Noob-"ish" Capacitor Questions

[Myriad Society]

Hello everyone, I've been thumbing through and reading this forum for a few years now and have always managed to find the answers to my questions without having to ask. I'm not exactly new to electronics although I never started messing around with audio stuff until I found this forum (thanks for all of the great info!) so I still have a lot to learn about how different subtle changes to a circuit affect the way your finished box sounds.

Anyway, assuming I'm using 50V Panasonic ECQ-V caps wherever possible and that all other caps whether they be electro, tantalum or whatever are going to be at least 16V-25V (giving myself a 20% safety margin) depending on the source voltage I'm using for the pedal - is there any logic or rule(s) of thumb to voltage rating selection on capacitors such as just enough voltage to give you the safety margin on power filters and a minimum of 63V-100V on metal films in the signal path. Does the higher voltage rating in the signal give any audible benefit?

Recently, I purchased a JFET distortion (this is definitely a distortion and not a fuzz!) from a rather well known boutique pedal maker who I won't name. I'll say it's not Mr. Keeley though as his name pops up on this forum quite often. Love the sound of the pedal and so, of course, I had to open it up to see what was going on under the hood. Without tracing out the layout which I don't really want to do, it looked like a relatively standard JFET distortion with a few exceptions but I noticed that about 3/4 of the resistors were carbon comp rather than metal film and there was a full mix of different types of caps in there from monolithic ceramic, ceramic disk, metal film - greenies & brownies, a couple different types of electros, a tantalum...most of which were high tolerance and I don't get why when most of the values could have been covered in the commonly accepted ECQ-V series. Is this a sum of all the parts thing to arrive at the best possible sound or a building to minimize production costs thing which I get by the way or what? If it is the the first, what are the rules of thumb for what caps to put where in a given circuit? If there are none, it had to be an endless amount of testing for the guy to come up with his design.

Sorry for the long post. Any help here is greatly appreciated or if someone could point me toward the thread I missed in this forum as I'm sure it's there somewhere.

[jpm83]

For the caps it sounds like he is using what he got rather than what gives the the tonal nirvana to him. When I buy electro caps I usually go for at least twice of the supply voltage. Lately I've been searching for a bit more "Mojo" caps for my builds simply because they look nicer and nothing beats the psycho acoustics.

Janne

[alanlan]

I think you've really got to ask what is the difference between all these different types of cap.  Why are they chosen for a given application?

Could be whatever is in the bits box, but in most cases there's are bit more to it than that. 

Consider:

- capacitance per unit volume - for a given type, a higher voltage rating means it must be able to store more charge and have a higher breakdown voltage so it will be bigger.
- price - needs no explanation
- leakage - high impedance circuits may be upset by an electrolytic versus a ceramic or other low leakage cap.
- tolerance - you may want repeatability between units in a filter for example
- stability - you don't want an oscillator to drift over time and temperature
- linearity - minimum distortion
- inductance - would be important for higher frequency applications - not usually a problem for guitar stuff
etc.

People quite often like using what they have used before without necessarily asking all these questions.  In a lot of cases, it doesn't really matter that much.  If someone is really careful, they may do a lot of research and buy the most cost effective or just the best quality for the job.

And then there is the so called "mojo" factor which may be part explained by logic, maybe part inexplicable or based on experience.

To really "know" if there is any difference you have to experiment yourself.

[Myriad Society]

Thanks a lot for the responses. Everything you are both saying makes a lot of sense. I also have the tendancy to use what I've used before in that if I need a metal film cap it's likely to be a Panasonic. It's also easier on the wallet and the parts bins to keep a few select common values of the same brand of capacitor around that I know I'll use all the time rather than have the same 12 values repeated in 8 different brands of film caps, for instance.

If I had my guess after hearing what you guys are saying I'd think that he's likely selecting components by cost or using stuff he has laying around. Possibly there was a bit of logic to him with it also in that he's used monolithic ceramic caps in a build before for a certain duty in the circuit and it worked out for him so now, as was stated, we stick to what we've previously used even though a metal film would also work. Perhaps even when he was prototyping and tweaking the circuit originally he just threw in spare parts that fit the values from the toolbox rather than blowing fresh stuff from the parts bins to get that random mix and he hit that mojo so he just orders all those components to build them with now.

Thanks for the input...i just wanted to make sure that there wasn't some important rule or subtlety of capacitor selection that I was missing in all of this.

[PerroGrande]

Don't forget footprint...

PCB designers are usually out to minimize space, waste, etc.  Some decisions may be driven based on the size of the unit relative to certain design goals.  For example, it might be nice to use a 220uF, 50V capacitor in a given circuit.  However, if the dimensions of the available units do not work with the design (and the right size is too expensive), the designers may have to compromise and go with a 100uF, 16V capacitor (or 220, 10V).

[R.G.]

...hmmm...

I know where to buy IC lead frames, on rolls of 1000 each. How about a tricky guy making up tiny PCBs that fit the leadframe with SMD components, then clamping the leaframe into a mold and pouring in black epoxy. Once set, he'd sand off the top of the epoxy where it was poured - that being the same face that would normally carry the manufacturer's name, part number, etc.

It's a little bit of a pain to do in quantity, but for up to 100 mojo pedals, a guy with a friend with a milling machine could make up the necessary molds and fixtures in an afternoon. PCBs can be made to be easily diced up in a 12" shear at home, especially if you get the 0.012" to 0.015" stuff that multilayer boards are made from.

The nice thing about this approach is that it is difficult even for the experienced reverse engineer-er to figure out.

Make your own mystery ICs for fun and profit.   

[PerroGrande]

R.G. -- you've given me some ideas!  (I'm getting a lot better at the smt stuff -- ironically, the biggest obstacle to small size are the electrolytic capacitors.)

I imagine you can get the frames in a number of different IC footprints.  PDIP-8 might be a little small for the homebrew SMT'er to put a lot on -- but something wider would be very interesting. A PDIP-40 size (or equivalent width with fewer pins) would be very nifty to work with.  One PCB that mounts the stomp switch, the jacks, the power jacks, and one "mystery" IC + enclosure = pedal...  Wow.

Oh -- one problem...  Can you get Mojo in an SMT package?