For a laugh, just wired up an unbypassed 100uF NP electrolytic in series with my speakers. Bass is rolled off (obviously) but the sound remains the same. No "blindingly obvious" distortions or bad sound at all, even with it turned up. All the treble is there. Imaging is still there.
Shorting the capacitor while playing music just brings back the bass. The rest of the spectrum sounds the same. In passages with little bass there is no change in sound.
My scary wacky theory: Electrolytics sound fine.
Shorting the capacitor while playing music just brings back the bass. The rest of the spectrum sounds the same. In passages with little bass there is no change in sound.
My scary wacky theory: Electrolytics sound fine.
This is a very silly experiment. 100uF changes the sound so dramatically that it's pretty much impossible to claim that everything but the bass remains the same - the entire perception of sound changes by removing bass. Use several thousand uF non-polar if you intend to make any worthwhile conclusions. You should also describe your system and your hearing acuity. Do you hear differences between audio components? Between cables? Equipment support? Passive parts?
And even more importantly: what's inside the crossover? A whole lot of really nasty electrolytics?
And even more importantly: what's inside the crossover? A whole lot of really nasty electrolytics?
analog_sa said:This is a very silly experiment. 100uF changes the sound so dramatically that it's pretty much impossible to claim that everything but the bass remains the same - the entire perception of sound changes by removing bass. Use several thousand uF non-polar if you intend to make any worthwhile conclusions. You should also describe your system and your hearing acuity. Do you hear differences between audio components? Between cables? Equipment support? Passive parts?
And even more importantly: what's inside the crossover? A whole lot of really nasty electrolytics?
My hearing is very good. I can still hear bats.
As for the change in sound, the speakers are Mordaunt Short Avant 902s, with socks stuffed in the ports. There isn't that much bass to remove in the first place so you only lose a teeny bit of low end information. Everything else sounds the SAME. Trust me.
My amplifier is a Radford STA-15 and I can easily hear the difference between it and a leak stereo 20
bigwill said:
That's like the distance from the Earth to the Moon in audio terms
What the undersized caps certainly do, apart from rolling the bass is shift the phase at a very audible frequency. Even the most hardcore objectivists would admit this to be potentially audible and certainly not good for soundstaging and other spatial effects.
Which means your setup is faulty. And certainly doesn't mean that you or anyone else would necessarily hear a difference with a properly sized cap. Quite possibly there are similar or worse caps inside the crossover.
We've done blind A/B tests using identical speakers with the same crossovers, with the exception that one pair had electrolitics and the other used PP. Statistically it was a draw as to correctly IDing the components. The only thing that the 'litics lack is the long term stability of their values, the tight tolerances and the eventual expiration of the cap itself.
Of far greater concern, sound-wise, are (IMHO) the inductors. These are audible if the wire guage is small and they're poorly wound.
I personally use electrolitics (with tiny PP bypass caps) for my crossovers and indulge myself with decent inductors. It works well and several people that are of a like mind have won speaker contests for best sound.
Best Regards,
TerryO
Of far greater concern, sound-wise, are (IMHO) the inductors. These are audible if the wire guage is small and they're poorly wound.
I personally use electrolitics (with tiny PP bypass caps) for my crossovers and indulge myself with decent inductors. It works well and several people that are of a like mind have won speaker contests for best sound.
Best Regards,
TerryO
TerryO said:
What exactly is audible in inductors? Different sound? Dstortion? Or just some dB lost?
I expect that, if you exchange an inductor with 0.3 Ohm with another with 1 Ohm you will loose some dB. (and crossover f will change)
But if you design crossover, considering inductor resistance, it can sound good with any wire. The problem with small gauge is that you loose power (dB).
Inductor DCR is clearly audible if used in series with the bass driver. Damping factor will be effected and as such a high DCR will give loose bass with a slight peak around FS, especially if the amp itself has a high output Z.
Caps can sound differnt but again I think its mostly down to series resistance, as 'better' caps usually have lower resistance and so if in series with a tweeter the level could go up by a fraction giving a perception of more detail. This can be designed out in the Xover, the inductors DCR can't always.
Caps can sound differnt but again I think its mostly down to series resistance, as 'better' caps usually have lower resistance and so if in series with a tweeter the level could go up by a fraction giving a perception of more detail. This can be designed out in the Xover, the inductors DCR can't always.
If I wound a series inductor for my woofers, it would contain vastly more wire than my interconnects. If the wire gage was sufficient to keep the DCR low, they'd be huge. Given the current price of copper, the whole approach makes a lot less sense than it did many years ago. IMO, even if you had to use chip amps, you'd be far better off building a good active crossover and bi or tri amping. I tri amp, and the difference in clarity and adjust-ability is so great, I'd never consider going back to passive crossovers.
TerryO said:
I would suspect that any electrolytic audibility would be on low level passages where the zero crossing effects might be audible.
But I am curiuos about what you mean by "poorly wound" and why wire guage would affect audibility. Is it the wire heating and resistance change that does this? What is your hypothesis?
gedlee said:
Hi Doc,
You've asked for a hypothesis and I'm sorry to relate that I don't have one. I leave that to smarter people, like yourself, to determine "the why's" of a phenomena.
As I did my graduate work in History, I will venture a guess, that historically it's proved to be cheaper using substandard or lower grade parts and that this may improve the profit margin. (Jeez Terry, that's a real news-flash!)
"Poorly wound" is my descriptive phrase for a coil that is not uniform in winding tension, has erratic layers that may even have the wire cross back and forth randomly, etc. IOW, it looks like something a very young Terry Olson would have produced.
(Note: I have seen coils that match this description in some speakers that I've looked at, so someone out there is actually selling them!)
I have used cheap inductors for prototyping, and at low levels there isn't much difference AFAIK, but when you start to play them louder, the sound becomes inconsistent, which doesn't occur when larger, better quality inductors are substituted. Again this is an opinion, based on what (I think?) I'm hearing, and as such is hardly passes muster as a scientific explanation.
Best Regards,
TerryO
Some comments about poorly wound inductors:
- Poorly wound can lead to loose turns, which can vibrate.
- Poorly wound will usually lead to higher DCR than a nicely wound inductor.
- If winding is not consistent, and one winds to a specifed L, DCR tends to be inconsistent.
Even nicely wound inductors can have surprising effects. In some cases, foil inductors can exhibit *much* higher AC losses than wirewound ones, due to layer effects. These are another eddy current effect related to skin effect. If you have an impedance analyzer available, try measuring one of those foil air-core inductors over the audio frequency range sometime and comparing its resistance to a measured DCR value.
John G
- Poorly wound can lead to loose turns, which can vibrate.
- Poorly wound will usually lead to higher DCR than a nicely wound inductor.
- If winding is not consistent, and one winds to a specifed L, DCR tends to be inconsistent.
Even nicely wound inductors can have surprising effects. In some cases, foil inductors can exhibit *much* higher AC losses than wirewound ones, due to layer effects. These are another eddy current effect related to skin effect. If you have an impedance analyzer available, try measuring one of those foil air-core inductors over the audio frequency range sometime and comparing its resistance to a measured DCR value.
John G
In my designs the DCR of the coil is considered and held very tightly so the DCR of the coil doesn't matter unless:
a) it varies from one inductor to the next, which has not been the case in my experience
b) it changes significantly with heat, which I have to conclude is not a big issue since the change is small and in my speakers the crossover never even gets warm
I have long wanted to look at inductor nonlinearities, but I can't seem to find any physices that would say that it is an issue. Of course I would agree that vibrating wires would be an issue, but thats easy to avoid and in the inductors that I buy they are all bonded wire.
Are the audibility issues stricktly limited to very poor inductors? Or has anyone found a decent inductor to be audible?
a) it varies from one inductor to the next, which has not been the case in my experience
b) it changes significantly with heat, which I have to conclude is not a big issue since the change is small and in my speakers the crossover never even gets warm
I have long wanted to look at inductor nonlinearities, but I can't seem to find any physices that would say that it is an issue. Of course I would agree that vibrating wires would be an issue, but thats easy to avoid and in the inductors that I buy they are all bonded wire.
Are the audibility issues stricktly limited to very poor inductors? Or has anyone found a decent inductor to be audible?
If the inductors "damping factor" or Q is taken into account, then its not a factor. The inductors Q has to be considered, you can't just ignore it and if you do design for it then it has been accounted for and should not be an issue.
When these audibility tests are done, are inductors just swapped? Is the Q difference taken into account and the crossover modified accordingly?
When these audibility tests are done, are inductors just swapped? Is the Q difference taken into account and the crossover modified accordingly?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.