DC Filament Supply questions

[jonasnoble]

Good day everyone, could you help me with a simple DC Filament Supply? I've searched the forums and don't see exactly what I need.

This is my first attempt at running DC heaters, and I thought it would be really simple. I have 12.6V into a bridge rectifier, filtered, the voltage divider feed to a 7806 or a Zener, and I cannot get either one to work. The Zener method will give me the 6.3VDC, but as soon as I load it with anything (even a light load) the supply drops to ~2V.

Here are some sims I'm working with. If anybody could direct me to a schematic or thread, I'd be very grateful. (I'm sure this has been covered.) Or, if you can tell me how to provide enough juice that my supply doesn't get dragged down so hard...

 

[jonasnoble]

It looks like a 7806 can take anywhere from 9V to 25V, so maybe I need to try it without the voltage divider. I thought I just didn't want to make it work that hard.

 

[Khron]

7806 - what's with the 27k / 68k divider, BEFORE the 7806? That series 27k will drop 1V for every 37mA 37uA (microamps) you attempt to draw through it.

Zener - you seem to have a 2.7k / 6.8k divider before the zener. Same as above, just dropping 1V for every 3.7mA 370uA / 0.37mA draw.

While it's true that zeners need a series resistor "upstream" of them, in order to work as intended, there are online calculators for that.

How about retry both simulations, but REMOVE those 27k/68k divider ahead of the 7806, and the 6.8k R3 from the zener schematic?

http://www.calculatoredge.com/electronics/zener.htm
Edit: (in addition to mental math corrections) In the case of the zener regulator, you'll also wanna keep an eye on the power the zener itself needs to dissipate. It's "customary" (and trivial) to whip up a discrete regulator with a fat NPN pass-transistor, that is more readily able to dissipate the needed watts (they come in packages that can more easily be bolted to a heatsink, than a little cylindrical diode).

https://sound-au.com/articles/regulators-2.htm#s2

 

[jonasnoble]

I thought I needed to bring the 7806 supply down to a manageable level, so it didn't have to throw off so much energy in heat. (I was playing with different values, high/low impedance to see if it would have any affect.

Just ran the sim again without the voltage divider. I'm much closer, but still getting loaded down when I put the "tube" in it.

 

[jonasnoble]

7806 - what's with the 27k / 68k divider, BEFORE the 7806? That series 27k will drop 1V for every 37mA you attempt to draw through it.

Nice, I didn't really think of it that way. I originally had very low values in there, but I see why you're telling me that.

 

[jonasnoble]

http://www.calculatoredge.com/electronics/zener.htm

This is great, thank you.

 

[Khron]

Just ran the sim again without the voltage divider. I'm much closer, but still getting loaded down when I put the "tube" in it.

6v into 10 ohms is 600mA - Ohm's law is your friend... Is this meant for a single power tube that needs 600mA for the heater?

I thought I needed to bring the 7806 supply down to a manageable level, so it didn't have to throw off so much energy in heat

That can't really be avoided with linear circuits (ie. not switchmode / PWM). You could either use a transformer with a lower secondary voltage (say, 9V AC), and/or adding (only) a series resistor ahead of the 7806. Once again, Ohm's law - figure out what current you need, decide how much voltage you want to drop across that series resistor (keeping in mind you'll want to end up with at least 8v at the input of the 7806), and calculate how much power that resistor will need to dissipate (and pick one rated for 1.5-2x that power).

 

[ccaudle]

Is this meant for a single power tube that needs 600mA for the heater?

Or if there are two tubes with 6V heaters, you can run them in series on 12V and not have to waste so much power on a 6V regulator.

 

[Khron]

Or if there are two tubes with 6V heaters, you can run them in series on 12V and not have to waste so much power on a 6V regulator.

... Assuming neither is running on a fixed bias derived from its own heater power - that might need some rewiring / resistor value alterations.

 

[jonasnoble]

6v into 10 ohms is 600mA - Ohm's law is your friend... Is this meant for a single power tube that needs 600mA for the heater?

Yes, I have heard of Ohm's law. It's a single 6N1P right now, I may add another stage later, but I need it to work like this first.

 

[Khron]

That 7806 is gonna need quite the chunk of aluminium for heatsinking...

Any reason you're simulating the transformer as well, though?

 

[jonasnoble]

That 7806 is gonna need quite the chunk of aluminium for heatsinking...

Any reason you're simulating the transformer as well, though?

No real reason, other than I like to experiment and see if I can make something work. In this case, it worked. I have dual 6.3V secondary windings, and I thought I could either parallel or stack them by changing the ratio of T1

There's obviously a way to get what I need here, it just might not be that efficient or elegant. Also, I don't always trust these simulations, so I'll take the voltage divider out tonight and see what happens.

 

[Khron]

Paralleling them should indeed be ok - assuming the transformer is rated for the current / power you need.

At the "nominal" current, a 6.3v AC should net you almost 9V DC after rectification & smoothing, which should be just fine for a 7806. That way it'll "only" need to burn off a bit under 2W on its own, which is quite reasonable.

 

[scott2000]

assuming the transformer is rated for the current / power you need.

+1 ..Can't forget the losses from the bridge rectifier

 

[jonasnoble]

I got something to work. In practice, with heater windings, I never see the output I expect. Maybe the secondaries just don't have enough juice. My sims weren't working because the documentation said I have to have a series resistance with the primary. As soon as I removed the 1ohm resistor I got it to work, loaded.

Here's my simplified model, I'll try it out tonight. I'll probably go 7805 with two diodes, or just a grounded 7806. I think because it's DC, I can probably get away with lower than 6.3V.

 

[jonasnoble]

I just want to say "thanks" for all the help. You guys are kings.

 

[Khron]

I might be misreading that signal generator, but you're feeding in 6.3v @ 1khz into the transformer - is that supposed to be a 1:1 ratio transformer, or....?

 

[jonasnoble]

I might be misreading that signal generator, but you're feeding in 6.3v @ 1khz into the transformer - is that supposed to be a 1:1 ratio transformer, or....?

Holy smoke, how did I miss that? No, in this case I'm stepping up to 12.6V, but it should be 60Hz, obviously not 1k.

 

[jonasnoble]

It didn't change much. Probably slowed the charge, but I didn't get granular enough to see it.

 

[Khron]

I rarely bother simming power supplies so "completely". Or when i actually have reason to, i don't bother with the transformer, just slap a sine source @ 50/60Hz at the voltage i'm interested in (in LTspice it's the peak value, while transformer voltage ratings are RMS).

I just want to say "thanks" for all the help. You guys are kings.

The devil, as is so often the case, is in the details...