Capacitor questions

[Johnny1234]

Hi all,
First the reason for this post, I want to re cap some Otari MTR-90 audio modules.

My question is three fold.

1) The Otari card is loaded with many bipolar electrolytic capacitors. We all know how pricey that can get.
Can I substitute polarized capacitors in the circuit?
2) If the above answer is yes, is it correct that the positive of the capacitor connect to the output of the op amp.
I have many schematics and that seems to be the way it's usually done.
3) And final question why is it that the positive of the capacitor is connected to the output of the op amp. It would
seem when operating normally the DC voltage at the output of the op amp is extremely low.

Happy Holiday's to all you have all been a great source of knowledge.
Johnny1234

 

[ccaudle]

is it correct that the positive of the capacitor connect to the output of the op amp.

The positive terminal of a polarized capacitor should connect to a circuit node which is more positive than the node to which the other terminal connects.

why is it that the positive of the capacitor is connected to the output of the op amp.

A typical reason would be that the circuit is powered by a single positive supply, and the capacitor is referenced to either the 0V reference node of the power supply, or to the mid-point which is half of the positive supply voltage.

We all know how pricey that can get.
Can I substitute polarized capacitors in the circuit?

That seems to me to be the equivalent of the question "were Otari engineers ignorant and used more expensive components without good reason?"
I would think that before anyone would be willing to answer in the affirmative to that question, any...well, any prudent person would want to make a thorough analysis of the original circuit to determine DC voltages and maximum AC voltages at that location in the circuit, as well as the distortion requirements for the circuit to determine if polarized caps are appropriate or not.
It would likely be an informative exercise whatever the outcome, but just based on the reputation and longevity of the equipment I would not start my bet with the assumption that Otari hired incompetent engineers.

 

[Johnny1234]

Can't the offset at the output drift negative with respect to the next stage?

 

[Ike Zimbel]

Can't the offset at the output drift negative with respect to the next stage?

I've certainly found it to be unpredictable whether you're going to see positive or negative offset in practice. WRT bi-polar caps being pricey, Panasonic SU series are reasonably priced, sound excellent and are very long lasting. I have been using them for over twenty years.

 

[ccaudle]

Can't the offset at the output drift negative with respect to the next stage?

That question has no context. As asked the question is impossible to answer because there is no reference to a specific circuit topology.

 

[Johnny1234]

Bi polar capacitors are also on their way to obsolescence. As per Mouser Electronics.

 

[Johnny1234]

"That question has no context. As asked the question is impossible to answer because there is no reference to a specific circuit topology."

How about simple inverting and non inverting op amps. Why is the positive of the cap always tied to the output of the op amp.
Sometime I feel it is impossible to get a simple answer here, but Drip questions will get a thousand answers.
Frustrated.

 

[JohnRoberts]

"That question has no context. As asked the question is impossible to answer because there is no reference to a specific circuit topology."

?

How about simple inverting and non inverting op amps. Why is the positive of the cap always tied to the output of the op amp.

it isn't ... Polar capacitors are connected relative to node voltages. Polar capacitors require a positive bias between + and - end.

Sometime I feel it is impossible to get a simple answer here, but Drip questions will get a thousand answers.

don't know what a "Drip question" is . I quick search did not reveal what that means.

Frustrated.

me too...

JR

 

[ccaudle]

Why is the positive of the cap always tied to the output of the op amp.

I will refer you back to my earlier explanation, not always the case but is likely the case in single supply designs:

A typical reason would be that the circuit is powered by a single positive supply, and the capacitor is referenced to either the 0V reference node of the power supply, or to the mid-point which is half of the positive supply voltage.

I did not write as clearly as I could have there. A more concrete example: older discrete circuitry sometimes ran from a single +24V power supply, so the output was generally around +12V DC. In that case the obvious connection is positive terminal of the output cap to the circuit, and the negative terminal would be to the output connector.
The same situation would exist with single supply op-amp based designs.

For dual supply designs you have to consider the specific op-amp (or other circuitry) design and whether the input bias current requirements will always cause the output to drift in one direction. If the op-amp design is such that it could be slightly positive or negative (as mentioned by Ike Zimbel) then it seems somewhat arbitrary which way you connect a polarized capacitor.
If the capacitor is sized large enough it may not may much practical difference. The voltage across the capacitor terminals is what matters, not the total output voltage. If the voltage across the capacitor is never large enough to reverse bias it to a substantial degree then the capacitor should not be degraded.

 

[ccaudle]

?

Johnny was quoting me there, he asked "can't the output drift negative?" I was just trying to point out that there is really no way to answer that question. Output of what? A triode? A single (positive) supply discrete transistor design? A single supply op-amp design? Then no. A dual supply design? Depends, is it discrete transistors, all NPN? Maybe, if the input is tied to 0V and you have several BE drops in series, sure, output would be negative.
If the circuit uses op-amps does the input pair always need bias current in a particular direction? If so that might develop an offset voltage across the input bias resistors that is always in the same direction. If the op-amp is FET input then there typically isn't a predictable DC bias current.

So I stand by my original answer, without context of a specific circuit to discuss there is no way to give a sensible (short) answer.

don't know what a "Drip question" is .

Drip questions

Abbey knows what Drip questions are.

Seems you were lucky to not get pulled into those threads.

 

[JohnRoberts]

Seems you were lucky to not get pulled into those threads.

not exactly luck...

JR

 

[Whoops]

Hi all,
First the reason for this post, I want to re cap some Otari MTR-90 audio modules.

My question is three fold.

1) The Otari card is loaded with many bipolar electrolytic capacitors. We all know how pricey that can get.
Can I substitute polarized capacitors in the circuit?

Hi mate,
not what you've asked but it's just my honest advice.

Otari MTR-90 is a valuable and good machine,
my advice is: replace those Bipolar capacitors with new Bipolar capacitors.
You can still get those from many different sources, just do an excel table and compare the prices between the different sources and choose the best for you. And change Bipolar for Bipolar, don't try to be "creative" and find "shortcuts or workarounds" in such a great and valuable unit.

You can gey Bipolar caps from these sources:
Mouser
Digikey
Newark
Farnell
Arrow.com
TME
Future Electronics
Banzai Music

This is good advice
My 2 cents

 

[Whoops]

Banzai Music, has a great selection of Bipolar Capacitors for good prices,
Check this link:

https://www.banzaimusic.com/search.php?mode=search&page=1&keep_https=yes






As you live in the US and Banzai store is in EU (Germany),
if you would like to order from there you can send me the list and quantities you want , I can order from the store,
and then on 22 of January I will be in the US for the NAMM show, I can take the caps with me to the US and post them to you from inside the US.
If this helps, let me know

 

[kidb]

I agree with those who say replace with bipolar, I’ll second the panasonics. Easier than digging into topology, for both of us lol and not thaaaat expensive.

 

[Khron]

Worst-worst case they can always be replaced with back-to-back series pairs of normal but double the value electrolytics (especially if the originals are axial but new ones are obscenely overpriced or something)...