Maybe.
Possibly.
To be honest, it depends on the cap
For other types, e.g. large shunt caps on a woofer leg, the ESR is arguably a bit less of an issue. Although it's not necessarily going to bring you many gains by replacing with an MKP either -chances are it's there because it's a lot smaller, cheaper and the cost / benefit ratio didn't justify film caps.
Thank you for the prompt answer.Well,one crucial advantage of film caps is their stability over time.Back in 1978 I bougt a pair of Jim Rogers JR 149s,which still accompany me.20 years later I replaced the Elcaps in the cross with Philips MKTs.I kept the elcaps and today they mesure ten times their nominal capacity.I think that,even if I had used electros,I could not restore the original voicing of the loudspeaker (which I had forgotten after all these years),since newer caps would have lower ESR.I asked because some people here are against replacing cheapish electro caps with film caps in a loudspeaker bought secondhand "because this will alter the parameters and the crossover circuit should be redesigned".I wonder if designers of relatively low cost loudspeakers go to such lengths,(i.e. taking ESR under considfration)when they have to work with cheap components.
Last edited:
Stability is one of the major advantages of film / F&F capacitors over electrolytics. Arguably the major one. However, it is also worth remembering that the majority of loudspeakers are not expected to last decades, so a degree of servicing every 10 years or so (which most decent electrolytics will do without major deviations) is not particularly unreasonable, whether it be the filter or addressing something else e.g. a perished gasket.
Old speakers that were cheap when new? Unlikely the designer factored ESR in, except sometimes in the case of tweeter caps. Most that were considered 'reasonable' or better? Almost invariable that they would, the JR149 being a classic example. This was bread & butter engineering for the period, some of which has been overlooked since the widespread availability of MKT and MKP type capacitors of moderate price, and it wasn't / isn't especially difficult to factor into the transfer functions. It has to be kept in mind that, with certain exceptions, driver & component quality is often higher now than it was in, say, the 1970s, so engineering was king -this sort of thing had to be considered, because there wasn't any choice in the matter.
Old speakers that were cheap when new? Unlikely the designer factored ESR in, except sometimes in the case of tweeter caps. Most that were considered 'reasonable' or better? Almost invariable that they would, the JR149 being a classic example. This was bread & butter engineering for the period, some of which has been overlooked since the widespread availability of MKT and MKP type capacitors of moderate price, and it wasn't / isn't especially difficult to factor into the transfer functions. It has to be kept in mind that, with certain exceptions, driver & component quality is often higher now than it was in, say, the 1970s, so engineering was king -this sort of thing had to be considered, because there wasn't any choice in the matter.
Not a problem. It's not an argument though, just, broadly speaking, a statement of fact. For example, BBC & most BBC type designs (e.g. Spendor, the JR149) accounted for [new] electrolytic ESR in the crossover design; this is a matter of historical record. KEF & B&W were the same; ditto for companies like Ferrograph, IMF et al. Same for JBL who were also one of the prime drivers during the 1970s - early '80s of shunting electrolytics with low value film caps; whatever stance one might have on that particular matter, you couldn't accuse them of lacking attention to detail with components. Naturally, as noted, at the cheap end there would be less interest, but quality designs accounted for as much as they could.
Apologies, but is it ok, alright, for total rookie/novice to piggyback on this?
A SPECIFIC EXAMPLE?
Reason is, month ago, I'd posted something vaguely similar, but my question(s) [different thread] were so random/diffuse that they were addressed, but not fully - in the particulars.
My bad - entirely my fault. No biggie. I'm LEARNING (I think Lol!) ... a little, e.g., to solder [not so hard] and de-solder [much harder! i.e., in close-confines tight OEM PCB]. Shoulda been warned Lol! "Spend $4.99 on solder-sucker or PAY consequences of 'wicking' up solder/tears with tissue/braid."
THAT'S HOW I'D ARRIVED AT:
May I offer an example re-cap, a BIT of a "problem," as things are not going to be as quick/easy/cheap as ... well, as one might suspect, and possibly, perhaps, explore resolution in [excruciating Lol!] detail?
I have these 1998/99 speakers, JBL, list about $600. US dollars. ... That was list/pr or around $300ea 20+ years ago.
JBL HLS 820 ... and ready for re-capping:
Radial N.P. electrolytic
50V 22μF ±10%
50V 43μF
Axial N.P. film
50V 2.7μF
Diagram [JBL Technical Manual] in case there's any doubts:
http://oldaudio.by/media/kunena/attachments/56/jbl_hls-820_250w_speaker_system_sm.pdf
and here's again likewise [previously posted last month] an image of one of the networks:
Between me, and my [new Lol!] g/f, OVER THE YEARS, we've collected 1/2-dozen pairs of "big old stereo speakers" w/ "good" cones/coils/surrounds (and "presumably bad" caps.)
These JBL are [very likely] the best-of-the-lot (we've been keeping basement/garage/lvg rm):
Been THINKING about maybe buying an ESR meter; Peak ESR70 Atlas ESR Plus Capacitor Analyser. Not sure if it's going to accomplish ...
Anyhow, and this is where any/all this (might?!) start to "matter."
The 43uF caps are discontinued. They're obsolete.
So here's where I got "stuck," and asked questions [here at diyAudio.com] and these are roughly the answers I got (pretty much verbatim) ... not starting a whole new thread and rather piggybacking because (guessing) this is "just about what the average Rookie" could run into:
1st, Rookie was told, "Film caps are inappropriate as replacements for the 22uF and 43uF [electrolytics]."
a) So was advised (again, diyAudio) to "Replace like for like," which is entirely consistent with what Scottmoose has said.
b) Was left implicit, I believe, to "Leave the other (film 2.7uF) alone, and leave other components be," as well, which makes perfect sense.
c) Was NOT told "why" Film caps are inappropriate replacements, but is likely inferred that at 2X's or 3X's the diameter, securing Film caps to the board ... well, the PCB wouldn't conveniently accommodate, size-wise, such largER caps, tho they [crossovers re-built with film] MIGHT fit the cabinet (opening)? ... still the matter of SECURING them (and any components in general); consequently the return-on-investment isn't worth the complication. That's not definitive, it's inference, and I'm 99% ok making it ... (very likely) all good.
2nd, Since 43uF caps are discontinued, was suggested (roughly) "Use 47uF, and at +/-5% from OEM, likely you're OK."
a) So, I'd also considered/suggested "parallel a 30uF + 13uF" in order to "= a 43uF," but again, with the chirping crickets, figured, isn't worth the complication.
b) That's where diyAudio forum members [w/ level(s) of expertise well-beyond anything Rookie might aspire, Lol!] basically left things: use 47uF instead of 43uF OEM.
c) Naturally, my "other questions" (embarrassing and/or not-well-posed and even diffuse in-the-extreme) may have left the experts (understandably) a little annoyed and/or very bored. Rookie did a TERRIBLE job of posing half-way-decent interrogatives.
3rd, So, I did some more reading.
a) I wanted the "right" values for my re-cap.
b) Who wouldn't?
c) but WHY?
4th, Here's some of what I read, elsewhere, and HOPING that maybe some of this is sorted out:
a) As with any other components being replaced or when building a set of speakers, the tolerance doesn't matter (as long as you aren't way off) as much as matching ... I have heard tolerance is based on batching which won't help with singular caps ... To get good stereo imaging, the speakers have to match ... A small meter that can measure capacitance will be better than nothing here. Bench measuring with load even better so you measure the whole crossover at one time to match the other ... So in ways tolerance does matter but not as much as component matching unless they are way off ... However, simpler networks are more tolerant ... The higher the "order" of the crossover, the more critical the tolerance. Parts Express has a nice, affordable crossover design program that will let you play with the values and see the effect it has on the filter's response, if you want to get into the bug dust.
b) [Nothing quantitative there; other posters at that forum offered mainly anecdotal 1% tolerance "goals" so it's not taken as liturgical truth ... tho the 3rd order Butterworth? mine? Rookie's? "Perhaps suggests the more critical (higher 3rd?) tolerance...? All that's speculative.]
c) Then I read this: ... That's one small drawback to the internet now, versus the old days when you could go to a parts shop and fish through the bins for what you wanted, but I can live with that ...
d) [Maybe that's misleading. Rookie WENT to the (single extant in major metro area) parts shop!]
e) Rookie FOUND a "pair" of so-called "matching" [nominal] 47uF (250V this time) that measured right-about 44.2uF and 43.9uF, i.e., found a pair (of essentially 44uF) to "replace" the 43uF. Or about 2% from spec. Fair?
5th, Rookie said he didn't own an ESR meter. True.
a) So, he borrowed the one at the Shop.
b) Wouldn't Rookie, then, naturally, beings he carried to Shop the pair of crossover networks with him [remember, this is all happening-pretty-fast] ask Shop Owner (or staff) if "we" could ALSO measure the capacitance of the (extant) capacitors on the (extant) network(s)?
6th, So Rookie did that, or rather, Shop manager was getting a little impatient, and extant caps showing no visible damage, only one (left-hand for argument's sake) nominal 43uF actually WAS measured at roughly 51uF -- while still in-circuit naturally.
a) ran out of time to check the others.
b) Rookie "Still has/had no idea whatsoever" where any of this is going.
c) i.e., don't we check for defects before R&R?
d) it's still obvious, however, nominal 43uF measuring 51uF means at least one crossover component is well out-of-spec. Naturally, one then wonders (we bought the speakers used last year) how
e) audible IS that? if indeed it is audible degradation, then
f) PRIOR to re-capping, could be ILLUSTRATIVE and one seriously considers if indeed
g) one might detect (somehow) and/or (less likely) measure the before & after [recap] "differences," the verifiable differential
h) beings it might be illuminating.
i) could ask similar questions about ESR variances, and XO "slope?"
j) however, ESR isn't really a variable under consideration (if like-for-like).
7th Somewhere on the diyAudio forum, Rookie remembers reading you could (theoretically?) "calculate" the shift in XO frequency due a -/+5% cap value and (from memory -- can't find post) it was a "sq.rt. LC" type calculation and represented a shift in crossover frequency of around 2.5%.
a) PROVIDED "Any of that type calculation is true, even in terms of orders of magnitude?" then, if these JBL HLS 820 were spec'd a crossover of 2600Hz, that'd shift (upwards?) to 2665Hz.
b) see where this is going?
c) Does a "marginal" shift in actual (not nominal) capacitance become either
d) XO frequency calculable, or
e) measurable, or maybe most importantly
f) audibly discernible?
g) I'd read somewhere, possibly, that: f) audible? depends highly upon (in this case) tweeter resonance frequency relative to XO frequency.
8th Yet, supposing I'd done "blindly" (relatively) as suggested and replaced a 43uF with a 50uF cap (what some of the nominal 47's in the bin actually measured)? Forget about "matching" for now,
a) IOW, Still curious, what's the "effect on XO freq" of shifting a cap value "up" aprox 14%? Even if it's just thought-experiment.
b) Recalling, one guy had claimed, Parts Express has a nice, affordable crossover design program that will let you play with the values and see the effect it has on the filter's response, if you want to get into the bug dust ...
c) Rookie has no way knowing anything "crossover design program" even if he has tinkered briefly with various crossover calculators so-called.
So Rookie can't help but wonder about (numerical?) VALUE in that "relatively exact capacitance?"
Can anything be LEARNED (making measurements) from re-capping when (if even necessary?) comparing "old" metrics to "new?" Does anybody (naturally everybody "sound" checks!) actually "electronically test" performance of refurbished, i.e, re-capped speakers?
Both before, and after?
Or, as "99% what I've read" suggests, most (novices) just (relatively blindly) re-cap, old electrolytics, then say to themselves, "Sounds great," and leave-it-at-that? I won't ask about before-after "performance" testing, if that's what's called, IOW driver's response to signal(s), as I see there's 2-D and 3-D and multidimensional graphs galore. But sometimes, with calibrated mic perhaps? IOW maybe (big maybe) there ARE some "relatively simple cost-effective tests" besides favorite CD cello, hard rock, symphony, 1W, 10W, 50W, et. al. The complexity (at the forums, especially diyAudi e.g, seems incredible) but MAYBE there's a "dummies guide" to "down and dirty numerical assessment" of speakers' response, namely, "before & after re-cap." Golly that'd be great, but every time I look for simplicity things tend to wade into growing complexity - OR - some (relative) subjectivity. In the dark a little on this one Lol!
I realize this is a very long, and relatively diffuse post. Still, it's a "real world" ENTRY LEVEL hobbyist-type inquiry. Comes from real-life, and involves actual metrics, mostly my own, extreme naive novice-as-they-may-be. You'll laugh, but I think we forgot to discharge the 43uF cap (the lone cap tested) in-circuit when we measured 50 or 51uF. It wasn't my Shop, wasn't my ESR, but if I had it to do all over again, I'd have put the ESR meter to ALL the caps on BOTH boards (before disassembly). I'd at least have some numerical values to "how out of whack" things might be. The Shop "tech" (a salesperson really) claimed he "couldn't get a reading" on one of the caps on the board. Then he (and consequently "we") gave up. Hand waving. I don't know why that might be. He didn't seem to think it was worth pursuing. I don't know if anybody does Lol!
Seems strange, but honestly, I'm trying like Hades to ask "practical questions," especially to things I've not been fortunate enough to locate answers, in practice. How does one KNOW the practicality of re-capping OBJECTIVELY? Rookie just hasn't come across a single "measured" result of re-cap. Might be millions of them out there, and I'm just ignorant of how to locate a few.
"Generally speaking for such speakers you're probably better off replacing like-for-like rather than simply aiming for a [theoretically] 'better' type of capacitor, unless you can add the appropriate amount of series R," has said Scottmoose, which makes perfect sense, but UNLESS you TEST the like-for-like, how do you KNOW - the parameters - of what you're doing? And since I've no earthly idea, is there ZERO VALUE in testing caps (on the boards, in-circuit) BEFORE re-capping? Just do things "on suspicion they've drifted from spec?" Maybe test all the components before re-cap? Why, or why not? The Rookie tends to get hung on "what or what not" Yes/No type questions ... Why-or-why-not questions seem troublesome. Yet need they be?
Apologies in advance, but I've seen people (facetiously) suggest "Buy (purchase) 50 electrolytic and use ONLY the ones you test!" I'm not moralizing, but "Internet" sometimes frames (shapes) thought process in desultory manner. In my google searches of "tolerances and electrolytic caps," and the hundreds of posts I'd read, not one person suggested, "Go to the bins, in the Shop, with a meter." I only found the counter-suggestion, the negative, which doesn't mean anything, except I'm not afraid to think counter-intuitively, like 7 billion others Lol!
So, consequently, 3 people on the Internet say, "Versus the old days when you could go to a parts shop," and suddenly nobody goes to an Electronics Parts shop anymore, EVER, when sometimes (maybe? possibly?) it's exactly the RIGHT thing to do? I know, that's hyperbole, to some extent, but we don't "live and operate" in forums, and can (sometimes) be very difficult for "purveyor" and "recipient" of information to get to same wavelength. As Scottmoose has kindly suggested, and I almost hate to suggest this, "cost / benefit ratio didn't justify..." getting along "the same wavelength." Us dumb guys (me, the Rookie) wanna know WHY the smart guys do XYZ Lol! Rookie also wants to know why not do XYZ Lol!
I know this is a lousy miserable post of mine, terribly naive; but ultimately only somebody w/ some slight time/inclination and considerable expertise could point me in direction of the 2 or 3 "expert tips or pointers" that'd HELP to START to resolve my own vagaries which is really about the VALUE (and numerical or rather quantifiable value - if feasible - in a few hours time or whatnot) DISCERNIBLE. Rather, discernible value in doing XYZ...
Audible testing seems to be a whole nother perhaps distinct bundle of wax.
Could have posed the whole question differently: which has greater potential weight; the tolerances (parameters) of (electrolytic?) caps, or the ESR values? Even that's probably not well-posed. But in some small respect, delimiting the value (going to the bins, in the shop?) of the tolerance(s) is one way of restricting the one variable? What good's delimiting one w/o the other, rhetorically?
Good heavens I'm gonna regret posting this, I can feel it in my bones, but hope against hope, hope springs eternal ... my heart's in the right place, even IF my brain's on vacation (thumb in his butt and mind in Arkansas)
Glad Tidings! I resign my post (no not that post, the other post...
PS: Ah, I'd forgot; "measurement testing" of individual drivers might be fairly straightforward, along some parameters, yet, contrariwise, assembled speakers might be very difficult to do "performance" testing barring the anechoic chamber or some-such-thing? So that in principle, even in practice, very few have the capacity to do valuable (realistically speaking) metrics on refurbished re-caps, w/r/t assembled multi-way speakers, aside from listening to passages of audio recordings, i.e., music? And maybe that's why not many REPORT? Beings it's just too difficult, aside from the corporate professionals?; sorry for the segue, but it's easy to get "lost" in electronics ... So I'm inclined ... suspecting ... that when comes to "performance measurements" in the real world (and esp. for Rookie) answers may be "No, no, and rarely if ever" Lol! ... That'd explain why "measured sound testing" w/r/t refurbish is seldom discussed. Duh.
A SPECIFIC EXAMPLE?
Reason is, month ago, I'd posted something vaguely similar, but my question(s) [different thread] were so random/diffuse that they were addressed, but not fully - in the particulars.
My bad - entirely my fault. No biggie. I'm LEARNING (I think Lol!) ... a little, e.g., to solder [not so hard] and de-solder [much harder! i.e., in close-confines tight OEM PCB]. Shoulda been warned Lol! "Spend $4.99 on solder-sucker or PAY consequences of 'wicking' up solder/tears with tissue/braid."
THAT'S HOW I'D ARRIVED AT:
May I offer an example re-cap, a BIT of a "problem," as things are not going to be as quick/easy/cheap as ... well, as one might suspect, and possibly, perhaps, explore resolution in [excruciating Lol!] detail?
I have these 1998/99 speakers, JBL, list about $600. US dollars. ... That was list/pr or around $300ea 20+ years ago.
JBL HLS 820 ... and ready for re-capping:
Radial N.P. electrolytic
50V 22μF ±10%
50V 43μF
Axial N.P. film
50V 2.7μF
Diagram [JBL Technical Manual] in case there's any doubts:
http://oldaudio.by/media/kunena/attachments/56/jbl_hls-820_250w_speaker_system_sm.pdf
and here's again likewise [previously posted last month] an image of one of the networks:
Between me, and my [new Lol!] g/f, OVER THE YEARS, we've collected 1/2-dozen pairs of "big old stereo speakers" w/ "good" cones/coils/surrounds (and "presumably bad" caps.)
These JBL are [very likely] the best-of-the-lot (we've been keeping basement/garage/lvg rm):
Been THINKING about maybe buying an ESR meter; Peak ESR70 Atlas ESR Plus Capacitor Analyser. Not sure if it's going to accomplish ...
Anyhow, and this is where any/all this (might?!) start to "matter."
The 43uF caps are discontinued. They're obsolete.
So here's where I got "stuck," and asked questions [here at diyAudio.com] and these are roughly the answers I got (pretty much verbatim) ... not starting a whole new thread and rather piggybacking because (guessing) this is "just about what the average Rookie" could run into:
1st, Rookie was told, "Film caps are inappropriate as replacements for the 22uF and 43uF [electrolytics]."
a) So was advised (again, diyAudio) to "Replace like for like," which is entirely consistent with what Scottmoose has said.
b) Was left implicit, I believe, to "Leave the other (film 2.7uF) alone, and leave other components be," as well, which makes perfect sense.
c) Was NOT told "why" Film caps are inappropriate replacements, but is likely inferred that at 2X's or 3X's the diameter, securing Film caps to the board ... well, the PCB wouldn't conveniently accommodate, size-wise, such largER caps, tho they [crossovers re-built with film] MIGHT fit the cabinet (opening)? ... still the matter of SECURING them (and any components in general); consequently the return-on-investment isn't worth the complication. That's not definitive, it's inference, and I'm 99% ok making it ... (very likely) all good.
2nd, Since 43uF caps are discontinued, was suggested (roughly) "Use 47uF, and at +/-5% from OEM, likely you're OK."
a) So, I'd also considered/suggested "parallel a 30uF + 13uF" in order to "= a 43uF," but again, with the chirping crickets, figured, isn't worth the complication.
b) That's where diyAudio forum members [w/ level(s) of expertise well-beyond anything Rookie might aspire, Lol!] basically left things: use 47uF instead of 43uF OEM.
c) Naturally, my "other questions" (embarrassing and/or not-well-posed and even diffuse in-the-extreme) may have left the experts (understandably) a little annoyed and/or very bored. Rookie did a TERRIBLE job of posing half-way-decent interrogatives.
3rd, So, I did some more reading.
a) I wanted the "right" values for my re-cap.
b) Who wouldn't?
c) but WHY?
4th, Here's some of what I read, elsewhere, and HOPING that maybe some of this is sorted out:
a) As with any other components being replaced or when building a set of speakers, the tolerance doesn't matter (as long as you aren't way off) as much as matching ... I have heard tolerance is based on batching which won't help with singular caps ... To get good stereo imaging, the speakers have to match ... A small meter that can measure capacitance will be better than nothing here. Bench measuring with load even better so you measure the whole crossover at one time to match the other ... So in ways tolerance does matter but not as much as component matching unless they are way off ... However, simpler networks are more tolerant ... The higher the "order" of the crossover, the more critical the tolerance. Parts Express has a nice, affordable crossover design program that will let you play with the values and see the effect it has on the filter's response, if you want to get into the bug dust.
b) [Nothing quantitative there; other posters at that forum offered mainly anecdotal 1% tolerance "goals" so it's not taken as liturgical truth ... tho the 3rd order Butterworth? mine? Rookie's? "Perhaps suggests the more critical (higher 3rd?) tolerance...? All that's speculative.]
c) Then I read this: ... That's one small drawback to the internet now, versus the old days when you could go to a parts shop and fish through the bins for what you wanted, but I can live with that ...
d) [Maybe that's misleading. Rookie WENT to the (single extant in major metro area) parts shop!]
e) Rookie FOUND a "pair" of so-called "matching" [nominal] 47uF (250V this time) that measured right-about 44.2uF and 43.9uF, i.e., found a pair (of essentially 44uF) to "replace" the 43uF. Or about 2% from spec. Fair?
5th, Rookie said he didn't own an ESR meter. True.
a) So, he borrowed the one at the Shop.
b) Wouldn't Rookie, then, naturally, beings he carried to Shop the pair of crossover networks with him [remember, this is all happening-pretty-fast] ask Shop Owner (or staff) if "we" could ALSO measure the capacitance of the (extant) capacitors on the (extant) network(s)?
6th, So Rookie did that, or rather, Shop manager was getting a little impatient, and extant caps showing no visible damage, only one (left-hand for argument's sake) nominal 43uF actually WAS measured at roughly 51uF -- while still in-circuit naturally.
a) ran out of time to check the others.
b) Rookie "Still has/had no idea whatsoever" where any of this is going.
c) i.e., don't we check for defects before R&R?
d) it's still obvious, however, nominal 43uF measuring 51uF means at least one crossover component is well out-of-spec. Naturally, one then wonders (we bought the speakers used last year) how
e) audible IS that? if indeed it is audible degradation, then
f) PRIOR to re-capping, could be ILLUSTRATIVE and one seriously considers if indeed
g) one might detect (somehow) and/or (less likely) measure the before & after [recap] "differences," the verifiable differential
h) beings it might be illuminating.
i) could ask similar questions about ESR variances, and XO "slope?"
j) however, ESR isn't really a variable under consideration (if like-for-like).
7th Somewhere on the diyAudio forum, Rookie remembers reading you could (theoretically?) "calculate" the shift in XO frequency due a -/+5% cap value and (from memory -- can't find post) it was a "sq.rt. LC" type calculation and represented a shift in crossover frequency of around 2.5%.
a) PROVIDED "Any of that type calculation is true, even in terms of orders of magnitude?" then, if these JBL HLS 820 were spec'd a crossover of 2600Hz, that'd shift (upwards?) to 2665Hz.
b) see where this is going?
c) Does a "marginal" shift in actual (not nominal) capacitance become either
d) XO frequency calculable, or
e) measurable, or maybe most importantly
f) audibly discernible?
g) I'd read somewhere, possibly, that: f) audible? depends highly upon (in this case) tweeter resonance frequency relative to XO frequency.
8th Yet, supposing I'd done "blindly" (relatively) as suggested and replaced a 43uF with a 50uF cap (what some of the nominal 47's in the bin actually measured)? Forget about "matching" for now,
a) IOW, Still curious, what's the "effect on XO freq" of shifting a cap value "up" aprox 14%? Even if it's just thought-experiment.
b) Recalling, one guy had claimed, Parts Express has a nice, affordable crossover design program that will let you play with the values and see the effect it has on the filter's response, if you want to get into the bug dust ...
c) Rookie has no way knowing anything "crossover design program" even if he has tinkered briefly with various crossover calculators so-called.
So Rookie can't help but wonder about (numerical?) VALUE in that "relatively exact capacitance?"
Can anything be LEARNED (making measurements) from re-capping when (if even necessary?) comparing "old" metrics to "new?" Does anybody (naturally everybody "sound" checks!) actually "electronically test" performance of refurbished, i.e, re-capped speakers?
Both before, and after?
Or, as "99% what I've read" suggests, most (novices) just (relatively blindly) re-cap, old electrolytics, then say to themselves, "Sounds great," and leave-it-at-that? I won't ask about before-after "performance" testing, if that's what's called, IOW driver's response to signal(s), as I see there's 2-D and 3-D and multidimensional graphs galore. But sometimes, with calibrated mic perhaps? IOW maybe (big maybe) there ARE some "relatively simple cost-effective tests" besides favorite CD cello, hard rock, symphony, 1W, 10W, 50W, et. al. The complexity (at the forums, especially diyAudi e.g, seems incredible) but MAYBE there's a "dummies guide" to "down and dirty numerical assessment" of speakers' response, namely, "before & after re-cap." Golly that'd be great, but every time I look for simplicity things tend to wade into growing complexity - OR - some (relative) subjectivity. In the dark a little on this one Lol!
I realize this is a very long, and relatively diffuse post. Still, it's a "real world" ENTRY LEVEL hobbyist-type inquiry. Comes from real-life, and involves actual metrics, mostly my own, extreme naive novice-as-they-may-be. You'll laugh, but I think we forgot to discharge the 43uF cap (the lone cap tested) in-circuit when we measured 50 or 51uF. It wasn't my Shop, wasn't my ESR, but if I had it to do all over again, I'd have put the ESR meter to ALL the caps on BOTH boards (before disassembly). I'd at least have some numerical values to "how out of whack" things might be. The Shop "tech" (a salesperson really) claimed he "couldn't get a reading" on one of the caps on the board. Then he (and consequently "we") gave up. Hand waving. I don't know why that might be. He didn't seem to think it was worth pursuing. I don't know if anybody does Lol!
Seems strange, but honestly, I'm trying like Hades to ask "practical questions," especially to things I've not been fortunate enough to locate answers, in practice. How does one KNOW the practicality of re-capping OBJECTIVELY? Rookie just hasn't come across a single "measured" result of re-cap. Might be millions of them out there, and I'm just ignorant of how to locate a few.
"Generally speaking for such speakers you're probably better off replacing like-for-like rather than simply aiming for a [theoretically] 'better' type of capacitor, unless you can add the appropriate amount of series R," has said Scottmoose, which makes perfect sense, but UNLESS you TEST the like-for-like, how do you KNOW - the parameters - of what you're doing? And since I've no earthly idea, is there ZERO VALUE in testing caps (on the boards, in-circuit) BEFORE re-capping? Just do things "on suspicion they've drifted from spec?" Maybe test all the components before re-cap? Why, or why not? The Rookie tends to get hung on "what or what not" Yes/No type questions ... Why-or-why-not questions seem troublesome. Yet need they be?
Apologies in advance, but I've seen people (facetiously) suggest "Buy (purchase) 50 electrolytic and use ONLY the ones you test!" I'm not moralizing, but "Internet" sometimes frames (shapes) thought process in desultory manner. In my google searches of "tolerances and electrolytic caps," and the hundreds of posts I'd read, not one person suggested, "Go to the bins, in the Shop, with a meter." I only found the counter-suggestion, the negative, which doesn't mean anything, except I'm not afraid to think counter-intuitively, like 7 billion others Lol!
So, consequently, 3 people on the Internet say, "Versus the old days when you could go to a parts shop," and suddenly nobody goes to an Electronics Parts shop anymore, EVER, when sometimes (maybe? possibly?) it's exactly the RIGHT thing to do? I know, that's hyperbole, to some extent, but we don't "live and operate" in forums, and can (sometimes) be very difficult for "purveyor" and "recipient" of information to get to same wavelength. As Scottmoose has kindly suggested, and I almost hate to suggest this, "cost / benefit ratio didn't justify..." getting along "the same wavelength." Us dumb guys (me, the Rookie) wanna know WHY the smart guys do XYZ Lol! Rookie also wants to know why not do XYZ Lol!
I know this is a lousy miserable post of mine, terribly naive; but ultimately only somebody w/ some slight time/inclination and considerable expertise could point me in direction of the 2 or 3 "expert tips or pointers" that'd HELP to START to resolve my own vagaries which is really about the VALUE (and numerical or rather quantifiable value - if feasible - in a few hours time or whatnot) DISCERNIBLE. Rather, discernible value in doing XYZ...
Audible testing seems to be a whole nother perhaps distinct bundle of wax.
Could have posed the whole question differently: which has greater potential weight; the tolerances (parameters) of (electrolytic?) caps, or the ESR values? Even that's probably not well-posed. But in some small respect, delimiting the value (going to the bins, in the shop?) of the tolerance(s) is one way of restricting the one variable? What good's delimiting one w/o the other, rhetorically?
Good heavens I'm gonna regret posting this, I can feel it in my bones, but hope against hope, hope springs eternal ... my heart's in the right place, even IF my brain's on vacation (thumb in his butt and mind in Arkansas)
Glad Tidings! I resign my post (no not that post, the other post...
PS: Ah, I'd forgot; "measurement testing" of individual drivers might be fairly straightforward, along some parameters, yet, contrariwise, assembled speakers might be very difficult to do "performance" testing barring the anechoic chamber or some-such-thing? So that in principle, even in practice, very few have the capacity to do valuable (realistically speaking) metrics on refurbished re-caps, w/r/t assembled multi-way speakers, aside from listening to passages of audio recordings, i.e., music? And maybe that's why not many REPORT? Beings it's just too difficult, aside from the corporate professionals?; sorry for the segue, but it's easy to get "lost" in electronics ... So I'm inclined ... suspecting ... that when comes to "performance measurements" in the real world (and esp. for Rookie) answers may be "No, no, and rarely if ever" Lol! ... That'd explain why "measured sound testing" w/r/t refurbish is seldom discussed. Duh.
Last edited:
Kimon:
It depends. If the cap is in series with a driver, it's effects are usually benign. At worse, you'll end up with a brighter tweeter or louder midrange.
If it's a shunt (going to ground) and you go from 2 Ohms to say 0.5 Ohms, it can really affect the sound in unexpected ways, as the impedance curve can be altered, dropping the impedance above the filter point, and leading to a depressed FR which most simulators won't capture.
Small changes (0.5 to 0.2 Ohms) are usually benign, but anything approaching an Ohm, especially in shunts, should be avoided.
Best,
E
It depends. If the cap is in series with a driver, it's effects are usually benign. At worse, you'll end up with a brighter tweeter or louder midrange.
If it's a shunt (going to ground) and you go from 2 Ohms to say 0.5 Ohms, it can really affect the sound in unexpected ways, as the impedance curve can be altered, dropping the impedance above the filter point, and leading to a depressed FR which most simulators won't capture.
Small changes (0.5 to 0.2 Ohms) are usually benign, but anything approaching an Ohm, especially in shunts, should be avoided.
Best,
E
If I may, variations in series R (as in, in series with the driver) is not automatically benign, and at worst, you won't just end up with 'a brighter tweeter or louder midrange' -level changes can also throw out the transfer functions & mess up phasing / response around the XO frequency into the bargain, with potentially increased distortion also if for e.g. more power is getting into the tweeter. The effects can also can vary with circuit topology, beyond a mere change in level. If you have, say, a 2nd order electrical high pass then level variation is the major effect, but with something like a 3rd order, it becomes more complicated as you have the potential for a shift in impedance altering the transfer functions. For e.g., the attached. This is a 3rd order high pass. Basic electrical sim only for the purpose of clarity. Grey assumes 2ohms series R added upstream of the filter (i.e. 2ohms ESR for the first cap). Blue assumes 2ohms series R added downstream of the filter (i.e. 2ohms ESR for the second series cap).
Resistance in shunts should be accounted for in any quality crossover design (whether ESR or specific components) just as much as in series components and any decent software modeller can account for / incorporate it. If it doesn't, you need better software.
Resistance in shunts should be accounted for in any quality crossover design (whether ESR or specific components) just as much as in series components and any decent software modeller can account for / incorporate it. If it doesn't, you need better software.
Attachments
Ah, apologies.
I did see ttps://www.diyaudioandvideo.com/Tutorial/SpeakerResponseTesting/
that apparently allows some testing of assembled speakers.
Maybe "diagnosing" old XO's is just "too much" as-they-say.
There's one (very definite) Q that I did have; the glue on the cap(s) drizzles over-and-onto one of the leads on the inductor coil; very tight space and some serious concern razor knife might sever ... anyhow is it ever used zylene or some such to dissolve glue from caps ? - or could that damage the PCB?
Sorry for the crazy segue on the other post. There's just so much detail in electronics.
I did see ttps://www.diyaudioandvideo.com/Tutorial/SpeakerResponseTesting/
that apparently allows some testing of assembled speakers.
Maybe "diagnosing" old XO's is just "too much" as-they-say.
There's one (very definite) Q that I did have; the glue on the cap(s) drizzles over-and-onto one of the leads on the inductor coil; very tight space and some serious concern razor knife might sever ... anyhow is it ever used zylene or some such to dissolve glue from caps ? - or could that damage the PCB?
Sorry for the crazy segue on the other post. There's just so much detail in electronics.
Not crazy, but most don't have time to read the forum equivalent of War and Peace so if you have a specific question or questions, it's usually a good idea to keep it a bit shorter / more immediately obvious. Most people don't have unlimited leisure time, and frankly, I for one took one look at the above and thought 'I don't have the energy to wade through all that'.
Re solvents, try to avoid unless you can be very specific in applying. It may well be fine, but it depends on the PCB material, and if you don't know exactly what was used (which most of the time you probably won't) it's probably better not to take the risk unless you're OK with a total rebuild.
so if you have a specific question or questions, it's usually a good idea to keep it a bit shorter / more immediately obvious. Most people don't have unlimited leisure time, and frankly, I for one took one look at the above and thought 'I don't have the energy to wade through all that'.Re solvents, try to avoid unless you can be very specific in applying. It may well be fine, but it depends on the PCB material, and if you don't know exactly what was used (which most of the time you probably won't) it's probably better not to take the risk unless you're OK with a total rebuild.
Not crazy, but most don't have time to read the forum equivalent of War and Peace
Re solvents, try to avoid unless you can be very specific in applying. It may well be fine, but it depends on the PCB material, and if you don't know exactly what was used (which most of the time you probably won't) it's probably better not to take the risk unless you're OK with a total rebuild.
You. Or yoos. war & peace or yadda yadda. clever style contest?
Nobody said anything yagga yadda about reading for comprehension.
No need er, uh nothing personal. My leisure time oughtta be unlimited else maybe it's wrapped too tight.
So if it's heavy lift then get paid. can't be that rueful. Can't possibly be that serious. Don't take it too serious. easy enough to take another long look and think whatever ya feel.
Unless yoos to givin a bad vibe. Nobody ta now had offered anything snide. Forget about the solvent. Or have a whiff. Whatever. Can't speak for others.
Low energy, then skip it ... seems free to choose. Anything along those lines. no credence in solving anything that could be clipped. No credence from yoos.
No, you wouldn't know exactly, most of the time, unless you knew better. Maybe. Not believable.
yadda, don't overthink yer skivvies-no value in it. no need and butter anybody's butt. Least i wouln't Wan't speak for "most people," regardless, and seldom do
Easily ask elsewhere, there's nothing of value here. Valid point about goin to bins, likely, and glad I did.
Nothing ventured really. Nominal gain I spse w/r/t speak for others. Avoid em like the plague - no sense of humor lol and if leisure time comes from such a premium sounds like hard work. It's not so hard to see -- what opinions and ar -holes have in common. They say everbody's got one and yer free to share yers w/ whomever you please
Last edited:
.Ask why.
Get told they get large and expensive with larger values.
Decide to make up your own mind.
In time, also learn that sometimes there is no audible difference.
Your loss, honestly, but I'm not going to argue.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.