Resistor power rating query.

[Shawnobi]

I'm replacing the Carbon comp resistors near the Power caps on my 70 watt Pro Reverb.
There is a 4.7K across the + rail of the two 20MFD/500V caps. Fender lists all resistors as 1/2W unless specified. This particular one isn't specified as being different, but looks larger than the other 1/2w.
Does anybody reckon it's a higher wattage of will a half watt be fine? It's got around 400v on that line.
Pic attached. Resistor in question is the larger one on the right.



Schematic:
fender_ultralinear_pro_rev.pdf

[Leo_Gnardo]

I'm replacing the Carbon comp resistors near the Power caps on my 70 watt Pro Reverb.
There is a 4.7K across the + rail of the two 20MFD/500V caps. Fender lists all resistors as 1/2W unless specified. This particular one isn't specified as being different, but looks larger than the other 1/2w.
Does anybody reckon it's a higher wattage of will a half watt be fine? It's got around 400v on that line.

I generally use 1 or 2 watt R's in this function. Doesn't much matter what voltage the R is at, but each end of that resistor, that will help you calculate the operating power. Some extra power handling should be added for that moment when the cap bank is charging. Carbon comps work well here because they tend to survive that short-lived overload without turning into fuses.

[Shawnobi]

Thanks, interesting point on the carbon comp. I was planning to change to carbon film.

[pdf64]

With this application, by far the most significant stress is at start up.
The power and voltage ratings need to accommodate this.
Resistors with a low 'limiting element voltage' rating may not have long life.
See Aiken Technical Q&A
However, I found out the hard way.

[Shawnobi]

With this application, by far the most significant stress is at start up.
The power and voltage ratings need to accommodate this.
Resistors with a low 'limiting element voltage' rating may not have long life.
See Aiken Technical Q&A
However, I found out the hard way.

Seems Aitken advises 5W 1% 750V metal oxide flameproof resistors for this application. Interesting he also advises using those for screen grid. I had recently heard that the school of thought that it's good for a screen grid to act like a fuse in the event of tube failure to protect the Tranny.

[pdf64]

I think that Aiken may have 'taken as read' that the B+ is suitably fused, in which case robust screen grid resistors are appropriate?

Where that isn't the case, eg Fenders, my perception is that the use of resistors that are likely to fuse when subjected to 'reasonably foreseeable' fault current (and in fusing prevent further collateral damage) may be beneficial.

As this thread is in the 'repair' rather than 'design' sub-forum, in the light of the above I think it would be a retrograde to uprate the power handling of screen grid resistors of eg a Fender, unless suitable additional fusing was also implemented.
Whereas swapping 1 watt carbon comp for 1 watt metal oxide (with a flame retardant coating) may be seen to be a good idea, as the carbon comps can 'light up' when they fuse and char insulation etc.

EDIT - Some great stuff regarding resistor power rating / use as 'ghetto' fuses, in a recent TAG thread http://ampgarage.com/forum/viewtopic.php?t=27296 including contributions from RG and Ian 'Gingertube'.