Polarized cap orientation

[croquet hoop]

I was observing the decimator schematic and it looks like one electro capacitor is implemented in reverse in the circuit:



Does the orientation of this cap serves a specific purpose? I know that it is generally safe to replace a polarized cap with a non-polarized one, but does this still apply in this case? (MLCC caps are so much practical than electro caps in small builds)

[rullywowr]

Hmm, I'm not a rocket scientist but that does look suspect!  Where did the schematic come from and is it verified?  A mlcc or tant 10uF would work there...id go with the tant personally. 

Of course you could always don some safety goggles, a helmet with a padded chinstrap, and a breaking bad yellow hazmat suit and try an electro in the more common orientation. 

[RobA]

Edited: actually it does look like it's not a split supply. Where's the whole schematic from?

This section on the input from the 2181 spec sheet might help to

The Input pin (pin 1) is a virtual ground with negative feedback provided internally ...

So, depending on what's happening with the supply voltages, the cap could be in the right way.

[croquet hoop]

It comes from here : http://www.efekty-diy.pl/pdf/Decimator%20G-String.pdf (source)

I triple-checked the schematic against a verified layout (Slade's) and it looks correct. There is just that capacitor which puzzles me — I do not doubt the orientation is correct, but I am curious about why it is turned this way.

[RobA]

The VCA's ground pin is attached to VAA, which is at half supply. The input to the the buffer op-amp is going to be biased a bit below this. From the quote above from the spec sheet, it looks like the input to the VCA should be held a bit above the output of the buffer. I would wonder if it's enough to keep the cap forward biased though. I would guess that a tantalum cap could be trouble here because the voltage swing could reverse bias it. A good MLCC might be fine, but a bad one could suck because of the distortion from the voltage swing at a raised voltage level.

[midwayfair]

Need to know the supply voltages and voltages on the pins of each chip.

If the op amp is on a split rail (suspected due to the op amp specifying "V-" instead of "ground"), then the output will be 0v, but the THAT chip could be running at a different voltage. What's the rest of the schematic?

[RobA]

Matt has the link to the full schematic a couple of posts up. The spec sheet for the VCA is at
http://thatcorp.com/datashts/THAT_2181-Series_Datasheet.pdf. The notes about the input pin of the VCA are on page 7.

[DutchMF]

Hmm, I'm not a rocket scientist but that does look suspect!  Where did the schematic come from and is it verified?  A mlcc or tant 10uF would work there...id go with the tant personally. 

Of course you could always don some safety goggles, a helmet with a padded chinstrap, and a breaking bad yellow hazmat suit and try an electro in the more common orientation.

Isn't a tant supposed to be orientated correctly as wel?

[rullywowr]

Yup!

[midwayfair]

Isn't a tant supposed to be orientated correctly as wel?

They don't blow up as easily.

Sometimes.

[RobA]

My personal experience (yes, I have blown them both up on purpose) is that tantalum electrolytics fail more easily than aluminum electros.  I have seen it said on several audio related sites that tantalums are better at handling reverse voltages, but this doesn't seem to be supported by the manufacturer's docs. A quote from Kemet's site:

All electrolytics, such as tantalums, are inherently polarized devices, and are subject to failure when exposed to reverse voltage. Such reversals may result from insertion errors, or from marginal circuit design. The degradation of tantalums begins with a reverse voltage at ~15% of rated voltage, and for aluminums, this begins at ~60% of rated voltage in reverse.

Here's a link to doc from Kemet on investigating failures in tantalum caps
http://www.kemet.com/kemet/web/homepage/kfbk3.nsf/vaFeedbackFAQ/DA8446492D74E7268525722500690E31/$file/2002%20CARTS%20Reverse%20Voltage%20Failures.pdf

The other point is that tantalum caps fail to a short circuit.

The main point is neither aluminum nor tantalum should be reverse biased, ever.

[kothoma]

I seem to remember having heard some decades ago that tantalum electrolytics may even blow up for no reason (other than age)?

[croquet hoop]

Thanks for all the info guys. I have also looked up build reports to see what has been done, and it looks that everyone has been using either an electro cap following the orientation of the schematic (see here or there), or even tantalums (here).

I'd rather use a MLCC cap for space reasons, but I'd rather play safe, so i'll bite the bullet and leave enough room on the board to fit a 10 uF electro cap.

[RobA]

Thanks for all the info guys. I have also looked up build reports to see what has been done, and it looks that everyone has been using either an electro cap following the orientation of the schematic (see here or there), or even tantalums (here).

I'd rather use a MLCC cap for space reasons, but I'd rather play safe, so i'll bite the bullet and leave enough room on the board to fit a 10 uF electro cap.

There's no electrical reason you can't use an MLCC. The cap doesn't need to be polarized. The only issue with an MLCC is that, since it is voltage biased, distortion issues from a ceramic cap will be more likely. It still might be fine sounding. A good audio grade Al electro could sound better. [opinion] A tantalum electro will sound like crap. [/opinion] You most likely will have no audio issues with a good ceramic.

[croquet hoop]

Noted, thanks! I think I'll just rearrange the board to allow all options, and use a socket to try different possibiities.