1/2w vs 1w resistors 5f1 build

[British1011]

Hey all. 

I'm diving into my first amp build.  Ive done a lot of repairs on my own amps and have built a lot of circuits but I'm excited to learn all I can about amps now and starting with an easy one the 5f1. 

The first question I have is..  I've seen on some posts here that some builders are using 1 watt resistors in place of the 1/2 watt that the schematic calls for... I'm wondering if theirs a reason for this?  I've heard that a lot of the newer production resistors have very small leads.. is that mainly the reason?  And is their a sound difference?

I already know the carbon comp vs metal film debate... I plan on buying both since they are so cheap and seeing what I like best. 

I know this is probably nit picking but I like to learn everything I can before I start a new project.  Which brings me to my next question. 

Do you have any recommendations on books or youtube videos that I can learn all there is to know about amps and how they work?  I really want to know everything I can possibly learn about amps. 

Thanks guys!

[labb]

http://www.aikenamps.com/index.php/resistor-types-does-it-matter

[j_bruce]

from the link above:

"Since the characteristics of thermal noise have a Gaussian probability density function, and the noise of the two separate sources is uncorrelated white noise, the total noise power is equal to the sum of the individual noise powers. If you model the individual resistors as noise generators, the output noise voltage will be equal to the square root of the sum of the squares of the individual noise sources."

Gotta love a Gaussian probability distribution!!

Jerry

[shooter]

Gotta love a Gaussian probability distribution!!

no you don't!, ya just gotta pass the test so you can get a good job 

[labb]

All that being said, I use 1 watt resistors in lieu of the 1/2 watt. Just what I do. Not sure I could justify doing it if I had to.

[echuta13]

My logic on using 1 watt over 1/2 watt is more of an effort to insure reliability over the long term (a little overkill is a good thing). 
More opinion here...  I prefer metal film if I can find them.  I use 1W/metal film for the plates and anywhere they may be in the signal chain if I can. 
 

[sluckey]

I like 1/2W MF because I can poke four leads into the top of a turret...

[shooter]

any recommendations on books or youtube video

start here, see what comes of it, btw, I only use MF 1w for most things, I pay Doug for extra turrets 
http://www.aikenamps.com/index.php/white-papers
http://www.valvewizard.co.uk/

[PRR]

You can't burn-up even a 1/4W in grid or cathode service.

If a preamp tube goes dead-short, a 100k plate resistor on 300V supply gets 0.9 Watts.

Toasted carbon-comp is bad mojo, always over-size for reliability.

[j_bruce]

Gotta love a Gaussian probability distribution!!

no you don't!, ya just gotta pass the test so you can get a good job 

Us professors of mathematics do! 

[shooter]

 
keep proving the math, I'll keep it in spec 

[British1011]

Thanks for all the info.  Seems like 1w is just the safer bet so that's probably what I will go with. Thanks again.

[d95err]

For manufacturers, every cent spared on the cost of a component amounts to a fortune.

For repairs and DIY builds, the price of small components like resistors is almost negligible, so we might as well go for the bigger tolerance.

[AmberB]

I have heard that one of the main reasons for resistance value drift with the old 1/2 watt carbon comp plate resistors is the heat that's caused from pushing the preamp tubes while playing loud.  I don't know how much truth there is to it, but it does make some sense.  The amps I've built from old PA amp chassis, and the couple of full amp builds that I've done all have 1 amp carbon film preamp tube plate resistors, just in case.
I've been using carbon film resistors because they're supposed to be more stable than carbon comp resistors, and I used to hear that the metal film resistors were "noisier".  I don't know if that's true or not, just what people used to say 20 years ago...

[echuta13]

I can't speak to the quality of metal films 20 years ago, but today they are considered lower noise compared to carbon comp & film. 
I know some people like to color their amps character by judicious placement of a carbon comp in the right place (phase inverter, or last gain stage perhaps). 
I've always felt that the components should let the tubes add the color and stay out of the way otherwise.  Keep the noise floor as low as possible in the early stages at the least.

[pdf64]

Here’s good info on carbon comp mojo http://www.geofex.com/Article_Folders/carbon_comp/carboncomp.htm

From that, the lower the CC resistor’s dissipation rating, the more pronounced its resistance voltage coefficient.
So what’s the goal? If it’s CC mojo, then use as low a power rating as is compatible with the circuit (eg 1/2W) CC type, as a 1W will have less of it. Whereas if low noise and stability are the goal, then may as well use a lower noise, more stable resistor type than a 1W CC (eg CF, MF).

I have heard that one of the main reasons for resistance value drift with the old 1/2 watt carbon comp plate resistors is the heat that's caused from pushing the preamp tubes while playing loud.  I don't know how much truth there is to it, but it does make some sense.  ... I used to hear that the metal film resistors were "noisier".  I don't know if that's true or not, just what people used to say 20 years ago...

Bear in mind that pushing preamp circuits with signal will have a negligible effect on the heat dissipation of their resistors.

Metal film resistors have always been a lower noise, more stable choice than typical carbon types. Reports contrary to that are probably either in error, or faulty MF resistors have somehow been used.

[HotBluePlates]

I have heard that one of the main reasons for resistance value drift with the old 1/2 watt carbon comp plate resistors is the heat that's caused from pushing the preamp tubes while playing loud.  I don't know how much truth there is to it, but it does make some sense.  ...

Carbon comp value drifts because carbon comp drifts in value.   

Since the source of the above info was someone making a claim, I suppose someone making an opposite claim doesn't carry much weight.

So look at a schematic for any of the big Fender amps: Super Reverb, Twin Reverb, Showman.  The largest signal in the amp that heats a resistor is the signal driving the output tube grids.  And we know from the schematic that bias voltage for these amps is in the neighborhood of -52v, which also means max clean output power happens when the output tube grids are driven with a 52v peak signal.

     52v peak / 1.414 = 36.8v RMS

The grid reference resistors are 220kΩ running from 6L6 grid to bias supply.  Power = Voltage² / Resistance = 36.8v² / 220kΩ = ~6.2 milliwatts.  For a 500mW rated resistor, this is nothing.

That driving signal for the grids had to come from the phase inverter plate, and the lower resistor value will have the greater dissipation as shown by the formula above.  So how about that 82kΩ resistor?  Plug in the numbers, and you get 16.5mW of resistor-heat due to signal.

The situation earlier in the preamp is even less heat, because the signals are so much smaller.

___________________________________

The exception to the above are amps that have a resistor, rather than a choke, feeding the screen filter cap in the power supply.

The 5E3 Deluxe would be an example, and has a 5kΩ resistor in this spot.  If you look at the Class AB 300v condition in this data sheet (the closest approach to tweed Deluxe values), screen current rises from 5mA to 13.5mA idle to full-tilt.  Add 4mA for the preamp tubes, and we have 9mA rising to 17.5mA.

Idle: 9mA² x 5kΩ = 0.41w
Max: 17.5mA² x 5kΩ = 1.53w

A 2w resistor here will get warm during sustained loud playing, but a 3w resistor is right where it should be for long-term operation.

[66Strat]

Something to keep in mind is that a resistor's power rating is temperature dependent. For example a Vishay RN65 1/2 watt resistor is rated at 70 degrees C. At 125 degrees C, the RN65 de-rates to 1/4 watt. Resistors should be de-rated to allow for heat, particularly resistors mounted to tube sockets such screen grid resistors.

https://www.mouser.com/datasheet/2/427/cmfmil-1762972.pdf

[Raybob]

Something to keep in mind is that a resistor's power rating is temperature dependent. For example a Vishay RN65 1/2 watt resistor is rated at 70 degrees C.

70C = 158 F.  I can't imagine any resistor in any amp ever getting over that heat level, much less 125C which is higher than boiling water.

[66Strat]

Something to keep in mind is that a resistor's power rating is temperature dependent. For example a Vishay RN65 1/2 watt resistor is rated at 70 degrees C.

70C = 158 F.  I can't imagine any resistor in any amp ever getting over that heat level, much less 125C which is higher than boiling water.

That was just an example. The Vishay has a relatively high tolerance for heat. Other resistors are are not as forgiving. A rule of thumb that my Dad taught me a long time ago was to calculate the power dissipated by the resistor in circuit and then to double that amount to allow for heat.

https://www.youtube.com/watch?v=sY4TJHTI9Xg

[HotBluePlates]

... A rule of thumb that my Dad taught me a long time ago was to calculate the power dissipated by the resistor in circuit and then to double that amount to allow for heat. ...

Oh, for sure!  If anyone wasn't clear on that, a resistor should always have a power rating of at least double the expected dissipation for cool operation & long life.  Hence:

...
Max: 17.5mA² x 5kΩ = 1.53w

A 2w resistor here will get warm during sustained loud playing, but a 3w resistor is right where it should be for long-term operation.

[66Strat]

The other piece of my post that was overlooked concerned the placement of resistors. A resistor mounted directly to the tube socket pins (heated by conduction) will get a lot hotter than the ambient air temperature (heated by convection) inside a tube amp chassis. If the internal components of a properly functioning output tube range between 130 C and 150 C, how hot does a component get, that is directly attached to a socket pin? Ever seen a Fender with a crispy screen grid resistor?

In my original post, I referenced the Vishay RN65 1/2 watt resistors that de-rate to 1/4 watt at 125 C. In comparison, Ohmite Little Demon carbon comps rated 1/2 watt at 70 C de-rate to 0 at 130 C.
https://www.ohmite.com/assets/docs/res_od_of_oa.pdf

So, what's the point? My point concerns the cost of being wrong. If I am wrong in my assumption that a component can approach 125 C, all that it has cost me is a couple of dollars in component cost. Conversely if my assumption is correct, I have avoided the time and cost of replacing the failed component, as well as the sunk cost of the failed component.

Something to keep in mind is that a resistor's power rating is temperature dependent. For example a Vishay RN65 1/2 watt resistor is rated at 70 degrees C. At 125 degrees C, the RN65 de-rates to 1/4 watt. Resistors should be de-rated to allow for heat, particularly resistors mounted to tube sockets such screen grid resistors.

https://www.mouser.com/datasheet/2/427/cmfmil-1762972.pdf

[PRR]

> assumption that a component can approach 125 C

Spit on it. Does it sizzle?

Tube bulbs, yes (and they may shatter).

Seriously power wasting resistors, yes.

Otherwise most parts in most tube amps run below the sizzle point (well under 125C).

[sluckey]

Ever seen a Fender with a crispy screen grid resistor?

I've seen several. And they were all toasted because of excessive screen current, not because of tube heat.

Fender also put the 1500Ω/.5W grid stoppers between pins 1 and 5. And they were mounted at the bottom of the socket, very close to the tube. I've never seen one of those fail. If heat was frying the 470Ω/1W resistors, you would think the little 1500Ωs would be cooked well done.

[66Strat]

Ever seen a Fender with a crispy screen grid resistor?

I've seen several. And they were all toasted because of excessive screen current, not because of tube heat.

Fender also put the 1500Ω/.5W grid stoppers between pins 1 and 5. And they were mounted at the bottom of the socket, very close to the tube. I've never seen one of those fail. If heat was frying the 470Ω/1W resistors, you would think the little 1500Ωs would be cooked well done.

How can you differentiate heat versus current related resistor failure? Current capacity is heat related. A control grid, under normal operating conditions, will have no grid current, hence no power dissipated in a grid stopper resistor.

> assumption that a component can approach 125 C

Spit on it. Does it sizzle?

Tube bulbs, yes (and they may shatter).

Seriously power wasting resistors, yes.

Otherwise most parts in most tube amps run below the sizzle point (well under 125C).

Will spit sizzle on a tube socket pin?

[sluckey]

How can you differentiate heat versus current related resistor failure? Current capacity is heat related. A control grid, under normal operating conditions, will have no grid current, hence no power dissipated in a grid stopper resistor.

I'm not trying to. You're the one that suggested tube heat fried those 1 watt screen resistors. In order for tube heat to fry those resistors which are about 1" away from the source of heat, the heat must first rise past the 1/2W resistors which are mounted much closer to the source of heat (only about 1/4"). My point is that rising heat would surely fry the smaller resistors before it reaches the larger 1 watters.

With that in mind, I'll stick with my theory that the fried 470Ω screen resistors were cooked with excessive current, not radiant heat from the tube. I think your crispy screen resistor is not a good example for the point you were trying to make.

[66Strat]

How can you differentiate heat versus current related resistor failure? Current capacity is heat related. A control grid, under normal operating conditions, will have no grid current, hence no power dissipated in a grid stopper resistor.

I'm not trying to. You're the one that suggested tube heat fried those 1 watt screen resistors. In order for tube heat to fry those resistors which are about 1" away from the source of heat, the heat must first rise past the 1/2W resistors which are mounted much closer to the source of heat (only about 1/4"). My point is that rising heat would surely fry the smaller resistors before it reaches the larger 1 watters.

With that in mind, I'll stick with my theory that the fried 470Ω screen resistors were cooked with excessive current, not radiant heat from the tube. I think your crispy screen resistor is not a good example for the point you were trying to make.

I never said that heat was the sole reason that the resistors failed. But I certainly think heat was a contributing factor. The only thing that I said in my posts were that resistors should be sufficiently de-rated to allow for heat. Also heat is transmitted directly via conduction through the tube pins as well indirectly via convection from the tube base.

I need to do a cap job on my Fender Concert. This would be a good opportunity to take some temperature measurements of the tube pins and the grid resistors to actually see how hot these components get. My DVM has a temperature probe that is good to 250 C. I've used it to test the thermostat in my oven.

[66Strat]

The historic rule of thumb was to double of the calculated in-circuit resistor power dissipation to allow for heat. The Little Demon Resistors that I previously linked de-rate linearly from 100% at 70 degrees to 0% at 130 C. Applying the half-power rule of thumb, the resistors would de-rate to 1/4 watt at 100 C (212 F). My guess is that the composition of the Little Devils is similar to or the same as carbon comps that were in use when the rule of thumb first came about. So was this rule of thumb developed based upon a cognizant decision using 100 C as a reference?