PSU for AC filament heating

[flatresponse]

Hi,
sorry, when this topic has been discussed already and sorry if it´s in the wrong place. I´m not a total newbie, but I very seldom go mind-funding on the fores. In this case the task exceeds my experience.

Given, you want to build a passive power supply of a 50ies tube design. You have a given numer of tubes, each one asking for 300mA filament heating. How do you design a power transformer for the 6,3V heating ? Does the AC supply have to equal the power consumption (VA) of the tubes ? In my case the problem is that even with connected filament the voltage is still too high.
I don´t want regulated power supply and I don´t want DC heating, because I want to recreate a historical design - call me out of date or whatever. But, does someone know a rule to it ? I have the possibility to design a torroidal transformer to my specs and yes, maybe I have to insert load resistance, but preferably I want to get along without it.

Thanks for helping, Manfred.

 

[rock soderstrom]

If you want to have a dedicated transformer for your project, talk to a transformer winder, they know how to do it, you just provide the target specifications.

The question is, what is your heater voltage under load? A case mounted load resistor or two can easily adjust the over voltage.

 

[Tubetec]

Maybe for preamps you might get away with resistances , for power amps your going to throw away even more energy as heat . Best get a suitable transformer for the job you want to do . Not all tubes draw 300ma , power tubes like EL34 draw a lot more current on the heaters .

Heres one possible option ,
Netztrafo für Röhrenverstärker, z. B. 2x EL34 Ultralinear | eBay

If your only powering a preamp tube or two from this your heaters are going to be overvolted by a large amount .

 

[abbey road d enfer]

Hi,
sorry, when this topic has been discussed already and sorry if it´s in the wrong place. I´m not a total newbie, but I very seldom go mind-funding on the fores. In this case the task exceeds my experience.

Given, you want to build a passive power supply of a 50ies tube design. You have a given numer of tubes, each one asking for 300mA filament heating. How do you design a power transformer for the 6,3V heating ? Does the AC supply have to equal the power consumption (VA) of the tubes ? In my case the problem is that even with connected filament the voltage is still too high.
I don´t want regulated power supply and I don´t want DC heating, because I want to recreate a historical design - call me out of date or whatever. But, does someone know a rule to it ? I have the possibility to design a torroidal transformer to my specs and yes, maybe I have to insert load resistance, but preferably I want to get along without it.

Thanks for helping, Manfred.

Usually, transformers' power rating is given for a nominal input voltage and a nominal load.
Let's say you order a 6.3Vac xfmr rated at 50VA, it will deliver 6.3V when the mains voltage is excatly the nominal voltage (230v in Europe) and when the load is such that it draws exactly 17.75A 7.9A.
If the mains voltage is higher or lower (the supplier is permitted a variation of +/-10%), the 6.3v secondary will vary accordingly.
As well, when the current draw is less than nominal, the voltage at the secondary will increase by a factor that depends on the construction of the xfmr. This parameter is sometimes included in the datasheet.
Tube manufacturers are well aware of these issues and tubes are usually quite tolerant to heater voltage variations.
However, the combination of high mains voltage and too big a transformer can be an issue that decreases tube life.
Please note that this is valid for resistive loads, which tube heaters are. Reactive or non-linear loads (rectifiers + capacitors) require a more comples analysis.

 

[flatresponse]

Maybe for preamps you might get away with resistances , for power amps your going to throw away even more energy as heat . Best get a suitable transformer for the job you want to do . Not all tubes draw 300ma , power tubes like EL34 draw a lot more current on the heaters .

Heres one possible option ,
Netztrafo für Röhrenverstärker, z. B. 2x EL34 Ultralinear | eBay

If your only powering a preamp tube or two from this your heaters are going to be overvolted by a large amount .

Thank you so much for all of your input. Maybe I should have specified what I am doing: I rebuild the tube compressor TAB U73b with Varicom. In my prototype version with passive PSU I have 3 torroidal transformers, one for each unit: 2 x U73b with secondary 214V - 0,1A, which is fine for the B+ with a bridge rectifier. 6,2V - 2A, which delivers too high voltage for the 2 x E99F (0,15A each) and 6,3V - 2A too high voltage for E88CC (0,3A) and E80CF (0,33A). With load resistors I came close enough to the real deal with 6,26V and 6,37V. But I want to achieve the same result WITHOUT load resistors and my questio is how to do that ?

 

[flatresponse]

Usually, transformers' power rating is given for a nominal input voltage and a nominal load.
Let's say you order a 6.3Vac xfmr rated at 50VA, it will deliver 6.3V when the mains voltage is excatly the nominal voltage (230v in Europe) and when the load is such that it draws exactly 17.75A.
If the mains voltage is higher or lower (the supplier is permitted a variation of +/-10%), the 6.3v secondary will vary accordingly.
As well, when the current draw is less than nominal, the voltage at the secondary will increase by a factor that depends on the construction of the xfmr. This parameter is sometimes included in the datasheet.
Tube manufacturers are well aware of these issues and tubes are usually quite tolerant to heater voltage variations.
However, the combination of high mains voltage and too big a transformer can be an issue that decreases tube life.
Please note that this is valid for resistive loads, which tube heaters are. Reactive or non-linear loads (rectifiers + capacitors) require a more comples analysis.

Thanks abbey, I hope I got you right: the transformer can be designed to meet the power consumption of the tubes. For an example, to make sure I really understood: let´s say we have two E99F with 6,3V - 0,15A each, I have to order a torrodial with specs 6,3V - 0,3A. Plus a little more, or direct 1:1 ?

 

[flatresponse]

Thanks abbey, I hope I got you right: the transformer can be designed to meet the power consumption of the tubes. For an example, to make sure I really understood: let´s say we have two E99F with 6,3V - 0,15A each, I have to order a torrodial with specs 6,3V - 0,3A. Plus a little more, or direct 1:1 ?

@ abbey: By reading a second time I have to admit I didn´t get you at all. No clue about the figures of your example. If you be so kind to explain the calculation ? Which relation between 6,3V / 50VA / 17,75A ? I´m standing on the tube, as we say in German.

 

[abbey road d enfer]

@ abbey: By reading a second time I have to admit I didn´t get you at all. No clue about the figures of your example. If you be so kind to explain the calculation ? Which relation between 6,3V / 50VA / 17,75A ? I´m standing on the tube, as we say in German.

Sorry for the miscalculation.
P=U.I => 50VA=(6.3V) x (7.9A).

 

[abbey road d enfer]

Thanks abbey, I hope I got you right: the transformer can be designed to meet the power consumption of the tubes. For an example, to make sure I really understood: let´s say we have two E99F with 6,3V - 0,15A each, I have to order a torrodial with specs 6,3V - 0,3A. Plus a little more, or direct 1:1 ?

You have to order the xfmr as capable of 6.3Vac when delivering 0.3A. The winder will then choose a core that allows passing at least 6.3x0.3=1.89VA. But you should tell him to use a larger core, let's say one capable of 5VA or more and use low induction. This will result in less stray field (less hum).
Anyway, to be precise, you need to tell what is exactly the mains voltage, which is some kind of an unknown, because it varies as a function of the time and day and the alignment of planets.

 

[flatresponse]

You have to order the xfmr as capable of 6.3Vac when delivering 0.3A. The winder will then choose a core that allows passing at least 6.3x0.3=1.89VA. But you should tell him to use a larger core, let's say one capable of 5VA or more and use low induction. This will result in less stray field (less hum).
Anyway, to be precise, you need to tell what is exactly the mains voltage, which is some kind of an unknown, because it varies as a function of the time and day and the alignment of planets.

@abbey: Thanks for additional explaining of multiple parameters to watch. I already suspected there would be no easy solution. As your second chapter says, you can only estimate on the basis of 230 V +/- 10%. But I´m not that strict, in your first response you said: >tubes are usually quite tolerant to heater voltage variations<
I will go for toroidal transformers custom made with secondaries for a) 2 x E99F, for b) 2 x E88CC, for c) E88CC plus E80CF - according to tube data.. In this case the manufacturer ("multi-circuit-boards" do toroidal xfrms on demand) should do the maths. But I will pass your advice on larger core and low induction, to be on the safe side.
Thanks again for clearing my view, best
Manfred.