the 'best' componenet types to use

[Narcosynthesis]

what are considered the 'best' types of components to use?

you get quite a few different types of caps for example... electrolytic, metal film, ceramic, etc... so what are the 'best' to use

same for resisters and any other components

David

[cd]

Caps: depends on the application.  Ideally, never use electrolytics, since they dry out and you'll have to replace them eventually (10-20 years).  That's impractical though, so polys work.  Basically anything except ceramic.

Resistors: use whatever's cheapest, in a 9V effect it doesn't matter.

[mikeb]

'Best' -> cheapest and most readily available you can get that has the performance and characteristics you require in the application. What application are you talking about?

Mike

[Peter Snowberg]

That's a fairly grey question.

For resistors, metal films are the lowest noise but carbon films are just fine too.

For caps I like silver mica best in small values followed by NPO ceramics and then X7R ceramics. Try to avoid Z5U ceramics.

For larger values than that (say from 0.001uF to 0.47uF), I like film caps. These come in tons of varieties, but really two sub-classes depending on internal construction. Some have high ESL while others have low ESL. In solid state distortions I almost think that high ESL caps may be superior while in tube amps I tend toward low ESL units. Maybe it’s just 6 of one; half a dozen of the other. One thing is for sure to my ears…. Cheap film caps can sound really great!

For larger values still the realistic choices turn into aluminum electrolytic or dry tantalum. Tantalums will last a very long time but they distort the signal quite a bit while Al electros have a very finite life span and less distortion. Tantalums are also very picky about polarity. Reversing them will kill them. The "clearest" electrolytics may be the non-polarized ones.

In the end.... just about anything will work and the differences are fairly small in most cases. There are however some places where the difference is striking so let your ears decide.

[Ge_Whiz]

Once again I read this '10 - 20 year lifetime' myth for electrolytics. Most of my electrolytic stocks are ten years old, many are surplus devices from the early 1970s, and I've never come across a dried-out one unused or used at low voltage. Do you stateside guys have 'use-by' dates on yours, or suffin'?

I agree that there may be other reasons for avoiding electrolytics, but lifetime ain't one of them. And neither is cost. Sound? Not to my tin ears.

[R.G.]

Once again I read this '10 - 20 year lifetime' myth for electrolytics. Most of my electrolytic stocks are ten years old, many are surplus devices from the early 1970s, and I've never come across a dried-out one unused or used at low voltage. Do you stateside guys have 'use-by' dates on yours, or suffin'?

There is no accounting for luck.

The fact is, makers of electros used to design for five year shelf life at common warehouse conditions. The object was to get 90+ percent to survive that amount of time to minimize returns and off-spec, dissatisfaction, etc.  No amount of engineering will make over 90% of anything survive over five years and all of them fail in the sixth year.  

Of course they survive longer than the anticipated shelf life. But how much time do *you* want to waste debugging a circuit where you have a bad one that's slipped in. I realize that this is a rhetorical question to you. By your comment, you're willing to spend any amount of debugging time on it, because you think, as backed up by your sampling, that you will never hit a bad one.

Here's another way to look at it: most electros are used in circuit conditions that are quite forgiving compared to the manufacturers' specifictions. The value specified is often deliberately large, and tolerant of drift and high internal resistance. Chances are, a number of your ten year old stock won't make the manufacturers spec for leakage, ESR, capacitance tolerance, and so on. But your usual effect circuit doesn't make those demands, so you don't notice it and think 10 year old stock is fine.

What happens when you get a **great** sounding effect with surplus reused caps over 10 years old, and then use it for ten years? If a year or two of life is OK, great.

I have worked on a lot of dead effects - particularly EH - where the only things needed to make the things work is to replace all the electros and remelt all the solder joints. I have fixed several transistor amps that were given away for a song when all that they needed was an electro cap replacement.

The point is this: think about the uses for your equipment and buy parts accordingly. If you play in your bedroom, build with salvaged, old, dusty parts that are known and designed for a limited life, count yourself lucky, and accept the repairs when and if they become necessary. If you work on stage and rely on your equipment, or more importantly build them for pro musicians who depend on them for a livelihood, use the most reliable parts you can.

No one ever said that an electro doesn't live a long time. But nothing lasts forever, particularly things that are known to have a wear-out mechanism built in.

[Peter Snowberg]

I'll just say that in over 20 years of fixing televisions, by far the most common failure I've seen is a dried up electrolytic. The second most common failure I've seen is in regulators where there are lots of electrolytics around. The third most common failure I've seen is horizontal output transistors.

There's no getting around that the manufactures specify a shelf life for these components.

One thing to keep in mind about electrolytics is that the less you use them, the faster they fail. If you keep a charge on them the plates attain maximum "health". If you don't use them, the dielectric seems to degrade over time, finally resulting in failure. There are plenty of sites with information on "re-forming" electros.

[Lonestarjohnny]

What sound's best to your ear's, I try all different kind's of cap's, sometime's you find just that right combination that make's ya go Shwing !
and I agree with Peter, the quiter a resistor is the better I like it !    
Johnny

[saros141]

Hey R.G., what happens when an electrolytic goes bad - does it slowly lose its capacitance or does it crap out all of a sudden?  If I check it on my capacitance meter and it reads above spec, is this a good indicator it should be fine?  This applies to checking recent-stock electros too...  10% failure in only 5 years is enough to warrant 5 seconds on a meter for ANY electro, I'd think.

[Ge_Whiz]

Naturally, old electros go on the meter before use, but I've never had a failed one. And I've fired up amps after twenty years of non-use, expecting electros to blow or show signs of failure, but they don't. Yes, I have also bought s/h amps where electros are already blown, so of course I check everything first. Got any 1980s hi-fi or radio gear in your house that still works fine? Take a look inside, count the electrolytics.

Okay, so all my pedals will crap out before I'm 60. But by that time, the British Government will have banned rock 'n' roll on health and safety grounds.

Like you say, ALL components have a design lifetime. Here's a tip - if any of you want to avoid problems, NEVER NEVER use germanium transistors - or valves - for anything...  

[Peter Snowberg]

Like you say, ALL components have a design lifetime. Here's a tip - if any of you want to avoid problems, NEVER NEVER use germanium transistors - or valves - for anything...  

I wouldn't say all components have a design lifetime. All components have a MTBF rating, but the only ones I can think of with a shelf-life are batteries, electrolytic caps, germanium transistors , and a few tubes that use radioactive materials like Co-60 and Ni-63 with relativly short half-lives.

[R.G.]

what happens when an electrolytic goes bad - does it slowly lose its capacitance or does it crap out all of a sudden?

That depends on the conditions the cap is used in. For signal capc, the capacitance usually drops modestly - 20 to 50% - and the Equivalent Series Resistance (ESR) rises until at some point the circuit performance is notably different than it once was. It may be hard to recognize a "failure" since the process is gradual. That's what I meant in my earlier post about tolerant circuits. It may develop leakage and throw bias off, but high ESR is the usual thing, especially in low voltage circuits.

In power circuits, it may fail in high ESR, but if it gets too hot repeatedly, it can either fail suddenly - and smokily - at a power on, or it may get hot, leak out its electrolyte, and get open.

For effects, especially old effects, it's like a breath of new life to replace the old electros. They start to sound like they did when they were new.

If I check it on my capacitance meter and it reads above spec, is this a good indicator it should be fine?

Not really. Electros are designed with a huge tolerance compared to other components. This is because the electrochemical forming process is less controllable than other processes, and because many electro uses are tolerant of high tolerances. +/-10% is getting more common, but power supply caps used to be +80%/-20%.

What you really want is an ESR meter. These measure the internal resistance, which is the best indicator of whether it's still acting like a cap or not. There are some DIY designs for these, as well as some kits.

This applies to checking recent-stock electros too... 10% failure in only 5 years is enough to warrant 5 seconds on a meter for ANY electro, I'd think.

Naturally, old electros go on the meter before use, but I've never had a failed one. And I've fired up amps after twenty years of non-use, expecting electros to blow or show signs of failure, but they don't. Yes, I have also bought s/h amps where electros are already blown, so of course I check everything first. Got any 1980s hi-fi or radio gear in your house that still works fine? Take a look inside, count the electrolytics.

It's like I said, there's no accounting for luck.

I've fired up amps slowly on a variac so the inrush surge wouldn't get anything, and watched as the power supply caps slowly failed and then started smoking, too. I've also counseled many users of old wah pedals when they found they were acting like a volume pedal: it's the 10uF electrolytic cap in the inductor circuit. Replace that and they work fine.
It is truly a testament to the cap maker's art that electrolytic caps work as long as they do but, MBA's being MBA's, they would claim far longer life in their tech literature if they could get away with it.

If you always have good luck with electros, go for it. But feel lucky, and don't expect capacitor immortality when the makers warn you of a shelf life and a wear out mechanism.

[saros141]

"What you really want is an ESR meter"

Don't these contain electrolytics too?   AAAAAARRRRRGGGHHHHJ!  I think I'm going to build my next effect with motor run caps!

Just kidding, thanks as always for the great info.  I do want an ESR meter, and an inductance meter too.  And an SPL meter.   :wink: