How critical is it to use an aluminum bipolar capacitor rather than a ceramic?

[deadlyshart]

Hi guys, I'm getting into making analog synths rather than effects pedals (I hope it's okay to ask about relevant stuff here!), and the beginners guide I'm following at one point has me using a 1uF aluminum bipolar nonpolarized capacitor. I don't have any of these; I have 1uF electrolytic caps but I assume the nonpolarized part is important here. I also probably have some 1uF ceramic disc caps, which are unpolarized... could I safely use those?

Also, why *do* people use aluminum ones instead? thanks in advance for any advice.

[GibsonGM]

I'd get a 1u nonpolarized electrolytic, shart.  Easy to obtain.

The trouble with ceramics is that "they sound junky".  They drift, they're temp. sensitive, microphonic, among other things.   

Whether or not it would affect the actual sound you hear isn't clear to me from the description - but hey, it calls for nonpolar, and you can get one pretty easy, so I'd just go with it.

[deadlyshart]

Ahhh, yeah... it's one of those things where the shipping for just them costs 3x more than the product itself  I ordered these ones ( https://www.aliexpress.com/item/10pcs-1uF-50V-ELNA-CE-BP-4x7mm-50V1uF-Bipolar-Audio-Capacitor/32548554823.html?spm=2114.13010608.0.0.t7AWLh ) off aliexpress, which doesn't have that problem, but I can also expect to get them in 1-2 months... sigh.

[GibsonGM]

Crap, THAT long?  Can you get Ebay items? 

I would certainly put the ceramic in temporarily just to finish a build! Won't hurt anything, but if it's in a sensitive place it could cause drift,  things like that.  Probably not eve then.   Just replace it when your good stuff comes! 

[Phoenix]

One of the biggest (potential) issues with using ceramic capacitors in an audio path that is not widely acknowledged/realised (certain things like temperature coefficient and microphonics are fairly well known) is their voltage coefficient. The common cheap types like X7R and Y5V have exceedingly high voltage coefficients, that is, their capacitance values plummet as the voltage applied across them increases. They can drop to as little as 10% of their rated value at their full voltage rating! The more expensive types like C0G and NP0 have much better voltage coefficients, but are also not available in the higher values (typically only up to around 100n - though you can get higher values for much more money, much more than a non-polar/bipolar aluminium electrolytic).

Now because your project specifies a non-polar electrolytic, it's probably safe to assume that there won't normally be a DC bias across it, so you might be tempted to think that the voltage coefficient won't matter - but you still have to consider signal amplitude. If it's being used as part of an RC filter with a defined -3dB point, as the signal amplitude increases the half-boost point will dynamically shift (which might be cool in certain applications, but usually isn't what you're aiming for!), similar to, for instance, how an undersized tube output transformer will lose some bass response as it is driven hard into saturation.

The solutions for this are to use ceramic caps with better dielectrics (C0G and NP0), massively over-rated voltage specs (you can find the voltage coefficient on the datasheets to see if the expected signal amplitude will cause their value to drop by a greater than acceptible amount), or abandon their use for a defined RC filter and just accept that a low-pass filter will always allow more treble through under playing conditions than you want, or massively increase their nominal value for high-pass filters (up to x10) to make sure you never lose bass under any conditions.

Now if they're used for local decoupling like the customary 100n you see strapped across the power pins of all sorts of IC's, particularly digital/high speed, this doesn't really matter, you just need a local reservouir of charge with low esr and esl and a self-resonant frequency above the operating frequency of the chip they're decoupling, and even the cheap dielectric types of ceramics do this very well, better than other types of capacitors.

So all different capacitors have their place - the market doesn't just make all these different types for audio and different peoples tastes, audio is really a tiny fraction of todays electronic market. Every choice is a compromise between inherent properties of the different technologies (voltage ratings, capacitance values, value tolerance, lifetime drift, volumetric efficiency, polarity, temperature coefficient, voltage coefficient, esr, esl, self resonant frequency, ripple current/self heating, operating temperature, vibration tolerance, piezoelectric effects, resilience to industrial cleaning methods/flux removal, environmental resilience, expected lifetime, mean time between failures, expected failure modes, resilience to transient outside specification operation, self-healing ability, etc, etc), price, size, availability, etc. Without knowing all the details of any particular choice, it's sometimes difficult to guess as to why it was made.

[deadlyshart]

Crap, THAT long?  Can you get Ebay items? 

I would certainly put the ceramic in temporarily just to finish a build! Won't hurt anything, but if it's in a sensitive place it could cause drift,  things like that.  Probably not eve then.   Just replace it when your good stuff comes! 

Ahh, I'm half joking. My most recent batch of aliexpress orders was placed on 3/24 and I got them a couple days ago, so that's only about 2 weeks. Not bad at all considering the prices are often 2-5x lower than amazon. I've found ebay to be pretty much the same; I either end up paying amazon prices for US shipping, or aliexpress prices and then it ships on the same time scale. It's kind of silly of me because I can probably afford any of this, but something bugs me about spending $5 on shipping for a $2 item repeatedly... what can I say, we just held seder here in our house 

[GibsonGM]

Good point, Phoenix, that's what I meant by "drift".   Good description.

Ha ha, I thought you might've lived in kazakhstan or somewhere, Shart...2 weeks is ok. I usually get things in 7-10 days.   Yes, ceramics are pretty much poo, mostly they're used in power supplies.  Audio-wise, too many unknowns in how they will act.

I went to Seder once, I even got to read out loud about bricks and mortar    (I was a bricklayer at the time, ha ha). Enjoy. 

[duck_arse]

why not just use a poly of small enough size? problems solved, nothing to dry out, no right direction to point.

[EBK]

The answer I would give has already been said: just order the part you need and wait.
For the waiting problem, start another project to pass the time.
For the parts-cost-to-shipping-cost-ratio problem, use it as an excuse to order more parts.

Since I love an academic detour (meaning this is a less than serious suggestion that I happen to find interesting), I'll throw this nugget in, although I've never used it: https://en.m.wikipedia.org/wiki/Capacitance_multiplier

[GibsonGM]

That concept is neat, Eric....kind of odd, but neat 

[samhay]

why not just use a poly of small enough size? problems solved, nothing to dry out, no right direction to point.

^this.

We are bemoaning how bad ceramic caps are while telling the OP to stick with a bipolar electrolytic?
1 uF polyester box capacitors would be my first choice.

[GibsonGM]

why not just use a poly of small enough size? problems solved, nothing to dry out, no right direction to point.

^this.

We are bemoaning how bad ceramic caps are while telling the OP to stick with a bipolar electrolytic?
1 uF polyester box capacitors would be my first choice.

Agreed.  Much better quality.   Nobody could REALLY design something to 'need' an electrolytic, LOL.   
Just that it would work fine, as will the poly - the poly has more utility in other builds too, so you can get a *few* at a time (I'd get 25 lol)

[ElectricDruid]

Alternatively, use a pair of 2.2uF electrolytic caps back to back to give you a bipolar cap of half the value.

This is a common enough trick where you need a large bipolar value.

But yeah, film caps are the way to go for audio quality.

HTH,
Tom

[bluebunny]

I'm with my esteemed colleagues.  Film, film, film, film, film.   

[Phoenix]

We are bemoaning how bad ceramic caps are while telling the OP to stick with a bipolar electrolytic?
1 uF polyester box capacitors would be my first choice.

I don't think anyone suggested bipolar electrolytic was the best choice, I certainly didn't, I just answered Deadlyshart's question about whether a ceramic was a suitable replacement for such.
I also agree that a film cap would be the best choice, but that's not what he asked.

[alanp]

Hi guys, I'm getting into making analog synths rather than effects pedals (I hope it's okay to ask about relevant stuff here!), and the beginners guide I'm following at one point has me using a 1uF aluminum bipolar nonpolarized capacitor. I don't have any of these; I have 1uF electrolytic caps but I assume the nonpolarized part is important here. I also probably have some 1uF ceramic disc caps, which are unpolarized... could I safely use those?

Also, why *do* people use aluminum ones instead? thanks in advance for any advice.

What is the circuit in question?

When I need a odd part like that, I usually put it in the queue for my next Mouser order (this occurs when 1, I have the spare munney, and 2, I have enough stuff I need to get free shipping.) eBay gambling is only worth it for rare parts that Small Bear, DIYGP.AU and the like don't carry.

As for capacitor types, it's amazing what you can get away with. I put one echo module together from a kit and the supplied caps from the seller were entirely ceramic caps where most people would use film box, with a few electrolytic ChongX Taydas for spice in the uF range.

[dschwartz]

You have these options
- use the non polarized 1uF electro..you wont hear a difference
- use the 1uF ceramic (i assume it is a multilayer, ceramic disk caps are mostly COG type and they are not available in 1uF values). If the cap is on the output or the power supply, you won't hear a difference.
- put 2 2.2uF in series or 2 470nF in parallel with opposed polarities.
- get a polyester cap.

The main question was if it was critical.. well. It is not critical.

[PRR]

Historically, 1.0uFd ceramic was impossible, 1uFd paper cost real money, 1uFd electrolytic could be quite low-price, and 1uFd NP electrolytic not much more.

If you build for profit, Pennies Matter!!

If you build a synth, you will be drowning in caps and at some point Pennies Matter.

Cost no object-- ceramic is "perfect" up to 1,000pFd but increasingly crappy past there. I can hear some 0.01uFd ceramics in audio paths. I would not use a 1uFd in audio even as power bypass unless there was a fat reliable electrolytic across it.

Film caps were historically all 100V and up, large, expensive. Improved film-making has made uFd size LOW-voltage film caps more affordable than they used to be.

BTW, most research suggests that back-to-back e-caps work "about" as the same value of one. So "two 1uFd polar" makes "one 1uFd bipolar". The cap that is "backward" at a given instant is near-short. At least this is true in 48V systems.