No problem, you're learning and adapting an existing circuit is a great way to experiment and learn.
Once you get the hang of how things work, common sense will be your best friend.
Now, in a power supply regulator, parallel smaller capacitors, within reason, are helpful. At point of load (your amplifier PCB) and in a perfect world we might have bypass capacitors that range in value or type. In your case if you're worried, I would replace the electrolytic capacitors with Poly-Aluminum types. That would get you exactly what you want.
-Chris
Once you get the hang of how things work, common sense will be your best friend.
Now, in a power supply regulator, parallel smaller capacitors, within reason, are helpful. At point of load (your amplifier PCB) and in a perfect world we might have bypass capacitors that range in value or type. In your case if you're worried, I would replace the electrolytic capacitors with Poly-Aluminum types. That would get you exactly what you want.
-Chris
I am quite happy here in the forum even beginners are tolerated.
And learned how to transform chip amps working then in current drive mode. People here explaining me patiently how to diy.
Sound is unbelievable good especially with fullrange drivers. Conversion is easily done for people in electronics a real non brainer.
Never got my hands on silmic, kz muse or the like poly types of electrolytics.
And learned how to transform chip amps working then in current drive mode. People here explaining me patiently how to diy.
Sound is unbelievable good especially with fullrange drivers. Conversion is easily done for people in electronics a real non brainer.
Never got my hands on silmic, kz muse or the like poly types of electrolytics.
The thing to know about components is that even with electrolytic capacitors, they have various designs to suit specific applications. For example, there are high frequency switching types and normal ones. One is no better than the other, they each have characteristics that make one more suited for a job than the other.
The takeaway is simple. You have to understand both the components, and what characteristics in whatever circuit you are installing them in to get the best performing part. Just sticking different things in doesn't cut it because if you do hear a difference, either something major changed, or your mind is playing tricks on you. If you really did hear a change, it means that one part is better suited to the job, and the other was either defective, or very poorly suited. We aren't talking about brands or model names here, we are talking about basic design and construction. So it is the basic type (construction and materials common to that group), not abc brand, xyz model.
There is no one "best capacitor" for every application.
The takeaway is simple. You have to understand both the components, and what characteristics in whatever circuit you are installing them in to get the best performing part. Just sticking different things in doesn't cut it because if you do hear a difference, either something major changed, or your mind is playing tricks on you. If you really did hear a change, it means that one part is better suited to the job, and the other was either defective, or very poorly suited. We aren't talking about brands or model names here, we are talking about basic design and construction. So it is the basic type (construction and materials common to that group), not abc brand, xyz model.
There is no one "best capacitor" for every application.
Certainly electronics is much more complex than building crossovers for loudspeakers, where I feel at home.
With them it is common to use some bypass foils and combine C's.
I saw the discussion about the sound of poly electrolytics but I am sure often the speaker setup and frequency response influences the choice of electronic C's.
But it's legitimate to do that as "do it yourselfers" - we can create better the sound we want than people who only buy ready made things.
Just omitting decoupling on a Lm1875 makes it sound much softer and forgiving. If you like this sound.
Voila!
With them it is common to use some bypass foils and combine C's.
I saw the discussion about the sound of poly electrolytics but I am sure often the speaker setup and frequency response influences the choice of electronic C's.
But it's legitimate to do that as "do it yourselfers" - we can create better the sound we want than people who only buy ready made things.
Just omitting decoupling on a Lm1875 makes it sound much softer and forgiving. If you like this sound.
Voila!
Well, in crossovers the capacitors are not coupling caps. They are filter elements and as such will sound different depending on many things. Original crossovers are designed compensating for Electrolytic characteristics, changing to foil types will certainly change the sound and sometimes not for the better.
for me an a lot of this forum we have opposite experience !
we are happy to use a good caps like a recipe
Hi nicoch58,
Well, I would say that if they make that big a difference, something is wrong. For sure using the wrong component will generate higher noise or distortion and you can "cook" that way I guess. This is what they do with musical amplification, but reproduction systems ought not to affect the signal. Or at least affect it as little as possible. So the goal is in fact minimum distortion and noise.
If your amplifier (signal or power) has very little overall transconductance you are most definitely going to have minimal correction and higher distortion as a result. Not a path I would follow seriously, but it is fun. I do design tube products as well and enjoy them from time to time. Accurate they are not. Kind of like an effects unit in series with your music.
Well, I would say that if they make that big a difference, something is wrong. For sure using the wrong component will generate higher noise or distortion and you can "cook" that way I guess. This is what they do with musical amplification, but reproduction systems ought not to affect the signal. Or at least affect it as little as possible. So the goal is in fact minimum distortion and noise.
If your amplifier (signal or power) has very little overall transconductance you are most definitely going to have minimal correction and higher distortion as a result. Not a path I would follow seriously, but it is fun. I do design tube products as well and enjoy them from time to time. Accurate they are not. Kind of like an effects unit in series with your music.
Interesting thread, folks, which I've just happened upon. I haven't read all 146 pages (my wife, for some reason, likes me to spend some time away from the screen), but I'll come back to a few pages at a time. For now, just a few thoughts on some topics that I have some experience of...
Firstly, my experience amounts to repair, restoration, and building other people's designs - I'm not clever enough to design anything myself. Over the past few years, I've built up what I think is a decent sounding system (with Audiolab 8000 series amps at its heart), all bought used and each piece I've either repaired, restored, upgraded, or built from scratch.
First step was recapping an Audiolab 8000A. It used Elna electrolytics (an obsolete range now), so they were decent enough to start with. But prpbably 25 years old, in a well-used amp. So I replaced them all with Elna Silmics and Cerafines (depending on what I could find at each value). The design left some voltage headroom, so I went with the original values - if I'm recapping an amp and there isn't much voltage headroom in the electrolytic selections, I'll go up a voltage level.
I also replace the power supply caps with Mundorf Mlytics. They're similarly priced to other good quality ones with similar specs.
I was a bit surprised when I tested all the old caps to find most still measured OK, with only a few being a bit off. I guess they were quality ones to start with.
The result was, I thought, a bit of an improvement in sound quality, but not massive. Presumably just because the caps were newer and probably a step up in quality these days.
The next step was the interesting one, as the amp had electrolytic signal coupling caps in two places (with values that would make film caps quite large and prohibitively expensive). So, time to replace those, but I did a bit of experimenting along the way.
Reading a few forums, I found a lot of people with strong opinions over the relative merits of different electrolytic caps in the signal path, so I thought I try a few. I had new Elna caps in by now, so I got some from several Nichicon ranges (including one bipolar), and I already had a couple of other good brands - can't remember which now. And I wasn't going to waste anything, as I knew I'd use all the caps in the future.
I couldn't do any quick A/B testing obviously, so I just swapped caps, powered it up, and listened each time. People in some forum discussions reckoned one brand had much smoother highs, another had firmer bass, and things like that. Well, I couldn't hear any difference. I don't have top-end equipment, or top-end ears, so I can't be definitive - but for what I want, different good brands of electrolytics are just as good.
The next step was to try some film caps, and I went for Wima caps to replace a pair of 47uF electrolytics, and pair of Mudorf Evo caps for 4.7uF ones (the "best" I could justify at the prices). In each case, I glued the caps to spare areas of PCB and wired them across to their connections - but they weren't far away.
The sound difference was striking - resolution, clarity, dynamics, linearity all improved by an order of magnitude. Easily better than original.
Trying different brands of film caps wasn't possible due to their cost, so that's how that amp is today.
My conclusion? If there really are sound differences between different good brands of electrolytic caps, it seems I can't hear them. But there can be very big sound differences between different types of cap.
Anyway, nothing earth-shattering here, but I just felt like sharing.
Firstly, my experience amounts to repair, restoration, and building other people's designs - I'm not clever enough to design anything myself. Over the past few years, I've built up what I think is a decent sounding system (with Audiolab 8000 series amps at its heart), all bought used and each piece I've either repaired, restored, upgraded, or built from scratch.
First step was recapping an Audiolab 8000A. It used Elna electrolytics (an obsolete range now), so they were decent enough to start with. But prpbably 25 years old, in a well-used amp. So I replaced them all with Elna Silmics and Cerafines (depending on what I could find at each value). The design left some voltage headroom, so I went with the original values - if I'm recapping an amp and there isn't much voltage headroom in the electrolytic selections, I'll go up a voltage level.
I also replace the power supply caps with Mundorf Mlytics. They're similarly priced to other good quality ones with similar specs.
I was a bit surprised when I tested all the old caps to find most still measured OK, with only a few being a bit off. I guess they were quality ones to start with.
The result was, I thought, a bit of an improvement in sound quality, but not massive. Presumably just because the caps were newer and probably a step up in quality these days.
The next step was the interesting one, as the amp had electrolytic signal coupling caps in two places (with values that would make film caps quite large and prohibitively expensive). So, time to replace those, but I did a bit of experimenting along the way.
Reading a few forums, I found a lot of people with strong opinions over the relative merits of different electrolytic caps in the signal path, so I thought I try a few. I had new Elna caps in by now, so I got some from several Nichicon ranges (including one bipolar), and I already had a couple of other good brands - can't remember which now. And I wasn't going to waste anything, as I knew I'd use all the caps in the future.
I couldn't do any quick A/B testing obviously, so I just swapped caps, powered it up, and listened each time. People in some forum discussions reckoned one brand had much smoother highs, another had firmer bass, and things like that. Well, I couldn't hear any difference. I don't have top-end equipment, or top-end ears, so I can't be definitive - but for what I want, different good brands of electrolytics are just as good.
The next step was to try some film caps, and I went for Wima caps to replace a pair of 47uF electrolytics, and pair of Mudorf Evo caps for 4.7uF ones (the "best" I could justify at the prices). In each case, I glued the caps to spare areas of PCB and wired them across to their connections - but they weren't far away.
The sound difference was striking - resolution, clarity, dynamics, linearity all improved by an order of magnitude. Easily better than original.
Trying different brands of film caps wasn't possible due to their cost, so that's how that amp is today.
My conclusion? If there really are sound differences between different good brands of electrolytic caps, it seems I can't hear them. But there can be very big sound differences between different types of cap.
Anyway, nothing earth-shattering here, but I just felt like sharing.
And one more experience before I have to go do something else...
I've been a big fan of active crossovers for many years. It started when I was living overseas, in a city with a thriving electronics district - you could get very nice amp and speaker designs, made locally for not much money, and one place made active crossovers to order.
I started with a passive system - CD, preamp, power amp, speakers, and it sounded pretty good.
Next, modify the crossovers and replace electrolytic caps with film caps, and that made a big difference. (I've done that with several pairs of speakers since, and also replaced iron core inductors with air core ones, with definite improvements each time).
I did this while I was waiting for my active crossover to be built, and I had a second power amp ready. And when I got that and put it together, with passive crossovers removed, I was blown away by how much better it all sounded. (And I don't think the bi-amping itself made much difference, as I'd tried passive bi-amping while I was waiting for the crossover, with only a minor improvement.)
My current system is waiting for me to finish building a new active crossover (a Linkwitz-Riley one, from a Rod Elliott design), and finish repairing one of my power amps (and Audiolab 8000P), and I'm really hoping for a big step up.
What's all this got to do with capacitors?
Well, in my experience, using film caps in speaker crossovers gives a much better result than electrolytics - but different good brands of electrolytics don't make much (if any) difference to my ears. And going for an active crossover (with just tweeter protection caps) takes things up a level.
I've been a big fan of active crossovers for many years. It started when I was living overseas, in a city with a thriving electronics district - you could get very nice amp and speaker designs, made locally for not much money, and one place made active crossovers to order.
I started with a passive system - CD, preamp, power amp, speakers, and it sounded pretty good.
Next, modify the crossovers and replace electrolytic caps with film caps, and that made a big difference. (I've done that with several pairs of speakers since, and also replaced iron core inductors with air core ones, with definite improvements each time).
I did this while I was waiting for my active crossover to be built, and I had a second power amp ready. And when I got that and put it together, with passive crossovers removed, I was blown away by how much better it all sounded. (And I don't think the bi-amping itself made much difference, as I'd tried passive bi-amping while I was waiting for the crossover, with only a minor improvement.)
My current system is waiting for me to finish building a new active crossover (a Linkwitz-Riley one, from a Rod Elliott design), and finish repairing one of my power amps (and Audiolab 8000P), and I'm really hoping for a big step up.
What's all this got to do with capacitors?
Well, in my experience, using film caps in speaker crossovers gives a much better result than electrolytics - but different good brands of electrolytics don't make much (if any) difference to my ears. And going for an active crossover (with just tweeter protection caps) takes things up a level.
Last edited:
Hi oscroft,
Moving to an electronic crossover makes a massive difference! Night and day. Make sure you pull the passive components out of the speaker boxes first of course!
In electronic circuits, film capacitors can be problematic due to size, and locations can cause other issues along with lead length. You can replace on type for another sometimes as long as they fit in the original space. Differences depend greatly on the circuit and impedance levels. There are many times when they simply do not make a difference.
With passive speaker crossovers, film type caps may open slowly over time due to current flow. You should be using foil types or electrolytic only. Inductors, DCR is an issue. It's a trade-off between wire length, core size (construction and type) and cost. The other thing is that the original design was "voiced" using the original parts, you're definitely going to change the sound, so what is right is anybody's guess. Whatever sounds best to you over the long term might be the right answer.
The Audiolab 8000A is a very good amplifier.
Moving to an electronic crossover makes a massive difference! Night and day. Make sure you pull the passive components out of the speaker boxes first of course!
In electronic circuits, film capacitors can be problematic due to size, and locations can cause other issues along with lead length. You can replace on type for another sometimes as long as they fit in the original space. Differences depend greatly on the circuit and impedance levels. There are many times when they simply do not make a difference.
With passive speaker crossovers, film type caps may open slowly over time due to current flow. You should be using foil types or electrolytic only. Inductors, DCR is an issue. It's a trade-off between wire length, core size (construction and type) and cost. The other thing is that the original design was "voiced" using the original parts, you're definitely going to change the sound, so what is right is anybody's guess. Whatever sounds best to you over the long term might be the right answer.
The Audiolab 8000A is a very good amplifier.
Hi anatech, interesting thoughts, thanks.
I can't remember which caps I used to upgrade my current passive crossovers, but they're ones sold for speaker use, so presumably foil caps. Getting ready for going active again, I have a spare set of binding post panels with crossover components removed (they're all on the back). So I can swap out the existing crossovers and keep them intact in case I ever want to put them back.
And yes, there are indeed differences in voicing - and adjusting frequency curves seems like a bit of a fine art with passive crossovers. But I tend to get used to small tonal differences fairly quickly, and I'm far more sensitive to resolution (and staging etc). Interestingly (I thought), replacing the inductors in my passive crossovers only made a relatively small (but worthwhile) difference - I did caps and inductors separately, so I could hear the individual differences.
Anyway, yes, the previous time I switched to an active crossover system, the difference was a mouth-opening one. I was stunned by the clarity and dynamics. I came to the opinion that passive crossovers were (to borrow a phrase Warren Buffett used to describe cryptocurrecy), rat poison squared
(not meaning to insult anyone who prefers them, I know more upmarket ones can sound a lot better than mine).
The crossover I'm building now (boards all complete, ready for test) should be significantly better than my old one. So, high hopes.
I can't remember which caps I used to upgrade my current passive crossovers, but they're ones sold for speaker use, so presumably foil caps. Getting ready for going active again, I have a spare set of binding post panels with crossover components removed (they're all on the back). So I can swap out the existing crossovers and keep them intact in case I ever want to put them back.
And yes, there are indeed differences in voicing - and adjusting frequency curves seems like a bit of a fine art with passive crossovers. But I tend to get used to small tonal differences fairly quickly, and I'm far more sensitive to resolution (and staging etc). Interestingly (I thought), replacing the inductors in my passive crossovers only made a relatively small (but worthwhile) difference - I did caps and inductors separately, so I could hear the individual differences.
Anyway, yes, the previous time I switched to an active crossover system, the difference was a mouth-opening one. I was stunned by the clarity and dynamics. I came to the opinion that passive crossovers were (to borrow a phrase Warren Buffett used to describe cryptocurrecy), rat poison squared
(not meaning to insult anyone who prefers them, I know more upmarket ones can sound a lot better than mine).The crossover I'm building now (boards all complete, ready for test) should be significantly better than my old one. So, high hopes.
Hi oscroft,
Just because something is sold for a purpose has no bearing on whether it is actually suitable or the right part. Often, people who sell things have little to zero knowledge of how the stuff works. Never assume anything.
Foil caps are larger, more expensive than film types. Air coils have much higher DCR than iron core, so that affects damping and certainly box tuning. How much the practical change is, depends. All you can do is try - and measure it. Low frequency measurements are very, very difficult. At least get an idea, try.
The first decade of my career I also designed loudspeaker systems. I tested the drivers by hand to design the enclosure, then measured the results. They agreed with the theory, I used T-S parameters I generated from a population of drivers and compensated for everything. These were sold as kits and beat the speakers of the day except for very expensive speakers. Those guys had more experience, better instruments and access to custom drivers.
A passive crossover simply makes the system go. It makes speakers affordable and greatly increases the WAF. Can you imagine a person coming home with three or four amplifiers and electronic crossover? Wouldn't fly. The expense alone would kill it. Today they try and build them in (and the vibrations and lack of cooling does them in). However, if you do get a good electronic crossover and couple that to good amplifiers and a decent speaker system - glory! A friend of mine just set up a pair of Klipsch Jubilees for his home ... I wish I could afford them. They come with a matched electronic crossover.
I would love to run my speakers with an electronic crossover. Maybe someday I will. They are not designed with that in mind and they could really use a boost in efficiency.
Just because something is sold for a purpose has no bearing on whether it is actually suitable or the right part. Often, people who sell things have little to zero knowledge of how the stuff works. Never assume anything.
Foil caps are larger, more expensive than film types. Air coils have much higher DCR than iron core, so that affects damping and certainly box tuning. How much the practical change is, depends. All you can do is try - and measure it. Low frequency measurements are very, very difficult. At least get an idea, try.
The first decade of my career I also designed loudspeaker systems. I tested the drivers by hand to design the enclosure, then measured the results. They agreed with the theory, I used T-S parameters I generated from a population of drivers and compensated for everything. These were sold as kits and beat the speakers of the day except for very expensive speakers. Those guys had more experience, better instruments and access to custom drivers.
A passive crossover simply makes the system go. It makes speakers affordable and greatly increases the WAF. Can you imagine a person coming home with three or four amplifiers and electronic crossover? Wouldn't fly. The expense alone would kill it. Today they try and build them in (and the vibrations and lack of cooling does them in). However, if you do get a good electronic crossover and couple that to good amplifiers and a decent speaker system - glory! A friend of mine just set up a pair of Klipsch Jubilees for his home ... I wish I could afford them. They come with a matched electronic crossover.
I would love to run my speakers with an electronic crossover. Maybe someday I will. They are not designed with that in mind and they could really use a boost in efficiency.
Hi Never Get Old,
Everyone has an opinion.
All I'll say is this. Buy any good quality one (not Philips or Jameco). Get the highest voltage rating that fits properly. Use axial where you had axial. That will be the best performing capacitor both by measuring total amplifier performance, and by listening. Those two things are the same. This is how I do the best rebuilds.
Now we'll open the floodgates of opinion. Stand back! lol!
Everyone has an opinion.
All I'll say is this. Buy any good quality one (not Philips or Jameco). Get the highest voltage rating that fits properly. Use axial where you had axial. That will be the best performing capacitor both by measuring total amplifier performance, and by listening. Those two things are the same. This is how I do the best rebuilds.
Now we'll open the floodgates of opinion. Stand back! lol!
For hi-fi cathode bypass caps, we don't need but about 50 volts but we do need ~470uF-1,000uF. They are generally considered part of the signal path because they directly affect the signal, even though they are not directly in the signal path. I'm in the process of building my fourth and fifth point-to-point wired stereo tube amps. Axial or radial isn't an issue as far as fit because everything is point-to-point wired. Physical size isn't an issue, but cost always is.
Over the summer I used Nichicon UKA series for cathode bypass in a new Dynaco Stereo 35 build because they are "audio grade" per Nichicon, but they do not list ESR for them. Weird. https://www.mouser.com/datasheet/2/293/e_uka-3082349.pdf I parallelled them with some small value WIMA caps that I had on hand. Just wondering what the "best" two 2 or 3 actually are now. 146 pages into this thread, some discussed here have become the rare element known as unobtainium.
I have been using Rubycon LXW 450V 120uF for power supply caps parallelled with small value WIMA. https://www.mouser.com/datasheet/2/977/e_LXW-1600626.pdf
Last edited:
Hi Never Get Old,
This is very simple. You don't often need more than a 35 volt cap. However capacitor characteristics improve with voltage up to about 63 V and 100 V. This is measurable. But don't ever install parts that are too large, that gets you into other problems. Also, do not change the form factor for the part, or add capacitors hanging off into space. They always mount where the orignal did in the same manner. In addition, external shields do not help one bit. It makes the part larger, heavier and more expensive, it doesn't do anything more than that. If you ground it as a shield, you just increased capacitance to ground, congratulations - another issue.
Now if you look at actual capacitor tests, capacitors do not affect an audio signal until the signal voltage appears across the part. The DC potential doesn't really affect anything. So that means distortion at very low frequencies. Now, before people get excited, if you drive the output transformer below it's ratings, distortion shoots way up, much higher than any capacitor distortion.
What does this mean? Increasing coupling capacitor sze can drive the output at lower frequencies and you were way further ahead rolling off bass above that. Something the engineers who designed the equipment knew and took into account. Engineers did things for a reason, unless you're talking about something really cheap, then cost became a larger issue.
This is very simple. You don't often need more than a 35 volt cap. However capacitor characteristics improve with voltage up to about 63 V and 100 V. This is measurable. But don't ever install parts that are too large, that gets you into other problems. Also, do not change the form factor for the part, or add capacitors hanging off into space. They always mount where the orignal did in the same manner. In addition, external shields do not help one bit. It makes the part larger, heavier and more expensive, it doesn't do anything more than that. If you ground it as a shield, you just increased capacitance to ground, congratulations - another issue.
Now if you look at actual capacitor tests, capacitors do not affect an audio signal until the signal voltage appears across the part. The DC potential doesn't really affect anything. So that means distortion at very low frequencies. Now, before people get excited, if you drive the output transformer below it's ratings, distortion shoots way up, much higher than any capacitor distortion.
What does this mean? Increasing coupling capacitor sze can drive the output at lower frequencies and you were way further ahead rolling off bass above that. Something the engineers who designed the equipment knew and took into account. Engineers did things for a reason, unless you're talking about something really cheap, then cost became a larger issue.
You should check out Doug's site as well. Lot's of great information there.
http://www.douglas-self.com/
http://www.douglas-self.com/
I think the "best" cap is a can of worms for sure.
My preamps have SOME Teflon caps.... I owned them before and after the change and it really did change the sound for the better.
They also needed significant break in...
Yet the rest of my components do not have them.