replacing ecap w/film cap on speaker crossover
- Thread starter oldman55
- Start date
Register to hide this ad
I would think so. Caps in crossover network determine the crossover frequencies, that means they use their reactance to form voltage dividers. Any voltage coef of the cap will affect the linearity of the signal around the crossover frequencies and create distortion.
Another thing is the ESR and ESL of the electrolytics, Remember the crossover is a very low impedance circuit ( 4 to 8ohm). Any series resistance and inductance with become significant and affect the performance of the network and create distortion.
I have done extensive experiment with THD vs speaker cable. My conclusion is the series resistance and inductance are the main cause of rising THD on the speaker side. You use caps with high ESR and ESL will do exactly the same as using smaller and longer cable.
I would use two or three film caps in parallel to further lowering the ESR aned ESL even with film cap.
Another thing is the ESR and ESL of the electrolytics, Remember the crossover is a very low impedance circuit ( 4 to 8ohm). Any series resistance and inductance with become significant and affect the performance of the network and create distortion.
I have done extensive experiment with THD vs speaker cable. My conclusion is the series resistance and inductance are the main cause of rising THD on the speaker side. You use caps with high ESR and ESL will do exactly the same as using smaller and longer cable.
I would use two or three film caps in parallel to further lowering the ESR aned ESL even with film cap.
oldman55
Well-Known Member
So if I need 2.5uf, you are saying that it is better to put a couple of 1 uf and a .5 uf cap in parallel vs using a single 2.5 uf cap (if I could find one)?
And it is OK to use a poly audio cap in lieu of the non polar ecap?
I don't want to give specific advice beyond this as I have not design a crossover. I just use the knowledge of electronics to give the suggestion.
Yes, I would use two 1uF and a 0.5uF in parallel to get 2.5uF.
I definitely would TRY to use film caps than non polar ecap.......BUT all depends on how much it cost. You you have a 100uF cap, it's not practical to try to replace with film. I would use say a use 3 33uF ecap in parallel to get 100uF. And even add a 1uF film cap to make up 100uF.
Yes, I would use two 1uF and a 0.5uF in parallel to get 2.5uF.
I definitely would TRY to use film caps than non polar ecap.......BUT all depends on how much it cost. You you have a 100uF cap, it's not practical to try to replace with film. I would use say a use 3 33uF ecap in parallel to get 100uF. And even add a 1uF film cap to make up 100uF.
In practical terms, yes replacing a bi-polar electrolytic with poly in a speaker crossover will make a difference. It is considered an improvement which should increase clarity in the tweeter and is a very common upgrade. To be clear, we are talking about the cap that is in series with a tweeter or mid-range driver. Changing any of the caps in parallel with any drivers (to poly) would likely not reveal any audible change.
Another factor is that a tweeter, or other driver, will likely be slightly louder because the poly cap has lower ESR than the electrolytic it replaced (especially if the electro cap is older). So the speakers will probably sound brighter. Some people offset this with a small series resistor, the value of which is hard to determine precisely, and opinions will vary, but I would guess anywhere from .5 ohm to 1 ohm. This step may not be necessary depending on the speakers, the room, your tastes, etc.
No offense to Alan, but I would not fool with paralleling caps to obtain a certain value (unless it is necessary to obtain the correct value). The idea of benefiting by lowering ESR even more than it would be with a single poly cap is stuff for theory. If you can find a polypropylene cap of the appropriate uF value then just go with that. Polypropylene caps are significantly superior to electrolytics, and by using a single poly in place of an electro you will realize the benefits you seek and that will (probably) be audible. And theoretical additional benefits to be had from paralleling poly caps would almost certainly not be audible.
Finally, be aware that there will be a significant difference in physical size of the poly caps vs electrolytics, so mounting considerations will come into play.
https://www.parts-express.com/cat/c...eca:matches(.,"P_Searchable","1")]&PortalID=1
Another factor is that a tweeter, or other driver, will likely be slightly louder because the poly cap has lower ESR than the electrolytic it replaced (especially if the electro cap is older). So the speakers will probably sound brighter. Some people offset this with a small series resistor, the value of which is hard to determine precisely, and opinions will vary, but I would guess anywhere from .5 ohm to 1 ohm. This step may not be necessary depending on the speakers, the room, your tastes, etc.
No offense to Alan, but I would not fool with paralleling caps to obtain a certain value (unless it is necessary to obtain the correct value). The idea of benefiting by lowering ESR even more than it would be with a single poly cap is stuff for theory. If you can find a polypropylene cap of the appropriate uF value then just go with that. Polypropylene caps are significantly superior to electrolytics, and by using a single poly in place of an electro you will realize the benefits you seek and that will (probably) be audible. And theoretical additional benefits to be had from paralleling poly caps would almost certainly not be audible.
Finally, be aware that there will be a significant difference in physical size of the poly caps vs electrolytics, so mounting considerations will come into play.
https://www.parts-express.com/cat/c...eca:matches(.,"P_Searchable","1")]&PortalID=1
Last edited:
Alan, your posting is going to make this poor guy believe that he has to go to the hassle of paralleling several poly caps, when in fact, probably 99.9% of the people who do this modest speaker mod would never fool with this, unless they had to do so to match capacitance values.
I don't doubt that paralleling poly caps reduces ESL and ESR, that is not the point. The point is no one would ever be able to tell the difference and it is a waste of time and effort. Not to mention that you are setting a precedent for this reader and for future readers based on theory alone, as though this is something that must be done. For practical purposes, not theoretical ones, changing from electrolytic to a single poly will yield all audible benefits obtainable.
Speaking of practicality vs theory, when there is an inquiry about replacing electrolytics with poly caps, it can probably be inferred that the speakers being discussed are not a high-end design, i.e. not high-end drivers. Higher-end stuff would have poly caps already.
I don't doubt that paralleling poly caps reduces ESL and ESR, that is not the point. The point is no one would ever be able to tell the difference and it is a waste of time and effort. Not to mention that you are setting a precedent for this reader and for future readers based on theory alone, as though this is something that must be done. For practical purposes, not theoretical ones, changing from electrolytic to a single poly will yield all audible benefits obtainable.
Speaking of practicality vs theory, when there is an inquiry about replacing electrolytics with poly caps, it can probably be inferred that the speakers being discussed are not a high-end design, i.e. not high-end drivers. Higher-end stuff would have poly caps already.
Last edited:
oldman55
Well-Known Member
Don't worry about me, I'll make my own decisions based on all of the facts that I can obtain. There are those that say that if it was designed for ecaps, go back with them.
The speakers were "cheap" by the time they got around to me but that doesnt mean that they dont sound great or werent built with quality components. I find that there is no correlation between $$ and the quality of sound you can produce.
Thanks all for the input.
Last edited:
ConradH
Lunatic Member
In my experience, replacing a tweeter crossover cap with film usually results in a brighter speaker and you have to add a bit of resistance to adjust the tweeter level. The easy explanation is that the film cap has lower esr, which it does, but it's not that simple. Esr is not constant with frequency, so adding a bit of resistance doesn't make a film cap have the same characteristics as an electrolytic cap. Fortunately our ears aren't very good with that sort of subtlety, so the solution works fine.
dzkfraser
Super Member
Those look like Utahs - maybe Knights?(Allied alnico drivers), If so, I have some Utahs that I recapped with polys, unfortunately, the better caps did not make much of a difference in the sound, I suspect because of the limitations of the drivers. If I had known, probably would have just used electrolytics, but then again, will never have to recap again
I might not have experience in crossover design, but I did a lot of work on speaker cables, I found the series parasitic inductance and resistance play a major role to the sound and the lower the better.
Yes, I notice the lower the ESR and ESL, the sound tends to be slightly brighter. BUT it made up by the noticeable increase in sound stage, clarity and separation. I am sure this applies to crossover as it's just down stream from the speaker cable. Far as I concern, when I get the sound stage, separation and clarity, I can always add back the warmth by picking the preamp and/or modify the preamp a little. There are ways to add like 1 or 2dB at frequencies below say 400Hz to add a little bottom without introducing much phase shift like the tone circuit or presence.
For home theater or watching tv, the added clarity, sound stage and separation really make the movie and tv show pop. So when you hear slightly brighter, don't just say no before you spend a little time playing with the rest of the system and spend some time with it.
This ESR and ESL has a lot to do with THD on the speaker side just like with the cable. I think that's where the real problem. Speaker is a non linear device, I proofed THD increase tremendously when driving a non linear load with cable( through crossover also) that has higher ESR and ESL.
Conrad can comment on this. To me, low ESL and ESR between amp and speaker is the KEY to get the best out of the speaker and SS amp.
I am too busy working on my second amp. I have plan to open the speaker, beef up the internal cables, work on the crossover and see how much I can improve the sound. If I am right, it will make a difference.
Yes, I notice the lower the ESR and ESL, the sound tends to be slightly brighter. BUT it made up by the noticeable increase in sound stage, clarity and separation. I am sure this applies to crossover as it's just down stream from the speaker cable. Far as I concern, when I get the sound stage, separation and clarity, I can always add back the warmth by picking the preamp and/or modify the preamp a little. There are ways to add like 1 or 2dB at frequencies below say 400Hz to add a little bottom without introducing much phase shift like the tone circuit or presence.
For home theater or watching tv, the added clarity, sound stage and separation really make the movie and tv show pop. So when you hear slightly brighter, don't just say no before you spend a little time playing with the rest of the system and spend some time with it.
This ESR and ESL has a lot to do with THD on the speaker side just like with the cable. I think that's where the real problem. Speaker is a non linear device, I proofed THD increase tremendously when driving a non linear load with cable( through crossover also) that has higher ESR and ESL.
Conrad can comment on this. To me, low ESL and ESR between amp and speaker is the KEY to get the best out of the speaker and SS amp.
I am too busy working on my second amp. I have plan to open the speaker, beef up the internal cables, work on the crossover and see how much I can improve the sound. If I am right, it will make a difference.
Those are nice looking drivers in your speakers. May I ask what brand and model those speakers are?
I totally agree with you that it is possible to get great sound without spending huge amounts of cash.
The e-cap to poly speaker cap upgrade is likely the most prevalent, most discussed, and best known audio gear modification there is. This was being done in pre-internet days, likely decades before internet forums even existed. The inherent advantages of poly (in particular polypropylene as opposed to polyester) caps vs electrolytic are significant and also well known.
But in internet forums, people get downright passionate (and obstinate) in their opinions in many matters regarding capacitors. My wording in previous posts was no doubt lacking, and I apologize for that. My intent was merely to separate practical, and well established practices, from the superfluous. Hopefully future readers of this thread will have the same kind of common sense that you obviously have.
oldman55
Well-Known Member
Those are nice looking drivers in your speakers. May I ask what brand and model those speakers are?
/QUOTE]
Kanazawa. I believe late 60s early 70s. From the quality of the other components and the cabinet, I am surprised that they would have used inferior caps.
Oldman
At the time these were made, film capacitors in those values were probably huge and very expensive. nonpolar electrolytic was likely the choice just because it made the most sense. They were not so much inferior as they were the only real option. Now we have different options, but a film cap is still going to be larger and more expensive than an NP electrolytic. You're also dealing with 2 speakers, not thousands as the manufacturer was doing.
Like you said, this is a one time thing, particular looking that the pictures, those components are not on pcb, parallel a few is not a big deal. What if it improves the sound? How do you know? Yes, opinion is just an opinion.
You try paralleling cap? I experimented extensively parallel speaker cables to lower ESR and ESL( same and other components, parasitic resistance and inductance) and see tremendous gain in sound quality and have a long thread on this. I don't just listen to others and form an opinion.
To Oldman55
I did extensive experiment in this thread on speaker cable and proof that ESL of the interconnect to the speaker is the KEY to get more transparancy, separation and clarity on SS amp.( don't know about tube amps only have SS amps). Crossover network is part of the interconnect between the amp and the speakers. I believe ( as educated guess) that the effect of ESL in components will be in play just like the speaker cable. If you have $20 to spare, at least try on the caps for mid and tweeter. From my measurement, ESL affect the higher frequency the most, it's a lot more forgiving for the Woofer. Yes, it's going to slightly brighter, but if you stop and listen, you might hear more clarity, separation and better sound stage. Getting back the little bottom is easy by doing it in the preamp......which I did also.
Yes, I am passionate, because a lot of people gave me the chart of sizes of speaker cables for different length. Even amp manufacturers provide that, they said if it is under like 8ft or so, it's even acceptable to use 16 gauge or 14 gauge. I did EXTENSIVELY test with instruments that show even a 10ft 12 gauge Monster cable raise the THD from 0.006% to 0.14% at 20KHz. I proof it's all because of ESL. Paralleling two cable reduced the THD by HALF, 1/4 by 4 cables. My final cable in the system is 6 pair of 16 gauge cables in parallel twisted together to minimize ESL just for the mid and tweeter, then I use 2 monster cable pair just for the woofer. The difference is sound in the system is staggering. It's all in this thread.
In post #202 of page 11, I explained using simple Ohm's law why ESL and ESR cause distortion. It is actually very simple. Simple Ohm's law explain this nicely. Theory do work.
http://audiokarma.org/forums/index.php?threads/effect-of-speaker-cable-on-thd.764633/
Last edited:
Alan, I never said anything like "this is a one time thing". Please do not attribute to me something that I did not say. And for you to use the phrase "...see tremendous gain in sound quality..." in this thread is extremely misleading as it implies that by following your suggestion, that significant improvements in sound will result, which is simply not the case. Your wording is becoming sensational like ad copy.
Question for you:
Given that ESL and ESR improve with a few paralleled caps, why not parallel 100 caps, or a thousand? What number of caps would you choose to parallel? And whatever number you decide upon, I could come back and use your own argument....but Alan, ESR and ESL will improve if you use more caps. But it has to end somewhere. There has to be a point where scientific theory balances with real world, practical considerations, where theoretical gains are offset by effort/cost/time required. I say that point is a single poly cap which is a significant improvement over the single electrolytic it has replaced. Beyond that any theoretical gains would be marginal and would not be audible
The potential improvements mentioned above would be realized by upgrading from e-cap to a single poly. It is not necessary to parallel poly caps in order to realize benefits, and to imply otherwise is misleading.
Upgrading from electrolytic to poly in series with tweeters or mid-range speaker drivers is very well established mod. Many, perhaps thousands of people, have done this simple and effective mod over the course of multiple decades. However, any additional audible benefits, to possibly be derived from moving from a single poly cap to multiple paralleled poly caps, are based on conjecture, and are extremely unlikely.
You already said that this is something that you would do yourself. Fine. But leave it at that and don't attempt to convince anyone (future readers), or instill fear, that if they don't use paralleled poly caps that they will be missing out on "tremendous gains in sound quality", because that is simply not the case.
Last edited:
