A Mistake Changing the TYPE Of Capacitors In Old Klipsch Speakers

[Invidiosulus]

That comment about the kkk was out of line. There is no need for that sort of referrence on this forum.

[freddyi]

yeah - that's me - I like to ask questions and in one note from PWK (which may not have anymore), I brought up Thiele & Small parameters - PWK may have said something to the effect - what are those? []

it would have been interesting in Tony Gee's cap survey if he had documented capacitance & ESR with good meters

[ka7niq]

SNAP...love the sound of a trap snapping shut.

First of all, I expect those Mylar caps were not to bad for ESR back when they were new. Bet they stayed good through the warranty period. Now they are over 20 years old.

Second, I have never found a motor run cap or a paper in oil cap in a Klipsch crossover. Some of those from the 50s might have been PIO because in those days, not much else was available. There would not be anything wrong with a motor run cap if made right. Just never saw Klipsch use them. The oil filled caps I have seen in Klipsch crossovers are film and foil in oil but way to low in working voltage to be motor start or motor run. Bet those tested really good for ESR when new.

You can prefer any sound you want. But remember that is you. What I prefer or you prefer has nothing to do with what is right. Right is to reproduce the sound fed to the speaker as closely as possible. Your preference then (or my preference) has absolutely nothing to do with that. That is what tone controls and equalizers are for. Those things let you make the speaker sound like you want it to even if that sound is not faithful to the reproduction. I am fortunate that a flat frequency response sounds right to me, so I don't have to mess with stuff to get some elusive sound that works for me.

Anyway, here is the reference you wanted. I believe that freddyi is here with us in this thread.

Since I have cornwall 2's, I am referring to them, not Klipschorns.

MOST Mylar are not Low ESR Caps, yet they are used in the Cornwalls.

Theoreticallly Bob, what effect does ESR have on a crossover ?

Does it change Q, slopes, etc ?

Floyd Toole demonstrated a dip is betteer to have then a peak.

So, since we dont know the original caps ESR, why put a lower ESR cap in, and risk a peak ?

[Don Richard]

Are the type A and type AA crossovers identical, except for the cap type, same slopes, cross frequency, etc ?

The circuit for the AA is the same except for steeper slopes on the tweeter (18 dB/oct vs 6 dB/oct, approximately) and zener diodes for tweeter driver protection. Different types of caps were used in the AA over the years, probably an availability/cost issue.

[freddyi]

here's a pdf where a guy was trying to pin down differences between several 4.7uF caps recording spl (from pink noise and sine) and waveforms from a dome tweeter- his electrolytic's value was on the lean side and his ESR meter the cheap MAT Electronics unit

"Spectral Waveform Analysis Of Capacitors"

http://www.classicspeakerpages.net/IP.Board/index.php?act=attach&type=post&id=2705.

Groomlakearea51 -re: paper and oil - did you try any of the KBG caps?

look at the Klipsch KBG cap crossovers on Ebay Germany (are those Orange Drop bypass caps?):

http://cgi.ebay.de/Crossover-Frequenzweiche-fuer-Klipsch-Eckhorn-La-Scala_W0QQitemZ220540766039QQcmdZViewItemQQptZLautsprecher_Selbstbau?hash=item3359410757

[BEC]

I can't go back in time and test the particular Mylar caps you have in the Cornwall II crossover when they were new. But I can make some assumptions about them based on testing a bunch of them that are over 20 years old. The average one of those shows ESR of around 0.5 ohms. A few of them test at around 0.3 ohms and some are at 0.6 ohms (some are much worse, but let's throw those out of the mix).

Even for those cheap mylar caps, I doubt they left the factory with a variation of that much between them. So, I assume that ESR increases with age and that the increase is not the same for every cap. Sort of like spark plugs in your car aging. At some point one of them gets bad enough to get your attention and you change all of them. OR, some might just change them all at some time before one get to the point of not working.

"Theoretically Bob, what effect does ESR have on a crossover ?"

ESR effects the crossover the same way a real resistor added in series effects the crossover. When it increased to some point, there is audible attenuation of the signal to the driver. It also effects impedance of that section of a crossover in exactly the same way a resistor would. So, as ESR increases, impedance increases and the crossover point changes. In a simple 1st order crossover's 8 ohm tweeter circuit, increasing ESR by 1 ohm would move the crossover point lower by about 500hz.

What is audible in ESR changes? Well, that doesn't give any peaks or dips electrically. Since increasing ESR moves the crossover points, acoustically, you might move into a non-linear portion of the driver's curve. The major audible effect of ESR in my own listening tests is that increasing ESR, attenuates the highest frequencies. My test was to try to determine my threshold for hearing increasing ESR. After a bunch of blind tests, I could always hear the effect of increasing ESR by 0.5 ohms. I could not reliably hear the effect of a 0.3 ohm increase in ESR. I think likely someone with younger ears than mine could likely hear the change with somewhat more precision.

On the question, could ESR ever be too low, I don't think so. For an equivalent, I just looked at a bunch of crossover schematics. In lots of cases, DCR for inductors, (another unwanted thing where lower is always better), the specs I saw for those always said something like "DCR < 0.25 ohms" meaning you could go lower but no higher.

By the way, just realized that i have some much more recent mylar caps in an AL-4 crossover. I just lifted one end of one of the 4uF mylar caps on that board and tested it for capacitance and ESR. These are about 2004 vintage, I think so, about 6 years old. Test in at about 0.13 ohms ESR now. I would expect these may have been below 0.1 ohms ESR when new.

Bob Crites

[freddyi]

hi Bob - does your B&K 885 perform its ESR measurement at 100KHz? - or at the lower switchable frequencies? Is ESR a frequency indendent quantity (other than lead inductance) that remains about the same at midrange horn frequencies as it does at the very top of the audio spectrum? Does capacitor value matter for ESR? For those with time to burn it might be interesting to see RTA sine sweeps of say Klipsch Cornwall or Heresy with old caps (or ancient NOS caps) verus fresh caps known to be good putting a dummy load across the woofer then another dummy load across tweeter or mid output of the crossover to examine losses on the midhorn or tweeter to see what can be easily measured.. btw - the cheap MAT meter thinks the old AA Aerovox are about as good as the new Obbligato film/oil caps... be nice to see the results on a pro ESR meter

[BEC]

I always use 1khz on my tests. Generally as you go higher in frequency ESR looks better. At 100khz, nearly all look good, but don't think that means anything to me for audio use.

We do what you are suggesting every day in our lab for rebuilds and new crossovers. Now, after doing several hundred rebuilds, we just do an after rebuild trace on most to verify all is in spec before we ship them out. In general, the before and after trace on the average 20 to 30 year old crossover just shows a slight shift on the tweeter crossover point and slightly higher output for the high frequencies after the rebuild. The change is more dramatic for more complicated crossovers (more than one cap in series) than on the simpler crossovers.

One thing we sometimes see that probably has little or nothing to do with the parts used in the rebuild is improvement of the "noise floor" on the analyzer comparing the before and after rebuild traces. Often it improves by 10 db or more. I think that is mostly from just resoldering all the joints.

Bob Crites

[ka7niq]

I always use 1khz on my tests. Generally as you go higher in frequency ESR looks better. At 100khz, nearly all look good, but don't think that means anything to me for audio use.

One thing we sometimes see that probably has little or nothing to do with the parts used in the rebuild is improvement of the "noise floor" on the analyzer comparing the before and after rebuild traces. Often it improves by 10 db or more. I think that is mostly from just resoldering all the joints.

Bob Crites

Usually, as the frequency is increased, component selection gets more critical.

One would think parts would show their azz more, not less, as frequency is increased ?

That is why us Hams never use Mylar Caps in critical HF circuits like VCO's for instance.

Ditto on the resoldering of all joints in crossover!

On my old Celestion SL 600's I did exactly that, and noticed an improvement in overall sound and imaging quality.

[ka7niq]

I can't go back in time and test the particular Mylar caps you have in the Cornwall II crossover when they were new. But I can make some assumptions about them based on testing a bunch of them that are over 20 years old. The average one of those shows ESR of around 0.5 ohms. A few of them test at around 0.3 ohms and some are at 0.6 ohms (some are much worse, but let's throw those out of the mix).

Even for those cheap mylar caps, I doubt they left the factory with a variation of that much between them. So, I assume that ESR increases with age and that the increase is not the same for every cap. Sort of like spark plugs in your car aging. At some point one of them gets bad enough to get your attention and you change all of them. OR, some might just change them all at some time before one get to the point of not working.

"Theoretically Bob, what effect does ESR have on a crossover ?"

ESR effects the crossover the same way a real resistor added in series effects the crossover. When it increased to some point, there is audible attenuation of the signal to the driver. It also effects impedance of that section of a crossover in exactly the same way a resistor would. So, as ESR increases, impedance increases and the crossover point changes. In a simple 1st order crossover's 8 ohm tweeter circuit, increasing ESR by 1 ohm would move the crossover point lower by about 500hz.

What is audible in ESR changes? Well, that doesn't give any peaks or dips electrically. Since increasing ESR moves the crossover points, acoustically, you might move into a non-linear portion of the driver's curve. The major audible effect of ESR in my own listening tests is that increasing ESR, attenuates the highest frequencies. My test was to try to determine my threshold for hearing increasing ESR. After a bunch of blind tests, I could always hear the effect of increasing ESR by 0.5 ohms. I could not reliably hear the effect of a 0.3 ohm increase in ESR. I think likely someone with younger ears than mine could likely hear the change with somewhat more precision.

On the question, could ESR ever be too low, I don't think so. For an equivalent, I just looked at a bunch of crossover schematics. In lots of cases, DCR for inductors, (another unwanted thing where lower is always better), the specs I saw for those always said something like "DCR < 0.25 ohms" meaning you could go lower but no higher.

By the way, just realized that i have some much more recent mylar caps in an AL-4 crossover. I just lifted one end of one of the 4uF mylar caps on that board and tested it for capacitance and ESR. These are about 2004 vintage, I think so, about 6 years old. Test in at about 0.13 ohms ESR now. I would expect these may have been below 0.1 ohms ESR when new.

Bob Crites

The B&W engineer told me that changing crossover caps to a different brand/type will effect the Q of the crossover circuit, he warned me against doing it, unless the caps were shot.

Wonder what the effect of high or low ESR has in a crossover besides attenuation from bad, high ESR caps ?

Does it shift crossover point and Q ?

And, does it shift higher or lower ?

[Marvel]

"Theoretically Bob, what effect does ESR have on a crossover ?"

ESR effects the crossover the same way a real resistor added in series effects the crossover. When it increased to some point, there is audible attenuation of the signal to the driver. It also effects impedance of that section of a crossover in exactly the same way a resistor would. So, as ESR increases, impedance increases and the crossover point changes. In a simple 1st order crossover's 8 ohm tweeter circuit, increasing ESR by 1 ohm would move the crossover point lower by about 500hz

Wonder what the effect of high or low ESR has in a crossover besides attenuation from bad, high ESR caps ?

Does it shift crossover point and Q ?

And, does it shift higher or lower ?

[:^)]

[ka7niq]

Yes, what did the B&W engineer mean when he said different types of caps can change the Q of the crossover ?

[BEC]

Yes, what did the B&W engineer mean when he said different types of caps can change the Q of the crossover ?

First get him to tell you what he means by "different type caps".

Bob

[jerohm]

...

Sometimes lossy, high ESR caps are WHAT you want, just the thing to tame a bright, shrill horn!

no, MAYBE to tame less than optimal recordings or electronics?... higher ESR is NEVER desirable.

[freddyi]

one thing I don't see come up much in discussion of the old stock networks is PWK's assessment of "Otala distortion" being reduced by high reflected Z in the midrange due to the autoformer attenuator - especially on Heresy - is that still considered to be a positive feature vs swamping?

[ka7niq]

Yes, what did the B&W engineer mean when he said different types of caps can change the Q of the crossover ?

First get him to tell you what he means by "different type caps".

Bob

He meant changing the stock caps out for different types.

I understood him to say that even different brands of poly or mylar caps were different enough electrically to change the crossover.

[BEC]

Yes, what did the B&W engineer mean when he said different types of caps can change the Q of the crossover ?

First get him to tell you what he means by "different type caps".

Bob

He meant changing the stock caps out for different types.

I understood him to say that even different brands of poly or mylar caps were different enough electrically to change the crossover.

The two things that could be different are capacitance and ESR. Is he claiming that there is something else?

Bob

[freddyi]

IIRC Stan Ricker once told me he liked several motor run caps in series as the highpass (~2uF total value or so) on Altec Duplex[6] and that he thought that the plate area of the cap had something to do with its tone (I beleive he let the woofer run wide open attenuated by its felt disc)

[BEC]

IIRC Stan Ricker once told me he liked several motor run caps in series as the highpass (~2uF or so) on Altec Duplex and that he thought that the plate area of the cap had something to do with its tone (I beleive he let the woofer run wide open attenuated by its felt disc)

Freddy,

You must know everyone. I haven't talked to Stan in a couple of years I guess. Need to check in with him.

Bob

[freddyi]

I basically don't know anyone nor have traveled at all but had the pleasure of talking to Stan one time some years back via phone (he offered to buy me a diy Karlson) and he related the way he uses (and re-voices) 604 - -say "hi" to Stan from me and tell him to work on that Milt Hinton grip