please critique my regulated tube heater circuit

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
done and done CJ, thank you.  I'll chock my new little LDO 6V regulator circuit up to experimenting.  I plugged in some batteries and had fun tonight...  ::).  Long term solutions will be optimal.  Next task, calculating voltage drop, resistance, and resistor power rating, after rectifying.  I am going to use 1N540 type diodes, which will have the added bonus of looking hearty  ;D in the 1940's tube, low-gain/low-freq friendly pre.

Thanks again everyone for your help, I have a solid plan now.


 
Here is simple circuit #2.
It will have to be tweaked on test.
given the data, I thiiiink the analysis is correct.  ::)

I asked my wife to solve for X, she's the best!!  ;D

I know this is such FUNdamentals stuff...

I'm waiting for some resistors.
 
Perhaps more a smoulder than a fire  ;)

Using PSUD2 can help indicate where you would likely end up.  Assuming a 6.3Vrms supply, and 1N4007, then the output across the cap is more likely circa 6.9Vrms.  Although the AC winding may be 6.3Vrms, with a peak of 8.9V, that peak gets reduced by the two diodes conducting at that time (their voltage drop will be more than 0.7V due to the approx 1.7A peak charging current), and the capacitor rms voltage will be lower due to the ripple.
 
> I asked my wife to solve for X

This isn't advanced math.

You "know" the load is 6.3V and 0.6A.

You assume this must be dropped from 8.5V (yet to be verified).

We need to drop something over 2V. If you can manage a decimal, 2.2V.

Let's go dumb. Try 1 Ohm.

0.6A in 1 Ohm drops 0.6V. Not enough.

Try 2 Ohms. 0.6A*2r= 1.2V. Not enough.

Try 3 Ohms. 0.6A*3r= 1.8V. Not enough.

Try 4 Ohms. 0.6A*4r= 2.4V. Too much. But just barely.

Since the "8.5V" is a guess, I would start with 4 Ohms and see what comes out.

Yes, you could also see that 2.2V drop needed at 0.6A is 3.67 Ohms. (which is quicker than your voltage-divider approach.) However the "2.2V" comes from "8.5V" which is a rough guess (my money is on 7-and-some V, even tr's 6.9V). So don't spend all day carving-out a 3.67 Ohm resistor, slap 4 Ohms in there and then try again as needed.

Observation: since the load is Known to be 0.6A, you can prove the Power is proportional to the voltage dropped. But the voltage is not known yet. But if you get a handful of 1 Ohm resistors, each one will dissipate 0.6A^2= 0.36W. So you could need up to four 1r 1W units, or maybe only two, even one. Put four 1r parts in series, measure the result, then if low short-out one two or three resistors until the result is right. If "right" is in between 1 and 2, parallel a couple to make a half-Ohm resistor. This will run cooler than the rest but at the price of 1W parts already in hand, it is no big deal. (Re-figure before you go to mass production.)

If you are doing this to reduce AC humm/buzzz around low-level tubes, you need LOTS of filtering. 1V of AC ripple buzz on DC may be more annoying than 6V of pure hum. The point after the ~~3 Ohm resistor begs for a big cap to add a second stage of filtering, giving -12dB/Oct reduction of the higher buzz harmonics.

 
Thanks PRR. 
per you and Sr. Trobbins, expect lower rectified voltage than anticipated, and add another elytic.
Thank you both for the input!


"end thread" *spoken to Computer*
 
thread 'after hours'/epilogue:

I finished this a couple weeks back.

Under load, the heater winding was
7.5VAC, and after the fwbr ~8. & some change (lol) VDC

after 2.5ohms resistor drop = 6.4VDC

I am applying the filter cap equation:

C=  I/Vp-p  x  T

from my text book reading to determine the theoretical adequacy of my capacitor value choice.

maybe a dumb question, is there an agreed upon allowable ripple? is it "if it sounds right" based? My book used 1 Vp-p as a reference to complete the equation.

I have 6k microfarad for the first, then 1k after the drop (space and stock on hand reasons).
How do i calculate the second stage?  Is 1k microfarad useless, or useful? If the answer is too complicated, don't feel the need to go into detail for me, but it is always appreciated.

anyhoo, i added some glue to stick the caps to the board, I feel like a major manufacturer... :p


(the pic is an early stage photo, for clarity)

 
> allowable ripple?

How much ugly-wave do you want running next to your sensitive audio stages?

You lost 6V of 60Hz, but you have gained part-volt of 120Hz, 240Hz, 360Hz, etc., which are far more audible.

I would not pour concrete without a bench-test. I have a feeling that is not enough capacitance.
 
Thank you PRR  :)

I have a handful of 1kuf caps, when I really turned up the speakers and switched from line to mic (aka removing two stages of pad, per my adda box) I could hear the noise floor well.  adding 4kuf on the first stage made a worthwhile improvement (I think. see next sentence  ::)).
As is, using this on the line/twice-padded setting and adding a bit of mixer gain which is how I use the 6n1p tube di for passive p/u bass, that noise floor is *essentially* inaudible.

I think I will make the first  cap 10k, and the second ? and call it a day.

Most importantly, I need to paint a cool screw-hiding cover finishing top for this thing.  I was thinking of calling it the Stoner Box...
and painting two bloodshot eyes on a black background with a green jewell lamp...
I'm wasting entirely too much time on this heater circuit! (j/k of course, it's been fun learning and I appreciate all the input)
 
Back
Top