12 to 5 AC Rectification

[Elvee]

Originally Posted by Mooly

I can't really see a practical problem either with the phase shift and dissipation.

Do you have an idea about the tanδ of real lytics? I mean the ones with messy things inside, like paper soaked in liquid chemicals, porous layers of dielectrics and aluminium, etc, not the polcap from LTspice.
Reactive power is something like 12VAR in this case, I let you make the calculations and guess the result:

[woodrough]

This is truly fascinating. Using the capacitors natural impedance to limit. I love the idea of truly minimized heat / wattage consumption.
I realized that this is a job for electrolytics being a high uF number. Afraid as usual about the negative terminal seeing potential positive current, I would be one to say I must have another electrolytic to block the negative side. But you mentioned that I don't need to in this case because of the lack of DC present? What precisely is going on and should I sweat about it?

[Andrew Eckhardt]

It can be made to last. I doubt it can be made much smaller than a dedicated transformer, though.

Type A series Bi-polar SU | Aluminum Electrolytic Capacitors/Conductive Polymer Hybrid | Passive & Electromechanical | Industrial Devices | Panasonic Global

[Elvee]

Originally Posted by Andrew Eckhardt

It can be made to last. I doubt it can be made much smaller than a dedicated transformer, though.

Type A series Bi-polar SU | Aluminum Electrolytic Capacitors/Conductive Polymer Hybrid | Passive & Electromechanical | Industrial Devices | Panasonic Global

Even with this high-tech, bipolar example, it would be difficult to make it last.

Normally, you would choose the 1000µF/10V for this application.
Since it is obviously inadequate, you would jump two size and go for the 25V.
But it only tolerates 680mA @120Hz. That is still far from 1.9A, and the spec is at 120Hz. For 50Hz, the duty is much harder.
In this series, you would have to parallel 10 100µ/35V just to arrive at the minimal life of 2000H.
With standard quality caps, that would be even harder.

[woodrough]

This is somewhat new news to me as we'll. electrolytics have a amp to Hz rating? If I just oversized the volt rating to something higher, would it make my time more worthwhile? A small transformer couldn't easily fit into my layout... But I can manage capacitors being sprinkled all about.

Any tips on how to properly use capacitors?

[tauro0221]

What about to add another dummy tube and wire both in series. I do not know if you have space for another tube socket.

[Andrew Eckhardt]

Originally Posted by woodrough

But I can manage capacitors being sprinkled all about.

Elvee is right. You would need a very large bank of caps to make this reliable. The cost and size (size even more than cost) would be way above a simple 10 VA transformer. It wouldn't make sense. Try a Hammond 166L5.

[woodrough]

Oh dang! That would be an unreasonable solution. Longevity is important.

So as a worst case scenario, I will end up getting that Hammond 5Vac CT transformer.

But in next best bet.. (Being that capacitors are out of the question):
Step down 7 volts using power resistors? That doesn't seem like a lot of energy waste to me... But this theoretically should work right? If I know that the tube rectifier filiment runs at 1.9A, I can put a resistor that stops down the voltage to be at 5V upon load. No capacitor hassle. Can someone verify if I can make that jump safely? (12V to 5v using only resistors)

Also another idea came to mind:
What If I constructed a half wave rectifier to get half the voltage and try to smooth it out with a capacitor. And then deal with the remainder few volts by resistors.
Or am I wasting time trying to deal with diodes. Any harm to the transformer doing half waves?

[Mooly]

Originally Posted by Elvee

Do you have an idea about the tanδ of real lytics? I mean the ones with messy things inside, like paper soaked in liquid chemicals, porous layers of dielectrics and aluminium, etc, not the polcap from LTspice.
Reactive power is something like 12VAR in this case, I let you make the calculations and guess the result:

I have one thanks... a tan δ meter... me and my tanδ meter
Me and my Tan Theta Meter (Capacitor tester)

I still think this could be made to work. In a hazy blur this was partly one reason why I kept the series trim resistor, to "lighten the work" the caps do.

My concern last night was trying to visualise the reactive load in combination with the inductive transformer winding. The voltage source in LTspice isn't inductive. I concluded that wasn't an issue in practice with the "real" resistive load.

[Mooly]

Resistors are fine, the heat generation is the big problem.

A diode solution... not sure thats a good idea. Have to think on that one