[Coco] WTB Orchestra 90 cartridge

[Zippster]

This is a pretty interesting topic, and I’ve learned quite a bit researching it.
It would seem that while ceramics are on the bad end of the scale (in general)
when it comes to harmonic distortion of audio, they are commonly used in the
audio path, especially where size is a factor.  

Not all ceramics are equal, and it’s pretty clear a lot of the bad reputation comes
from larger, through hole components, especially using the less stable dielectrics.

I saw some information in manufacturer application notes indicating SMT packages
using X7R dielectric to be similar to tantalum SMT packages in THD of audio, around .1%
in their comparison.  Which doesn’t sound too bad, but didn’t compare to polyester 
at .001% or less.  Electrolytic came in ~.05% or so.

Some ceramics seem to be more suitable for the audio path, Class I-C0G-NPO type,
with their more stable capacitance.  I would guess closer to electrolytic in distortion.

So relatively speaking, film is best, then electrolytic, then tantalum and ceramic.
Another interesting question though is, how much THD is too much?  One manufacturer page
added some perspective by noting that the worst ceramic in their test, a through-hole using X5R
dielectric with THD of .2% at -3dB, still paled in comparison to the THD caused by audio DACs
and codecs.

Though a lot is made of capacitor induced distortion in audio, I’m guessing it’s somewhat
overstated.  At least in a lot of situations.

Anyway, since we aren’t especially space constrained on this board, I’ve picked out some non-polar
SMT electrolytic caps for the final PCB version, since there’s certainly no reason not to go that
route.  It’ll make an interesting comparison when I have boards built both ways in hand later too.

Learning is fun…  :)

- Ed



> On Aug 1, 2016, at 4:13 PM, Simon Jonassen <simon at roust-it.dk> wrote:
> 
> You *can* use a nonpolar mylar cap with good results... (I know from playing with various audio preamp circuits over the years (mostly riaa))
> 
> /Simon :-)
> 
> 
> -----Oprindelig meddelelse-----
> Fra: Coco [mailto:coco-bounces at maltedmedia.com] På vegne af Zippster
> Sendt: 1. august 2016 20:05
> Til: CoCoList for Color Computer Enthusiasts
> Emne: Re: [Coco] WTB Orchestra 90 cartridge
> 
> It’s certainly standard operating procedure.  Not a big deal to change it.
> 
> - Ed
> 
>> On Aug 1, 2016, at 12:59 PM, Simon Jonassen <simon at roust-it.dk> wrote:
>> 
>> I would tend to agree that any audio OUTPUT that needs to be AC coupled (eg dc filtered out) should use an electrolytic cap 
>> 
>> just my 2c
>> 
>> /Simon :-)
>> 
>> 
>> -----Oprindelig meddelelse-----
>> Fra: Coco [mailto:coco-bounces at maltedmedia.com] På vegne af Zippster
>> Sendt: 1. august 2016 15:02
>> Til: CoCoList for Color Computer Enthusiasts
>> Emne: Re: [Coco] WTB Orchestra 90 cartridge
>> 
>> Things can get a little weird when you get into audio land with opinions on capacitors, but…
>> 
>> I think ceramic will be fine here considering the value (.1uF), and use (ac coupling).
>> If we were using a much larger value, I’d probably agree and go electrolytic or film.
>> 
>> In any case, I’ll put a scope on it and take a look at the signals.  
>> As well as the highly subjective test of human ears.   :)
>> 
>> - Ed
>> 
>> 
>>> On Aug 1, 2016, at 2:44 AM, Jayeson Lee-Steere <cocolistemail at titaniumstudios.com> wrote:
>>> 
>>> On audio, there are a lot of options about which caps are appropriate for
>>> audio, but all references I have seen say that ceramics are not an
>>> appropriate choice in the audio path. Given their strong piezoelectric
>>> properties, it is not clear to me what are appropriate uses for the
>>> recently available, high value MLCC ceramic caps. Too bad, because they are
>>> smaller and less costly than tantalum.
>>> 
>>> This reference on audio caps seems fairly balanced:
>>> http://sound.westhost.com/articles/capacitors.htm
>>> 
>>> Film capacitors are the thing to use where practical. It would seem some
>>> film types are better than others, but for this use, polyester would be
>>> more than sufficient. That is probably what the small value caps are in the
>>> original Orch 90. Next down the line are electrolytics, which are
>>> apparently fine for audio coupling if an appropriately large value is used.
>>> 
>>> Jayeson
>>> 
>> 
>> 
>> -- 
>> Coco mailing list
>> Coco at maltedmedia.com
>> https://pairlist5.pair.net/mailman/listinfo/coco
>> 
>> 
>> -- 
>> Coco mailing list
>> Coco at maltedmedia.com
>> https://pairlist5.pair.net/mailman/listinfo/coco
> 
> 
> -- 
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco
> 
> 
> -- 
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco