Capacitor upgrade in crossover - You CAN'T handle the TRUTH! - Part 3

[restorer-john]

Well, a cap that fails that way wouldn't be one an engineer would choose for that purpose, right?

I'm sure you may have come across as many shorted electrolytics as I have. And they are by far the most commonly used type of capacitor used in commercial loudspeaker crossovers and have been for basically ever. When they do fail short-circuit, the tweeter or midrange fails very rapidly, due to full range content.

And under-rated capacitor in a loudspeaker network (series or parallel) which fails (open circuit, or worse, short) in the presence of high voltages is not remotely a 'good thing'. Driver protection can be achieved much better ways than sacrificial capacitors with unpredictable failure modes.

 

[SIY]

And under-rated capacitor in a loudspeaker network (series or parallel) which fails (open circuit, or worse, short) in the presence of high voltages is not remotely a 'good thing'. Driver protection can be achieved much better ways than sacrificial capacitors with unpredictable failure modes.

NHT disagrees and I have some personal experience that their approach works. Which is how I inherited two pairs of "blown" 3.3s and got them running again with about a dollar in electrolytic replacement.

 

[Davide]

I have read the whole series, but the opinions of various users have left me with doubt. In summary, from the point of view of reliability and correct signal filtering (according to the design I mean), are film capacitors better or not?

 

[Killingbeans]

Electrolytics suck. They are a necessary evil in most electronics. It's not so bad in low power applications, but in anything that gulps wattage and gets hot, electrolytics will cook and die eventually. So, not the best in terms of reliability. But I bet they work just fine in terms of signal fidelity, especially if the design accounts for ESR and what have you. Film capacitors will be "better", but also overkill in most instances.

 

[fineMen]

Electrolytics suck.... evil ... not so bad in low ... not the best ... reliability. ... just fine ... if the design accounts for ESR and what have you. ...

Don't know if the Falcon crossovers were mentioned already. I had a pair of KEF Calinda which after about 43y was in perfectly operational condition. If all values of the capacitance mach the imprinted numbers within 5% max, I would argue that those electrolytics stood the test of time. The speakers in total complied to specs anyways. Only that the specs of the design as a whole doesn't comply to contemporary wisdom.

 

[Killingbeans]

Some electrolytics are definitely better quality than others (not because they are "audiophile grade", just better quality in general), and some electrolytics gets tortured a lot less when good design practices are being utilized. Good electrolytics in a good design will last a very long time without problems.

Crap electrolytics in bad/cheap designs, on the other hand, will almost always be the first thing to die.

 

[fineMen]

Crap electrolytics in bad/cheap designs, on the other hand, will almost always be the first thing to die.

Not to bother You, but isn't the last statement just tautology? Logically crap isn't useful, as being not useful is the definition of crap ;-)

But how much crap is out there with a quite common product? I doubt there is many. But, admittedly, I'm neither an e/ engineer nor an attorney.

 

[dfuller]

Not to bother You, but isn't the last statement just tautology? Logically crap isn't useful, as being not useful is the definition of crap ;-)

But how much crap is out there with a quite common product? I doubt there is many. But, admittedly, I'm neither an e/ engineer nor an attorney.

In speakers, not much - passive crossovers are not what I would consider high stress. But in amplifiers and power supplies? The failures are common. Worse with tubes and with linear regulators, as both give off a lot of waste heat.

 

[Doodski]

linear regulators

Very handy devices that can be applied with minimal muss and fuss. My comment though revolves around the small package switch mode power supply. More difficult to apply but more better all around.

 

[Killingbeans]

But how much crap is out there with a quite common product? I doubt there is many

If I say "subwoofer plate amp", I'll bet most audio repairmen in here will feel the pain

 

[fineMen]

If I say "subwoofer plate amp", I'll bet most audio repairmen in here will feel the pain

First, I think to depart towards the power supply terrain isn't straight in line with the original suspicion. Namely that electrolytics don't have a place in speaker crossovers.

What too often isn't considered is the current limitation. Insofar high grade elctrolytics are rare and expensive. Since speakers are rated based an a statistical average over frequency, specific operation conditions could excced limits. E/g electrolytics in impedance equalizers in the bass network. It may work with pink noise, but will most likely fail with a boom boom program fed to it.

 

[Killingbeans]

First, I think to depart towards the power supply terrain isn't straight in line with the original suspicion. Namely that electrolytics don't have a place in speaker crossovers.

Absolutely. Like @dfuller says, electrolytics live a relatively peaceful life in crossovers.

My "electrolytics suck" comment was aimed at the component in general. I do not disapprove of the use of electrolytics in crossovers. It's not a critical application.

 

[fpitas]

Absolutely. Like @dfuller says, electrolytics live a relatively peaceful life in crossovers.

My "electrolytics suck" comment was aimed at the component in general. I do not disapprove of the use of electrolytics in crossovers. It's not a critical application.

They have improved tremendously over the years. So many millions are used in switching power supplies now, which drove a lot of engineering.

 

[dfuller]

They have improved tremendously over the years. So many millions are used in switching power supplies now, which drove a lot of engineering.

And yet, a lot of premature failures still happen. But that's off topic.

I don't see any inherent disadvantage to using electrolytics in passive crossovers if they make sense to use. They are ever so slightly higher distortion than film caps, but it's to a point that I don't think it particularly matters.

 

[Killingbeans]

They have improved tremendously over the years. So many millions are used in switching power supplies now, which drove a lot of engineering.

True, and they've become a lot smaller. It looks weird when you do a recap on something old and see, how much space the new electrolytics free up on the PCB.

But I'd still really, really like the industry to develop a more rugged alternative. I know it's a wet dream for them too.

...sorry for off-topic.

 

[MAB]

I have read the whole series, but the opinions of various users have left me with doubt. In summary, from the point of view of reliability and correct signal filtering (according to the design I mean), are film capacitors better or not?

Simply stated, the OP showed that the type of capacitor doesn't matter with respect to audible performance. And, changes in actual sound are more due to changes in the resistances of the capacitors (goes by many names... Equivalent Series Resistance - ESR, DC Resistance - DCR, etc.) which are trivial to model and simple to address if the new ESRs are an issue.
Yes, there is lots of chatter in the comments, but the OP showed in this experiment that capacitor rolling is a huge waste of time and money and the only audible impact on sound is due to the inadvertent changes in ESR due to the new components.

 

[levimax]

Don't know if the Falcon crossovers were mentioned already. I had a pair of KEF Calinda which after about 43y was in perfectly operational condition. If all values of the capacitance mach the imprinted numbers within 5% max, I would argue that those electrolytics stood the test of time. The speakers in total complied to specs anyways. Only that the specs of the design as a whole doesn't comply to contemporary wisdom.

You need to check the ESR on old electrolytic caps. The capacitance can be fine but often the ESR goes way up as they age.

 

[fineMen]

You need to check the ESR on old electrolytic caps. The capacitance can be fine but often the ESR goes way up as they age.

You (!) prove it, thank You. As I said, no degradation regarding the performance; none. The o'l KEFs complied to specs, which were even back in the day well documented.

 

[valerianf]

Electrolytic capacitors are cheaper and smaller.
That is a design choice for the crossover.

 

[levimax]

You (!) prove it, thank You. As I said, no degradation regarding the performance; none. The o'l KEFs complied to specs, which were even back in the day well documented.

I think you misunderstood me so I will rephrase. "When testing old capacitors both their capacitance and their ESR need to checked in order to determine if they still meet their specifications". It is entirely possible that 43 year old electrolytic capacitors are fine. Testing capacitors by measuring old speaker output against specifications is a very indirect way of testing their condition and not reliable. It doesn't mean the speakers don't work fine.