Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Ground on tube filaments or floating?

Status
Not open for further replies.

neazoi

Advanced Member level 6
Joined
Jan 5, 2008
Messages
4,119
Helped
13
Reputation
26
Reaction score
15
Trophy points
1,318
Location
Greece
Activity points
36,918
Hi, in this transmitter, two 220v to 6v transformers are used back to back to feed the 12au7 filaments on parallel mode (6.3v) and also HV for it's anodes.

I wonder, shall I ground (connect to chassis ground) one end of the 6v side of the transformers (i.e. the pin9 of the 12au7) or shall I leave it floating (balanced)?
Which is the best way to do it?
 

Attachments

  • Capture.JPG
    Capture.JPG
    46 KB · Views: 312

hi,
In the past I have used a 'hum dinger' pot across the heaters, with the pot slider to chassis, trim to null any hum.
E
Look here:
http://www.valvewizard.co.uk/heater.html

extract:
Humdinger
The ways in which hum are induced into the audio circuit are rarely perfectly symmetrical. Often you will find that minimum hum occurs with an 'off-centre' tap. This can be done using a trimpot with its wiper grounded to create the artificial centre tap. Again, a low value is preferable, and a 500 ohm pot will dissipate less than 80mW at 6.3V. This 'humdinger' pot can then be adjusted on test for minimum audible hum, which seldom occurs at the exact centre setting.
 
  • Like
Reactions: neazoi

    neazoi

    Points: 2
    Helpful Answer Positive Rating
hi,
In the past I have used a 'hum dinger' pot across the heaters, with the pot slider to chassis, trim to null any hum.
E
Look here:
http://www.valvewizard.co.uk/heater.html

extract:
Humdinger
The ways in which hum are induced into the audio circuit are rarely perfectly symmetrical. Often you will find that minimum hum occurs with an 'off-centre' tap. This can be done using a trimpot with its wiper grounded to create the artificial centre tap. Again, a low value is preferable, and a 500 ohm pot will dissipate less than 80mW at 6.3V. This 'humdinger' pot can then be adjusted on test for minimum audible hum, which seldom occurs at the exact centre setting.

This is a so nice page, I am reading it now and thank you!
So you say I shall better leave the fillaments floating without a ground connection and twist their wires?
And then maybe use this humdinger if I want to reduce hum?
 

Neazoi, that's a strange way to power it up! It looks like the incoming AC is stepped down to 6.3V then stepped back up again for the HT supply. Why not feed the AC directly to two transformers, one for the heaters and one for the HT (or combine them in to one transformer), it would be far more efficient.

As for floating the heaters, it really depends on the capacitive coupling between the heating wires and the cathode. As one cathode is grounded and the other is at very low impedance to mains frequency, it shouldn't really matter whether you ground the heater supply or not. For safety, I would either ground one side directly or through a resistor (~10K) just to prevent any high voltages from the transformer leakage building up on the heater supply. The "humdinger" or similar circuits are a good idea on sensitive audio pre-amps but not really essential for RF applications.

If this isn't a tried an tested circuit, I would be careful about the current allowed to flow into the anodes, there is no bias on either triode so if it doesn't oscillate for any reason, the current will be very high.

Brian.
 
  • Like
Reactions: neazoi

    neazoi

    Points: 2
    Helpful Answer Positive Rating
When I used to build valve/tube amplifiers I always used twisted wire, say about 5 twists to the inch and also the humdinger did make a difference.

E
No direct connection of the heater wiring to chassis.
 
  • Like
Reactions: neazoi

    neazoi

    Points: 2
    Helpful Answer Positive Rating
Neazoi, that's a strange way to power it up! It looks like the incoming AC is stepped down to 6.3V then stepped back up again for the HT supply. Why not feed the AC directly to two transformers, one for the heaters and one for the HT (or combine them in to one transformer), it would be far more efficient.

As for floating the heaters, it really depends on the capacitive coupling between the heating wires and the cathode. As one cathode is grounded and the other is at very low impedance to mains frequency, it shouldn't really matter whether you ground the heater supply or not. For safety, I would either ground one side directly or through a resistor (~10K) just to prevent any high voltages from the transformer leakage building up on the heater supply. The "humdinger" or similar circuits are a good idea on sensitive audio pre-amps but not really essential for RF applications.

If this isn't a tried an tested circuit, I would be careful about the current allowed to flow into the anodes, there is no bias on either triode so if it doesn't oscillate for any reason, the current will be very high.

Brian.

The way with the two transformers has to do only with transformer availability. Obtaining 6.3v transformers from any electronics shop is easy and cheap. But custom made HV transformers are not found nowadays easily in local electronics shops.
I will try the humdinger only if I see significant hum. after all this is a longwave TX and hes it has been tested on an old magazine article of popular electronics. The tube is the ecc82 (12au7). However I cannot be sure that it will oscillate untill I build it!

By the way the oscillator section anode choke is specified as 10mH.
I am thinking of using the FT50-75 (longwave TX) because it is easier to find than an old 10mH choke.

The FT50-75 must have 60 turns for 10mH and the max wire size is 0.36mm or #28AWG. It is specified from about 500KHz and up as a choke, but it may work at 160KHz as well without problems.

The 12au7 draws 10mA at 250v, so I guess the 0.36mm wire should be fine.

Do you really think I can use this core in such high voltages?
If so, it will make things easy for me and as a bonus the winding capacitance should be kept quite low with only 60 turns, that barely touch at the outside perimeter of the core.
 

The choke should be fine, all it has to do is present a high impedance at the oscillation frequency so you can probably put more turns on it to make it work better. The wire and it's resistance isn't the issue, what concerned me was what would happen if it didn't oscillate for some reason. The 12AU7 may be specified to work at 250V and 10mA in normal operation but then there would normally be a negative voltage on the control grid (actually 0V on the grid but some positive voltage on the cathode). In that schematic, both triodes are run without grid bias or any cathode resistors so the current could go much higher. The oscillation in the first triode will limit it's current by developing a negative grid voltage (positive half cycles will be conducted from the grid to cathode as though it was a diode) and the signal reaching the second triode will similarly be rectified to produce a negative grid voltage. If the oscillation stopped the grid bias would be zero and they would both conduct as hard as they could!

Brian.
 
  • Like
Reactions: neazoi

    neazoi

    Points: 2
    Helpful Answer Positive Rating
The choke should be fine, all it has to do is present a high impedance at the oscillation frequency so you can probably put more turns on it to make it work better. The wire and it's resistance isn't the issue, what concerned me was what would happen if it didn't oscillate for some reason. The 12AU7 may be specified to work at 250V and 10mA in normal operation but then there would normally be a negative voltage on the control grid (actually 0V on the grid but some positive voltage on the cathode). In that schematic, both triodes are run without grid bias or any cathode resistors so the current could go much higher. The oscillation in the first triode will limit it's current by developing a negative grid voltage (positive half cycles will be conducted from the grid to cathode as though it was a diode) and the signal reaching the second triode will similarly be rectified to produce a negative grid voltage. If the oscillation stopped the grid bias would be zero and they would both conduct as hard as they could!

Brian.

I said 250v as I thought it would produce more output power. The original circuit works on 120v and the amplifier tube will draw 8.3mA (i.e 1W output). Should I stick to this lower voltage better for protection do you think?

About this you say: "to prevent any high voltages from the transformer leakage building up on the heater supply." Would the hum reducing potentiometer circuit help on that?
 
Last edited:

I said 250v as I thought it would produce more output power. The original circuit works on 120v and the amplifier tube will draw 8.3mA (i.e 1W output). Should I stick to this lower voltage better for protection do you think?
That would be 1W input power not 1W output. Provided the oscillator runs, it should be OK on 250V but personally, I would add a resistor in the anode of the oscillator stage, even if only 10K to limit the current and I would add a resistor of say 100 Ohms, bypassed with 100nF in the cathode of the output stage. It shouldn't change performance much but it would offer some protection.

About this you say: "to prevent any high voltages from the transformer leakage building up on the heater supply." Would the hum reducing potentiometer circuit help on that?

It would because it provides a leakage path to ground but I doubt you would notice any hum in that schematic anyway. It really matters in sensitive audio pre-amplifiers where even the tiniest of pick up would be audible after amplification. The reason I suggest fitting a resistor or grounding one side is that otherwise the heater circuit is 'floating' and could potentially rise up to quite a high voltage because of capacitive leakage through the transformers. The performance is not influenced by the heater connection because it is not connected internally in the 12AU7 but there is a risk if the heater floats to high voltage that it could flash over (arc or spark) to the cathode. Adding a resistor or ground link would leak away any charge before it could reach damaging voltages. For almost zero cost you add extra protection.

Brian.
 

That would be 1W input power not 1W output. Provided the oscillator runs, it should be OK on 250V but personally, I would add a resistor in the anode of the oscillator stage, even if only 10K to limit the current and I would add a resistor of say 100 Ohms, bypassed with 100nF in the cathode of the output stage. It shouldn't change performance much but it would offer some protection.



It would because it provides a leakage path to ground but I doubt you would notice any hum in that schematic anyway. It really matters in sensitive audio pre-amplifiers where even the tiniest of pick up would be audible after amplification. The reason I suggest fitting a resistor or grounding one side is that otherwise the heater circuit is 'floating' and could potentially rise up to quite a high voltage because of capacitive leakage through the transformers. The performance is not influenced by the heater connection because it is not connected internally in the 12AU7 but there is a risk if the heater floats to high voltage that it could flash over (arc or spark) to the cathode. Adding a resistor or ground link would leak away any charge before it could reach damaging voltages. For almost zero cost you add extra protection.

Brian.

Like this?

For CW keying the author used a switch (key) to break up the connection of the cathode of the final tube to the ground. He bypassed the key with a 10nf capacitor, I think this was for soft envelope shaping. How can I use this key circuit in conjuction to the cathode resistor and capacitor you propose?
 

Attachments

  • Capture.JPG
    Capture.JPG
    55.8 KB · Views: 149

That is exactly what I meant. I still think you can ground one side of the heater supply directly or through a resistor, I doubt you would notice any difference at all if you adjusted the potentiometer.

Keying the cathode will still work but do it this way: Keep the capacitor permanently connected from the cathode to ground, put the key in series with the ground side of the resistor. If you notice key clicks, wire another resistor and capacitor in series across the key (suggest 100nF & 100 Ohms). What you want to avoid is a very suddent rise or fall in cathode current and a high voltage across the key contacts, wiring it as I suggest should fix both problems. I would advise against connecting a capacitor directly across the key contacts (without its series resistor) as it will charge to ~25V with the contacts open then be shorted out as they close, you might get sparking and contact erosion.

Brian.
 

That is exactly what I meant. I still think you can ground one side of the heater supply directly or through a resistor, I doubt you would notice any difference at all if you adjusted the potentiometer.

Keying the cathode will still work but do it this way: Keep the capacitor permanently connected from the cathode to ground, put the key in series with the ground side of the resistor. If you notice key clicks, wire another resistor and capacitor in series across the key (suggest 100nF & 100 Ohms). What you want to avoid is a very suddent rise or fall in cathode current and a high voltage across the key contacts, wiring it as I suggest should fix both problems. I would advise against connecting a capacitor directly across the key contacts (without its series resistor) as it will charge to ~25V with the contacts open then be shorted out as they close, you might get sparking and contact erosion.

Brian.

Like this?
 

Attachments

  • Capture.JPG
    Capture.JPG
    57 KB · Views: 146

Thats correct.

If you want to further improve stability, you could stabilize the oscillator supply voltage by adding a Zener diode (100V to 150V) across the capacitor between the oscillator choke and load resistor. At the moment, as you close the key the supply will drop slightly and pull the oscillator frequency a little, the Zener would make sure the oscillator supply stayed constant. If you do that, calculate the load resistor so the Zener passes about 2mA, taking into account the current through the oscillator anode as well. so for example, if the oscillator consumed 5mA and the Zener passed 2mA the total through the resistor would be 7mA and if it dropped say 250V to 150V Zener voltage, it would be R=V/I =(250 - 150)/.007 = 14.285K or 15K as the nearest standard value.

If you want to avoid semiconductors (for the fun of it - although you are using a semiconductor bridge rectifier already) you could use a gas tube voltage stabilizer such as the 0A2. Not as efficient, much bigger, more expensive, but with a nice blue glow to look at!

Brian.
 

Thats correct.

If you want to further improve stability, you could stabilize the oscillator supply voltage by adding a Zener diode (100V to 150V) across the capacitor between the oscillator choke and load resistor. At the moment, as you close the key the supply will drop slightly and pull the oscillator frequency a little, the Zener would make sure the oscillator supply stayed constant. If you do that, calculate the load resistor so the Zener passes about 2mA, taking into account the current through the oscillator anode as well. so for example, if the oscillator consumed 5mA and the Zener passed 2mA the total through the resistor would be 7mA and if it dropped say 250V to 150V Zener voltage, it would be R=V/I =(250 - 150)/.007 = 14.285K or 15K as the nearest standard value.

If you want to avoid semiconductors (for the fun of it - although you are using a semiconductor bridge rectifier already) you could use a gas tube voltage stabilizer such as the 0A2. Not as efficient, much bigger, more expensive, but with a nice blue glow to look at!

Brian.

I am amazed about your answer. In a similar LW transmitter (using bigger tubes), the author mentioned regulated voltage for the oscillator section, but no other info. This explains now why he did so!
I would like to use an OA2 (I have lots of them). Can I shunt connect the OA2 after the oscillator anode resistor (close to the oscillator tube, before the choke) you have proposed for protection, or does it need an extra resistor and connected before it?
 

The 0A2 should be connected to the junction of the choke and resistor and to ground. It works best with a little more than 5mA flowing through it so adjust the resistor value accordingly. The capacitor across it must not be greater than 100nF, I would suggest 10nF. If you use one too big, the 0A2 will oscillate by itself at a few Hz!

What frequency are you using this at? I might be able to listen out for it on the air!

Brian.
 

The 0A2 should be connected to the junction of the choke and resistor and to ground. It works best with a little more than 5mA flowing through it so adjust the resistor value accordingly. The capacitor across it must not be greater than 100nF, I would suggest 10nF. If you use one too big, the 0A2 will oscillate by itself at a few Hz!

What frequency are you using this at? I might be able to listen out for it on the air!

Brian.

The original design was for 160-160KHz, but in EU we only have 135-137KHz so I will scale it there. My intention is to build this (if time permits) for more bands 160m and up. It should be only a matter of changing the oscillator and output coil and the choke.
I do not think it can radiate that far, it is only 1W or so and Greece to UK is a long distance. It can be easily done with a few 10s of W though. Once I have spoken to an operator in uk with a simple wire and antenna tuner.
 

I do not think it can radiate that far, it is only 1W or so and Greece to UK is a long distance. It can be easily done with a few 10s of W though. Once I have spoken to an operator in uk with a simple wire and antenna tuner.

It may be possible to get it across a long distance; the shortwaves can travel across the world after atmospheric reflections but these long waves can travel direct and bend across. If you have a decent antenna...
 

The antenna at this end isn't very good, all I have is a 'long wire' stretched between trees in a forest and an automatic (Icom) ATU. It works reasonably well on 160m but I doubt it would work at 135-137KHz, even if my transmitter covered that frequency. I would have to bypass the ATU - that involves ladders and climbing trees so it would be a last resort, I'm not as flexible or bounce-proof as I used to be!

Brian.
 
  • Like
Reactions: neazoi

    neazoi

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top