Capacitors.. which ones for which applications?

[belwar]

Hey All,

I've been teaching myself electronics for the last several months now. I started at the very begining not really knowing what reistors, capacitors, or inductors did. Well I now have that basic understanding and im working towards designing my first very basic cuircuit.

One thing I dont understand is WHY different types of capacitors are used at different time. For example, I understand that in any audio circuit you should avoid ceramic because of its harshness.

Can someone explain for me why AND what are the circumstances you use a particular type of cap in? or more importantly, when to avoid using a particular type. I am a circuit I am trying to scematic out which uses Ceramics, Tantalum, and Metallisized Film Box Capacitors.

Any help is appriciated for this noob.

Bel.

[tedmich]

there are very many sites which describe basic capacitor types and their use, perhaps someone can post their favorites, as I can do later today.

In general all the currently made capacitor types exist because there is a market for them. In some cases these markets are huge: many billion ceramic, film, tantalum and electrolytics are sold every year for a multitude of uses. At this volume they are usually quite reasonably priced, while there are also much smaller amounts of hemp/silk dielectric, teflon/silver film and beeswax/paper/oil capacitors made at MUCH higher prices because they are deemed to be superior for specific audio applications. Its really a continuum of uses. Ceramic caps are very good to filter the AC lines feeding your power supply and you can find them in very high quality audio designs. Tantalums seem to work very well as cathode bypass caps in Mesa Boogies tube amps and most every power supply filter requires the large high voltage capacitance only electrolytic caps can economically give. There is good electronic design (use what works and costs lowest), and then good audio electronic design (use what you think sounds best and may cost more) and then mysterious audio design voodoo (use what people feel is best at any cost). As an example a 4.7uF/200v cap can cost $0.20 as an electrolytic, $2.00 as a polypropylene film audio cap, or even $2000 as a teflon film cast silver foil audiophile cap. Its all what you want and in some audio applications it may be impossible to honestly tell these choices apart.

General audio path biases on capacitors are as follows:

ceramic<tantalum=electrolytic< metallized film< film/foil

but smart people use what works for them!

[Glasgow]

What exactly are the properties that make certain types of capacitor better for audio? Leakage? ESR? Inductance? Temperature handling? Warm fuzzies?

Thanks,
Brian

[Steve Conner]

99% of the information on capacitors for audio is, as you say, warm fuzzies.

I do believe that you should avoid ceramic caps in an audio signal path, though. They can be non-linear (capacitance changes with voltage) and microphonic.

Non-polar electrolytics in crossovers are another of my pet hates. But beyond those two applications, I'd say anything goes.

Tantalum capacitors are a kind of electrolytic, except the electrolyte is solid instead of goop.

Film/foil capacitors are overkill for virtually any audio application. They're only really needed for RF and pulsed power applications. In a metallized film cap, high RF current can overheat and blow out the thin metal film where it meets the end contacts, but film/foil caps can take much more because the foil is thicker. I've used them in Tesla coil projects where metallized film caps really did explode after a few minutes.

You can use them in audio if you want, but the physical size makes things difficult, they're considerably bigger than a metallized film part of the same capacitance and voltage.

[belwar]

For for audio applications would those little box capacitors .. Metalized poly whatchamacallit work well? Im slowly trying to design my own pre-amp. These little box caps fit in tight spaces, and are fairly affordable. Well most of them. There seems to be alot in WIMA brand

[Steve Conner]

Well, I'm sure other people have other opinions, but I think those Wima box capacitors would work just fine. Wima capacitors use a three letter code:

MKT is metallized polyester (Mylar), these tend to be the best value for money, and similar to most other plastic film caps you'll find in current production guitar amps.

MKP is metallized polypropylene, a premium dielectric with better performance, but more bulky and expensive. You might use it in a really swanky hi-fi amp or a speaker crossover, but I'm not convinced anyone would hear the difference.

FKP is polypropylene film/foil, super heavy duty as I described above.

These are the three main food groups of plastic film caps, generally. You also get polycarbonate, polystyrene and Teflon, but they're rarer.

Other plastic film capacitors worth checking out are orange drops and green boogers. Vishay/Sprague hold the "Orange Drop" trademark and use it on their type 715P capacitors, but you get generic polywhatever caps dipped in orange epoxy from a few manufacturers, like BC Components.

The ones I think of as green boogers are cheap Chinese mylar film things that you might come across on Ebay.

[belwar]

Steve, Thank you very much for your very helpful responses!

So in designing an audio circuit, I couldnt go wrong using the FKP or MKP caps. Tedmich above mentioned a general view that Film/Foil is "viewed" as superior. Assuming size isnt a major factor, in a good quality bass pre-amp, I probably couldnt go wrong just using Polypropylene, PIO, or Film/foil?

I'm just starting at this existing pre-amp and trying to figure out why several different types of caps were used.. from cheap ceramics to some nice box caps. Makes no sense to me.

If anyone has good links talking about "where" to put certain types of caps, i'd love to see them. I've already learned so much from this thread!

[hasserl]

Just keep in mind that most of what you'll find on the net, or elsewhere, regarding "best" caps is highly subjective, and IMO is mostly people beliefs due to the power of suggestion with no supporting objective evidence. You'll also find plenty of anecdotal stories, and psuedo-scientific analysis or reasoning.

but I'm not convinced anyone would hear the difference.

Hooray Steve, coming from an ex-audiophile this is a big statement. I share your skepticism.

[Enzo]

Amen to that.


DO yourself a favor and avoid listening to arguments that include always and never when it comes to cap selection. The real world is not as hard core as that. We may all have our preferences, but you know, Fender has been making amps with ceramic caps in the tone stack and on the bright for 50 years. They sound just fine.

You need a variety of caps in an amp circuit. A tube amp will need some filter caps for the B+ and the bias supplies. Maybe 40uf/500v and 100uf/100v for those. Then you will need some coupling caps, typically 600v and common values might be .047uf, .022uf, .01uf and others. And the occasional small cap like the treble cap and brightness caps in the tone circuits, maybe a stability cap somewhere.

Now as an exercize, look up some 40uf 500v caps or maybe 450v are easier to find in catalogs. DO you find any that are not electrolytic? Darn few if any. And if you do, what do they cost and how large are they physically? Plug that data into your design decisions.

Now look at coupling caps. How many types can you find of for example .022uf 600v? Not many electrolytics I bet. Certainly ceramic ones and film. FInd any silver micas?

Look for the little caps like 250pf, 120pf. No lytics. Tough to find caps in that range that are not ceramic or mica.

SO there are size constraints, there are simple availability constraints, and there are cost constraints.

And construction. Those rectangular caps are fine, and they mount on a circuit board easily and can snuggle up next to each other in a row and save space. They are not nearly as friendly to mount on an eyelet board. And on the other hand, I am not so sure they use a pick and place machine to mount parts on an eyelet board like than do on a circuit board. SO in your shop it may not be an issue, but a commercial amp maker has to consider the construction methods.

Now tone is important, but the stuff I just discussed didn;t consider tone. You have to look at the whole project, not just part of it.

[Steve Conner]

Thanks for the great explanation, Enzo. These constraints of availability, cost and size are how amp designers choose capacitors. I don't think Leo Fender would have worried for an instant about the "tone" of a 470pF silver mica vs. a 470pF ceramic. He'd have used the ceramic because it was cheaper.

Nowadays a lot of boutique amp builders seem to think that different 470pF capacitors would sound different, and they use the brand of capacitor as a selling point. But I think this is just hype.

I've worked with power electronic circuits where you absolutely have to have the right make and model of capacitor: if you use the wrong one it catches fire. Enzo's example of trying to use a "greenie" in the tank circuit of a radio transmitter is the kind of thing I mean. It might be the same value as on the schematic, but a cheap Mylar cap isn't going to take 25 amps of RF. A good schematic should contain notes to explain where cheap generic components aren't good enough, for example "C4: 100pF 5kV vacuum capacitor, Jennings part no. 12345"

But I'm pretty sure tube audio isn't as critical.

PS: That's also a good point, the box capacitors tend to have little stubby legs for PCB mounting, which makes them a real pain to install on eyelet boards. You often have to solder on extra lengths of wire.

[Tom Phillips]

Bel

In choosing a capacitor type there are many more considerations than the capacitance and voltage rating. Glasgow touched on this in his response. In addition to the capacitance and working voltage rating the full set of specifications for a capacitor will list the Equivalent Series Resistance (ESR), Self (series) Inductance, Temperature Coefficient, Ripple Current and more. There will also be a family of curves provided showing how certain performance parameter are affected by other things such as temperature. Then there is construction and quality control which become more important when you are building something that must work for a long time in a location that is difficult or impossible to service such as on the bottom of the ocean or in space.

Fortunately, you don’t need to understand all this to choose your caps for a guitar amp project. It’s actually much more difficult to sort out the hype from the truth. The recommendations you are getting in the other replies here will allow you to pick your caps. For those who want to read more I have attached a field guide to capacitors paper written by Harry Bissell.

Now a couple of comments from the soap box.
There is a lot of hype out there. Following is my opinion about a few examples. You could also just consider these a few of my pet peeves.
1)“I changed all the caps in my amp from type X to type Y and the sound improved from dull to better sound heaven”
Comment: May be true. However, when you change everything at once you may have included the replacement of one bad cap that made most or all of the difference. Sometimes values get changed inadvertently. Furthermore, all the capacitance values actually change a little due to the manufacturing tolerance. Bottom line is that changing everything at one time is not exactly a controlled experiment and, although it may work great in one instance, it may not be the greatest approach for someone else.
There are exceptions: If you are restoring a 50 year old Fender with those tubular yellow astron caps you might as well replace all of them because most, if not all, will have excessive DC leakage. Whatever the materials and construction used in those caps was, it does not stand the test of time.
2) “I use silvered mica caps for the treble cap in my tone stack because they are better for high frequencies”
Comment: The silver mica may sound better to most people in that application. However, that’s mainly because of what’s being replaced (The ceramic disk) A few people like the ceramic better in a blind test. Remember that you are often replacing a very old 20% tolerance ceramic with a new 5% silver mica so the capacitance and the type changed at the same time. Note that silver micas are indeed better in “high frequency circuits” but the high frequency comment originally referred to radio circuits. The treble frequencies in our guitar amps are not really “high” in general electronic terms. A radio designer will also want the tighter tolerance and stability of the silvered mica even though ceramic could be used and in fact are used in cheap radios.
3)“I use orange drops”
OK. Orange drops are good caps but that’s not a very specific statement. It’s kind of like saying “I prefer two door cars” That still leaves a lot of undisclosed information about the car. There are over 20 types of “Orange Drop” caps available. Around five of these types are commonly shown in parts catalogs and they are made with different materials and different construction techniques. Therefore, if you believe what people say about the sound difference between a polyester cap and a polystyrene cap, then you need to know which orange drop you are talking about.

Anyway… This post may not directly answer your question but I thought it would be useful food for thought. Enjoy your learning experience. It will be interesting to see how the thread has progressed while I have been writing this.

Regards,
Tom

[Enzo]

Amen Tom.

One of my favorite claims is a guy who has an older amp and he:
Pulls out all the old film caps and installs new "orange drops" in their place.
Rebuilds the tone stack with diffferent value parts for "better tone."
Puts in a new premium output transformer for huge money.
Does the Joe Schmoe gain mod.
Puts all new filter caps in, raising the value of many.
And who knows what else.

And then talks of just how great those orange drops sound in his amp.

[Tele-Cat]

I'm a Mallory 150 user when it comes to coupling caps. No orange drops for me.

[Glasgow]

Tele-Cat,

Tell us what you like better about the Mallory 150s.

[Tele-Cat]

Originally Posted by Glasgow

Tele-Cat,

Tell us what you like better about the Mallory 150s.

I build Tweed clones when I build, and I think they sound better. ODs sound hi-fi to me. I have A/B'd otherwise identical amps. Maybe I was hallucinating, but since M150s remain available, I'll keep using them.

[Tom Phillips]

Which type Orange Drop are you using for comparison?

[Tele-Cat]

Originally Posted by Tom Phillips

Which type Orange Drop are you using for comparison?

716.

[Tom Phillips]

No discussion of Orange Drops is complete unless we know what type is being evaluated. Now I understand how you got that opinion of Orange Drops.
The type PS Orange drop, IMHO, sounds good in a guitar amp.

BTY, I do like the 150s. I also like the 150 form factor better for hand wired amp work.

Cheers,
Tom

[Tele-Cat]

Originally Posted by Tom Phillips

No discussion of Orange Drops is complete unless we know what type is being evaluated. Now I understand how you got that opinion of Orange Drops.
The type PS Orange drop, IMHO, sounds good in a guitar amp.

BTY, I do like the 150s. I also like the 150 form factor better for hand wired amp work.

Cheers,
Tom

Yeah, I definitely like config of the 150s better than the OD config.

Are the PSs polystyrene?

[Glasgow]

Thanks!

[fezz parka]

For guitar amps, especially tweed circuits, polyester is the shizzle.

[JoeM]

And what does shizzle sound like?

[hasserl]

Originally Posted by JoeM

And what does shizzle sound like?

Is that that frying bacon sound?

[Tom Phillips]

Originally Posted by Tele-Cat

Yeah, I definitely like config of the 150s better than the OD config.

Are the PSs polystyrene?

PS type rated 200 thru 1000 VDC are Polyester film.
The higher voltage PS type are different dielectric and different construction but those higher voltage one are not applicable to guitar amp circuits anyway. It is interesting that there are so many different types and variations within the type. There are over 20 type versions of "Orange Drops"

Tom

[Enzo]

Just the thing when working on your Marshall JCM.

[melvin]

Try this: take a .01 or .02 uF cap of each type and switch them out in a Fender amp where the preamp is coupled to the PI. I wonder if you can hear any differences. Keep in mind the tolerances may skew the test.

I tried this once with .02's and found that I favored this cap i found called a "TINY CHIEF", anybody know what these things are?
Anyway, it may have been the tolerances but I heard differences with each.

I have also subscribed to the idea that 400V coupling caps sound different than 600V. Am I going crazy? Maybe I've been handling lead solder too often.

[Enzo]

Where? In what applications? Under what circumstances?

[melvin]

I'm talking about playing a guitar through the amp and listening. The 400/600 in the same position I mentioned... is that what you mean? did I forget to mention that Kush was also preferred to Northern Lights during the test?

[Enzo]

Oh. Well I certainly would carefully measure the actual capacitance of each as well as its ESR.

[melvin]

Yes, you're right, I think it has little bearing until I do so.

[tubeswell]

FWIW due to being on the arse-end hemisphere of the tube-building world, I use what ever I can get my hands on that I think will be okay.

I go for mallory 150s or silver mica for the pf and low nf range (up to say 5nF), but if I can't find them, then ceramic disc go in as temporary installations until I can obtain the other types.

then film caps for the 1nF to 1uF range, - sometimes mallory 150s, sometimes orange drops, or brown turds, but I found a source of 22nF mustard caps that I have used in about 5 amps so far. I like these - but maybe that's mojo factor I dunno

the bigger ones 1uF and above are electrolytic for physical size considerations 2CW

[gbono]

Note that the melting point of polypropylene (capacitors) is 160 C. I believe polystyrene is over 260 C. Not that hard to damage these while soldering them into the application.

As previously stated many other non-audio applications created the market for the current technology used to make capacitors - low ESR for power supply filter, low dielectric absorption for sample and hold circuits, multilayer ceramics for RF, etc. "Subjectavist" high end audio applications for capacitors only started to show up in the mid-1980's - but unfortunately they still persist.

[Tele-Cat]

If you like the brown turds, that's all that matters.

[guitician]

As far as e-caps go, use the highest voltage rating you can find for your application and if they are also of the higher temp versions, like 105c, they will last longer.

[fezz parka]

Originally Posted by JoeM

And what does shizzle sound like?

It depends on your ears.