Wattage rating for series resistors in PSU?

[trans4funks1]

I have built a bench test circuit for a PSU for tube mic DIY and have a question about the wattage ratings on the series resistors in the RC circuits.

The design is based on a 10mA current draw at 120vDC with the idea that it can support the choice of a wide variety of tubes. I am having trouble finding tube data for anode current at 120vDC.  I have looked at various tubes (12ax7, 12at7, 12ay7, Ef86, Ef800) and it seems like the current draw can vary from 0.5mA thru to 8mA depending on which tube you select, so I'm designing for 10mA with the idea that the load, represented by R5 (12kOhm) will actually be a combination of a tube and some particular value of resistor wired in parallel.

I am writing to request that someone please review my thoughts about rating the series resistors, R1 through R4, for wattage and correct me if I am making a mistake.

Each of the series resistors, R1, R2, R3 and R4 drops the voltage approximately 10-15 volts.

15volts x 0.01Amps = .15 watts

R5 drops 120volts. R5 will be sized to compliment the load from the tube, so each particular choice will experience a different current through it but the worse case is

120volts x 0.01Amps = 1.2 watts

I will appreciate correction or confirmation regarding the wattage rating of the resistors.

 

[dmp]

power = V*I = V^2/R = I^2*R  (from Ohm's law)

Keep in mind that if the tube draws LESS current than your 10 mA estimate, the voltages will be higher.
You might want to calculate the power requirement for no load, and size the resistors accordingly.
For instance, worst case R5 could see 165v, which would be more like 2.2 Watts dissipation (165^2/12k)

Also, it is pretty common to have a 100k plate resistor in mic circuits, whichever tube is used ((12ax7, 12at7, 12ay7, Ef86, Ef800), in which case, the maximum current draw is 120/100k = 1.2 mA
This is the current draw if the entire voltage is dropped across the plate resistor, when typically only 1/2 or so is.
So your 10 mA estimate might be high unless you will be powering non-standard mic circuits.

 

[trans4funks1]

Keep in mind that if the tube draws LESS current than your 10 mA estimate, the voltages will be higher.

Yes, That is why I plan to use a resistor in parallel with the tube to maintain a 10mA draw. That way I can swap tubes and hopefully replace a single resistor in the R5 position and leave the series resistors alone.

your 10 mA estimate might be high unless you will be powering non-standard mic circuits.

Here is some anode current info I collected from various tube data sheets, some of the values are estimated via proration because the voltages on the tube data sheet are not specifically listed for 120vDc:

6072:  3mA
12AX7: 1mA
12AT7: 5mA
EF86:  3mA
E80F:  2mA
EF800: 8mA
5840:  8mA

Maybe I'm not appreciating how the plate resistor will effect this?

I'm trying to establish a base design for my DIY PSU and hope to make a few PSUs to use a with a few yet to be made DIY mics where each mic uses a different tube.



In any event, I am hoping to learn if the approach I took to arrive at the wattage rating of R1, R2, R3, and R4 is correct.

 

[trans4funks1]

Hi dmp,

I realize that I was so focused on my original question that I did not recognized the significance of your comment about the plate resistor... Thanks for pointing this out. I'll make time to give this some thought.

Thank You.

 

[PRR]

You "never" run 10mA in a small-signal triode.

> draw can vary from 0.5mA thru to 8mA

Those are factory "show-off" numbers. Yes, 12AT7 can be used profitably at 5mA... in a TV tuner, with coil loading. In audio with resistance loading, more like 1mA-2mA. With a B+ as low as 120V, probably even less.

You are the boss. The tube sucks what YOU tell it to suck. You pick appropriate cathode and plate resistors.

Any tube you would use is just as happy (or happier) at 150V as at 120V. Classic designs run 250V. The B+ voltage is *not* critical, it only has to be "smooth".

Increase your filter resistors to maybe 2K each, that's still plenty for any one or two small triodes. Use over-size resistors-- the running power may be tenth-Watt but there is a turn-on surge and huge momentary power. 22ufd is fine.

Study small-signal triode design. No, *steal* values from a Resistance-Coupled Amplifier chart.... the best designers did this.(*) That particular table does not show current demand. Assume the tube is in a fair-fight with its plate resistor. If B+ is 120V, then tube and resistor both have 60V across. If Rp is 100K, then current is 60V/100K= 0.6mA. If B+ were 140V and Rp 47K, then 70V/47K= 1.5mA.

(*) Nearly all Fender guitar preamps and most clones use 12AX7 with Rp=100K Rk=1.5K and about 200K audio load with about 300V of B+. Look at Chart 9 (page 11), second line for 300V. That's where Leo got it from.

 

[dmp]

Any tube you would use is just as happy (or happier) at 150V as at 120V. Classic designs run 250V. The B+ voltage is *not* critical, it only has to be "smooth".

Microphones often use 120v to make the capsule bias easy (0v / 60v / 120v gives all three patterns).
Most capsules don't like more than 60v bias.

 

[PRR]

> Microphones often use 120v

Yes; but microphones "never" use 10mA, because 1.2 Watts inside the small case would cook the diaphragm.

 

[trans4funks1]

Thanks to both of you.

I am going to extend my test bench circuit to include an example of the mic circuit so I can plug some tubes in and see what happens.

That should confirm to me first hand that my estimate of current draw is as off base as you suggest.



Back to my original question. Discounting the ideas about adding a safety margin and the idea that there is a power surge that occurs in the first 1/2 second; Did I use the right reasoning to size the wattage ratings of the resistors?

Voltage drop across resistor and resistance value determines amps via ohms law. Then amps time volts to arrive at Watts?

I've done a lot mocking and or copying of designs and this time I'm trying to learn the real basics via calculations and prototype bench testing. I am enjoying building and testing as a learning exercise. So far my test bench work has matched the calcs I have made very closely. Like I say, I'm trying to move beyond copying old examples that are known to work, but I'm also not trying to reinvent the wheel. I'm just trying to learn the stuff with some first hand work and need help with simple questions now and then.

Thank You.

 

[PRR]

> the idea that there is a power surge

Power off. All caps charged to zero voltage.

Power on. Assume that C1 rises to 160V "instantly" (because it is connected to city power). At that hypothetical instant, C2 is still at zero. There is 160V across R1. (160^2)/1K is 25 Watts of power.

It is not that bad. There's resistance in the transformer so C1 can't charge instantly. Also C2 will charge (simplified) as 22uFd+1K or 22 milliSeconds to 101 Volts, and then dissipation is 3.5 Watts; after 44mS to 140V and only 0.43 Watts. Can R1 stand a dozen Watts for a dozen milliSeconds? With old fat-mud resistors, it usually did. But we keep making resistors smaller. Thin-film has less thermal mass, though they do seem to work fine.

On breadboard, it cooks, you replace it. But for paying studio work it is wise to pick that resistor large.

 

[DaveP]

Trans4funk1,

With the kind of experimental work you are doing, Steve Bench's old site is a great place to start.
http://www.jacmusic.com/techcorner/SBENCH-PAGES/sbench101/

Here is his comparison of the noise in various tubes for example:-
http://www.jacmusic.com/techcorner/SBENCH-PAGES/sbench102/RAT_Noise/noise4.txt

Steve is no longer with us but his work lives on in other sites, it was thought to be that important.

Also Ian at Ruffrecords on this site has done some interesting work on noise in mic pre tubes that is worth investigating.

best
DaveP

 

[trans4funks1]

Edited to remove dumb question.

It took me while to realize that dmp had already provided the answer:

power = V*I = V^2/R = I^2*R  (from Ohm's law)

Thank You!

 

[trans4funks1]

Hi Dave P,
Thanks for the links.

I am not specifically trying to do experimental work. I'm trying to learn more about traditional tube mic designs in a way that leads to a better understanding of electronics and mic design.

I've built many small tube guitar amps by essentially copying existing designs.

I've had a long standing fascination with how power supplies effect the sound of the factory gear I have purchased through the years.

At some point I realized that tube mics were fairly simple circuits and it is possible to DIY one and so I've begun the process of trying some DIY mics by working up a DIY power supply rather than simply mimicking one. I expect to end up with something remarkably similar to one that I may have mimicked but I hope to understand what's going on a bit better by taking this approach.

Thank You!