing of the peaks..

Ash Small wrote ...

I don't need to worry about the 5V floating supply for the GZ37 rectifier tube as the filament is tied to the cathode internally anyway?

That's why the rectifier's 5V cathode heater has a transformer winding all of its own, as the whole winding will be at HT potential.

Ash Small wrote ...

I just need to worry about the 6.3V 300mA supply to the ECC82 and the 6.3V 900mA supply for the 807.

Are you making an HF CW transmitter? If so, I suggest putting some home-made VHF RF chokes in the heater lines close to the valve sockets, and adding some bypass capacitors (say 1000pF) from the heater pins to chassis to divert to Earth any unwanted RF currents induced in the heater supply.

PS: Directly heated valves have filaments, indirectly heated ones have heaters.

Ash Small

Thu Mar 24 2016, 08:43PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Proud Mary wrote ...

Are you making an HF CW transmitter? If so, I suggest putting some home-made VHF RF chokes in the heater lines close to the valve sockets, and adding some bypass capacitors (say 1000pF) from the heater pins to chassis to divert to Earth any unwanted RF currents induced in the heater supply.

Thanks for the info, PM.

I'm building an AF amplifier, using an 807 strapped as a triode, because I have plenty of them. Link to amp thread here:

I think the 807 is basically a 6L6 with the anode connected to the top cap. I'm not trying anything new here.

I'm sort of counting on the rectified peak voltage being high enough to cope with the diode drop. I don't need to do this now, I can try AC first. I'd even thought of tying the centre tap to the cathodes of the ECC82. I have two 6.3V windings, so can use the other for the 807. I also have 12.6V heater windings which I could also use for the ECC82, I'm not sure if this would increase hum though.

So many choices. (I've been advised by the designer of the circuit to just cut out the centre tap connection, and tie the heater supplies to the cathodes, as he says that eliminates hum as well, but I'd assume it would make more sense to tie the centre taps to the cathodes.)

Proud Mary

Thu Mar 24 2016, 11:32PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

The 807 is an excellent valve, and robust too.

I'll have a look in my shed and see if I have any of the other parts indicated on the circuit diagram that I could give you.

Hazmatt_(The Underdog)

Fri Mar 25 2016, 03:30AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Valve amps are a little tricky, we learned this with the amps I showed-off in your other thread, they are typically noisy.

Center tapped transformers are better for hum cancellation then non-tapped transformers, but you can add a hum-balance control to try to null the hum out of the circuit.
But I think a center tapped transformer offers more leeway when trying to eliminate noise.
The typical method is to chassis ground the center tap, then the hum noise are phase cancelling, but this doesn't always work, and some preamp tubes are noisier then others, typically the higher gain tubes add a lot of noise, so for a lower noise system you would want something more like a mu of 40, rather than 75 or 100.

To get around part of the noise issue, one method is to source a DC bias to the center tap of the transformer and ground, I've never tried this, but a DC voltage under 50 volts is supposed to eliminate the hum caused by the filament.

DC power to the filament does not eliminate hum, what it does is shift the hum up the spectrum, so this is not really desired. Over there, rectifying the filament supply shifts the voltage up x 1.4 so now you have too much voltage into the tube, AND the added consequence of 100 Hz charging ripple, causing 100 Hz hum in your audio rather than the 50 Hz hum from the AC.

All in all, I think the lower gain tubes work well with a center tapped transformer, and you can get rid of most of the noise without too much trouble.

Ash Small

Fri Mar 25 2016, 09:58AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Proud Mary wrote ...

The 807 is an excellent valve, and robust too.

I'll have a look in my shed and see if I have any of the other parts indicated on the circuit diagram that I could give you.

That's very kind of you, PM.

I think I pretty much have everything I need, I'm not certain about the potentiometers, though.

I'm interested in interstage transformers at the moment, I'm certainly after more of those to play around with, also suitable cores upon which to wind some.

I've been researching the three main types, and their advantages/disadvantages. Bifilar ones tend to use the stray capacitance for extra coupling, and aren't suitable for phase reversal, according to my research.

Interleaved and segmented windings also seem to have problems of their own. I've read a number of threads regarding DIY IT's, and, like OT's, no-one seems to be getting it right.

I've found a supply for FormVar enamelled wire, coated with urea-formaldehyde-alkyd, which has a lower dielectric constant than 'modern' enamelled wire, and even found the original Westinghouse patent

Plumber's teflon tape (PTFE) also seems to have a low dielectric constant. Like Nomex, the dielectric constant depends, to a point, on density. Apparently plumber's tape is less dense than the variety that gas technicians use.

@Hazmatt: I plan to try all those methods if necessary, but initially I'll follow the advice of the designer of the circuit. He, like me, understands the advantages of 'keeping things as simple as possible'.

I read an interesting article describing how to connect a 500 Ohm trim pot, wiper to the centre tapped wire, and ends to the filament wires, to 'fine tune' hum cancellation.I don't really like the idea of adding extra resistors, or extra wiring to tie it to a higher DC voltage.

The idea of using a 6V lead acid battery, or similar, seems to make more sense if hum turns out to be a problem.

The designer assures me that if I stick to the design he's perfected over the years I won't have any problems, but I like to research stuff in depth.

It's common practice to put a fuse in the HT return (centre tap), but that won't work with the ex military Parmeko power transformer I'm using, due to all the centre taps being wired together. The screen is also tied into this.

Proud Mary

Fri Mar 25 2016, 11:42AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

I see from a datasheet that the max allowable peak heater to cathode voltage for 807 is 135V. Your design won't reach that figure, I take it?

I've had a rummage through my VRs and couldn't find anything close to the specs in the circuit diagram.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob