best choice for longest life from preamp cathode bypass caps?

[mxrshiver]

i like to get 10,000 hour caps whenever i can across the board, or at least 5,000 hour, as some of the best brands like Nichicon and Rubycon often have them for as good or better than the best prices for the standard 2,000 hour caps.

pretty much the only available choices for the most typical value of 22uF with lifespans that long, are 50V and above... and i've read quite a few times that caps that aren't loaded to a voltage at least fairly close to their rating, will suffer a reduced lifespan as a result. i usually try to select caps that are rated for anywhere between 125%-200% of the voltage they'll actually see across them because of that.

but with no more than a volt or three on most cathodes in the preamp... even the lowest 6.3V caps wouldn't make that ideal range, and even going up to 16V where there are more options, that puts them at a rating about ten times of the voltage they're likely to see.

assuming all other factors (such as heat and temperature rating) are equal... which would you go with? the 6.3-16V 2,000-3,000 hour cap for a better rating-to-actual voltage ratio? or the 50V 5,000-10,000 hour cap for a better lifespan rating?

[kagliostro]

Oh, I didn't know that caps must be rated at that voltage set

As to have a longlife, if possible, I prefer film capacitors

---

As a rule of thumb use the temperature as a choice, the higher, the better for longlife

Franco

[mxrshiver]

good rules for sure - film caps are always the better choice. i usually shell out the extra cents for them if it's <5uF, and bypass cap values 470n to 5u are usually what i use if given the choice, at least for guitar. but 22uF film caps start at about $5 a pop 😫 and so many vintage amps that badly need a recap at this point, just use 22uF across the board... so the cost starts adding up quick when someone doesn't want the values changed at all.

i don't think the voltage rating thing is something i've seen in an actual article or published work, but i've seen several folks post about it and at least one lengthy discussion somewhere... from what i remember, it was about undervoltage less than 50% of the rating, leading to uneven forming of the dialectric layer, which causes uneven performance and more ESR... this would be exacerbated by infrequent use and long dormant periods, and perhaps by inrush current as well.

but yes, heat is the biggest factor - unless it's going in a doghouse or another spot where i know there's no heat, i will take a cap rated for 3,000 hours at 105 degrees, over a cap rated for 10,000 hours at 85 degrees, any day!

[Merlin]

i've read quite a few times that caps that aren't loaded to a voltage at least fairly close to their rating, will suffer a reduced lifespan as a result.

This is not really an issue for modern caps, or for audio amplifiers. If you think about it, audio amplifiers spend most of their lifetime 'off', so the caps spend most of their lifetime with zero volts across them, 100% below their rating! If the caps are going to lose oxide it will be mainly for that reason, which you can't do anything about. If the old advice matters at all (I'm not conviced that it does with modern electros), it matters in equipment that is permanently powered. You can use 50V cathode bypass caps with confidence. If you want the longest life, buy 105 degrees instead of the standard 85 degrees.

[kagliostro]

Panasonic: FM, FR, FC
Nichicon: PW, HE, HD, HM, HN
Rubycon: ZL, ZLH, ZLG

For that brands the indicated acronyms specify they are for Switching PS

And that is indication of superior quality

As to improve the life of standard e-caps may be of some help to add a small polyester cap in parallel

Franco

[mxrshiver]

thanks Merlin! i thought that the dialectric layer may become uneven over long periods of time if unused, which iirc is the reason why old caps may need re-forming in the first place... but that if the cap isn't loaded close to its rated voltage, that re-forming may not happen evenly either, and lead to spots that have more or less thick of a dialectric layer. totally makes sense that this would be an issue that's sorted out by now in our manufacturing, though!

Franco, fantastic information and an excellent suggestion, thank you! am i correct that a value as small as 1nF would be suitable for this? also, would a ceramic cap be suitable for this?

[acheld]

am i correct that a value as small as 1nF would be suitable for this?
would a ceramic cap be suitable for this?

Interesting question.  Before you asked, I was not aware that ceramic caps were available as high in sizes appropriate for cathode bypass caps. 

I could not find a Class I cap; all (looking at Mouser) were Class II, ie their capacitance will drift +/- 20% with temperature (the exact amount varies with the ceramic material used).

To answer your questions:  1nF would be minuscule as a bypass cap.  Check out  https://www.ampbooks.com/mobile/amplifier-calculators/ and plug your numbers in to see what happens.

While I've never seen a ceramic cap used as a bypass cap in a guitar amp, unless I'm missing something it should work.  Unless you can find a Class I cap, it will drift with increasing temperature -- but so what! -- it's a guitar amp.

Personally, when I've used good quality electrolytic caps in my builds, I have not yet had one fail over 15 years.   Of course they WILL fail someday . . .  just not yet.

[kagliostro]

About ceramic caps I usually don't see it as cathode bypass capacitors

Electrolitic caps - Bipolar Electrolitic caps and film poly capacitors are those you see there

Franco

[mxrshiver]

sorry, i wasn't clear - for the 1nF cap, i was referring to Franco's idea to put a smaller polyester cap in parallel with an existing electrolytic, in order to extend its lifespan, not for use on its own. i do think i've heard of this method before in power supply filter caps... and i was wondering what capacitance would be appropriate for this, and if a ceramic would work as well as a polyester.

1nF would indeed be miniscule as its own bypass cap... though sometimes i'll use them in that 1-10nF range, if i just want a presence range boost from the stage! some of the 5150 designs have a 3.3nF bypass cap on the last stage before the phase inverter, and i think that's the mentality - there's been so much presence range rolloff due to all the 1M series resistors, that they boost it back a bit there.

[kagliostro]

I don't think a ceramica is a good choice for that use (may be it is but I don't know)

Give a look to the schematic of the VOX AC30CC2

Franco

[acheld]

OK, honestly I think your best bet is to find a 10,000 hour electrolytic cap rated for 105 deg and call it a day.   There's no real magic in this other than the specs.   

Adding another cap in parallel just adds complexity and another point of failure. 

Think about it -- what is the worst that can happen when a bypass cap fails?  Yeah, the amp will sound different --  but nothing else will be injured.