Power supply question

[E biddy]

In the link below.
http://www.thevintagesound.com/ffg/s..._6g3_schem.gif
The 27k resistor in the power supply shows as 1 watt? They show a voltage drop of 55v across it, so the current should be 2mA. So power dissipation should be I^2R which is .108W. I also figured it using Enzo's (rule of thumb) approximation of 1mA per triode and come up with .432W for a high end.

Is this just over engineering or smart practice?

[R.G.]

Power supply resistors ought to be selected for foreseeable overloads.

Think about this: what if one of the triodes served by that resistor shorts? What happens to the resistor?

[E biddy]

Power supply resistors ought to be selected for foreseeable overloads.

Think about this: what if one of the triodes served by that resistor shorts? What happens to the resistor?

Thanks for the reply. If it shorted, wouldn’t the whole 325v drop across the resistor? Then the resistor would have to dissipate 3.9W which would blow right through the 1 Watt. So should we have a 5W resistor there? Or would we rather have the resistor act as a fuse and stop the current through the tube?

[Old Tele man]

Doubling that "high end" 0.432W is darn close (and 'rounds up) to 1W. It's not uncommon to 2X wattage in power circuits for exactly the reasons R.G. elucidated above.

[cerrem]

Resistor wattage size is usually determined by the temperature rise desired....
In some industries there are derating guidelines for resistor size selection...based on air-flow in LFM to determine the temp rise at highest expected ambient conditions..
Sometimes it is also required to account for transients and overloads for a given time frame...
For example an aircraft arrestment on a carrier by take x amount of seconds... So the resistor must be able to hold on overload to complete the longest arrestment time...

[Dave H]

Thanks for the reply. If it shorted, wouldn’t the whole 325v drop across the resistor? Then the resistor would have to dissipate 3.9W which would blow right through the 1 Watt. So should we have a 5W resistor there? Or would we rather have the resistor act as a fuse and stop the current through the tube?

If a triode shorts the current is limited by its plate load resistor (100k?) so the fault current will be 325/100k = 3.25mA which will only drop 88V across the 27k resistor.

[Mick Bailey]

There's perhaps a different reason for using a 1W resistor; if 1W covered a range of uses it would be more sensible to economize by buying just one rating rather than two or three. Maybe at the time they got a better deal on 1W resistors rather than 1/2W (I've often seen this - I just bought some tantalum caps where the 35v ones were cheaper than 25v). There are plenty of economical or other reasons outside of selecting a component based on theoretical calculation.

Another reason sometimes for a higher wattage is the voltage rating is increased. Maybe not a problem in this particular circuit but sometimes it's necessary to go bigger to get the voltage spec you need.

[cerrem]

I believe sometimes a larger size resistor can have lower Johnson noise...ie lower thermal noise...for the same class and type of resistor..

[Leo_Gnardo]

I believe sometimes a larger size resistor can have lower Johnson noise...ie lower thermal noise...for the same class and type of resistor..

I believe that's true, somewhere in one of my old textbooks there's an equation that shows noise is relative to temperature. All else being equal, a larger resistor will shed more heat - assuming its ambient environment is cooler than the resistor - and a cooler resistor will generate less noise.

[Old Tele man]

Leo_Gnardo, here is a Vishay data sheet on Johnson(thermal) and current noise: http://www.vishaypg.com/docs/49997/49997.pdf

[Leo_Gnardo]

Leo_Gnardo, here is a Vishay data sheet on Johnson(thermal) and current noise: http://www.vishaypg.com/docs/49997/49997.pdf

Ahhh, foiled again, in a good way! Thank you Tele man, from the also old, LG.

[Mick Bailey]

Would thermal noise be a big deal in a low gain amp such as this? The plate resistors would generate more noise than the PSU dropper if this were a consideration and they're 1/2W

[g1]

I don't think it would be a big deal, but one more reason to go with 1W if you already have them (rather than buying 1/2W's), along with the higher voltage reason.
I'm glad to see both the noise and voltage correlations brought up here. I'd seen someone mention the noise thing before, but only once. The voltage issue I was aware of, but up til 10 or 15 yrs. ago I'd never heard of it. We certainly never looked at voltage specs for resistors when ordering at any shops I worked in prior to then and we just took it for granted they would be fine for tube applications.