You don't have to worry. Modern electrolytic caps are really low ESR. The polymer and hybrid caps sometimes have ESR that is actually too low to damp any resonances.
I can certainly check them out. But what is up with the el-cap property's that - generally speaking - film caps can't deal with, specifically that which J.Didden pointed out about ripple etc ?
The film cap values are just too small. They may also cause resonance issues because they have such low ESR. I would have to see the schematic, but other than a coupling cap, I'm not sure where you could replace an electrolytic with a Wima FKP and have it make sense - either you or the original designer made a mistake. Replacing a 47uF cap with 0.047uF or something is generally not advisable...
Just to correct one thing. When it comes to PSU and decoupling, I am looking at WIMA DC-Link which they themselves state was developed to replace el-cap.
I am have no intention of replacing any values, the only thing that was or is the goal is to replace components so that the over all quality increase, endurance, stability etc.
DC link are not for decoupling though. They are for DC link applications in off-line SMPS or inverters. I find it hard to believe you can replace an electrolytic with even a DC link cap of the same value and have it physically fit. Most DC link PP caps are absolutely huge. Running long leads is not a great idea.
You won't achieve any of these things modifying it. Remember putting a soldering iron down in there means you can screw things up really bad. Far more risk there than leaving a working module alone! Similarly, you don't know whether your replacement part is actually an improvement! Capacitors are better thought of as complex impedance passive devices than two parallel lines on a schematic representing a perfect capacitance.
Please trust the folks who built these modules to have done their job and done it well. We're not talking a hack job of componentry in there.
Please trust the folks who built these modules to have done their job and done it well. We're not talking a hack job of componentry in there.
If it is a scratch build then you should actually do DESIGN, as in figure out what parameters matter in your specific circuit then pick parts to suit, sometimes they will be film, sometimes ceramic, sometimes electrolytic, maybe an occasional tant (really a type of electrolytic), it will depend on your design and what is important at that location.
DC link film caps are generally designed for high voltage at lowish capacitance use cases, they are not classically bulk storage caps but serve an important function in reducing the bus impedance in the hundreds of KHz region where the large elcos are starting to become annoyingly inductive (and lossy) in situations where there is a lot of circulating current.
These things are meant to provide the circulating current path in the HV DC side of switched mode comverters used in motor drives and such, they are physically annoyingly large, heavy and expensive, but if you have many amps flowing at harmonics of your IGBT switching frequency they may be just what is needed, otherwise there are generally better choices.
Modern Electrolytics are NOT a problem if you design using sane margins and parts from a reputable manufacturer (I generally spec Panasonic, with Nippon-Chemi as an alternate and Rubycon as a second alternate, cap availability can be 'interesting' for the sourcing guys). The scaling factors for life are well understood, and even the surface mount parts (That used to have really serious issues with the seals in lead free reflow) have gotten much better. Thru hole is still generally better then SMT IME for electrolytic caps (which is boring, that seal issue and cooking the things during reflow).
Regards, Dan.
DC link film caps are generally designed for high voltage at lowish capacitance use cases, they are not classically bulk storage caps but serve an important function in reducing the bus impedance in the hundreds of KHz region where the large elcos are starting to become annoyingly inductive (and lossy) in situations where there is a lot of circulating current.
These things are meant to provide the circulating current path in the HV DC side of switched mode comverters used in motor drives and such, they are physically annoyingly large, heavy and expensive, but if you have many amps flowing at harmonics of your IGBT switching frequency they may be just what is needed, otherwise there are generally better choices.
Modern Electrolytics are NOT a problem if you design using sane margins and parts from a reputable manufacturer (I generally spec Panasonic, with Nippon-Chemi as an alternate and Rubycon as a second alternate, cap availability can be 'interesting' for the sourcing guys). The scaling factors for life are well understood, and even the surface mount parts (That used to have really serious issues with the seals in lead free reflow) have gotten much better. Thru hole is still generally better then SMT IME for electrolytic caps (which is boring, that seal issue and cooking the things during reflow).
Regards, Dan.
No. Most circuits would not notice.Oneminde said:
OK, now we know which design school you belong to.
No. You sometimes need a little loss, as others have explained.
No.
No.
Strange comment. Why paint 'the best component for the job happens to be cheap and plentiful' as though it was a bad thing which has to be tolerated?
If you have room you could use a film or foil cap with a small resistor in series. Hopefully it won't pick up too much stray RF.
In order to improve a circuit (and its physical implementation) you first have to understand it even better than the original designer. I think this thread shows you still have some way to go on this.
In order to sell a circuit you just need to write a good story which will impress potential customers and journalists. It may help if the circuit includes some expensive or large components, because you can draw attention to these in your story. The story does not have to be true, but people have to believe it; it may help if the story writer believes it.
I would like to add that the OP apparently does not understand where ripple in a power supply comes from.
It is NOT some sort of shortcoming from a capacitor but it is inherent in the rectification process.
Possibly this not knowing 'colored' his comments on the various posts.
Suggest to review the rectification process and find out where ripple comes from and why it is (largely) triangular in shape, and how its shape and harmonic content changes with capacitor value and load current.
Then many comments will fall into place!
Edit: 20 minutes with PSUDII will make it all abundantly clear...
Jan
It is NOT some sort of shortcoming from a capacitor but it is inherent in the rectification process.
Possibly this not knowing 'colored' his comments on the various posts.
Suggest to review the rectification process and find out where ripple comes from and why it is (largely) triangular in shape, and how its shape and harmonic content changes with capacitor value and load current.
Then many comments will fall into place!
Edit: 20 minutes with PSUDII will make it all abundantly clear...
Jan
Last edited:
Thanks for taking the time to explain. I actually like the Hybrid Electrolytic Caps that Panasonic offer. Nichicon, KEMET and Rybycon have similar products and I want to spend some time going over what they offer. I think it could be a really nice upgrade to std electrolytic and at the same time increase durability and longevity. I forgot about Nippon-Chemi so will include them in my research (for PSU and decoupling). WIMA FPK is really interesting on the coupling / signal pat besides C0G 10 ppm/C.
The trick with coupling caps is just to make them have a very high value so the corner frequency is way below the audio band, then they (by definition) have bugger all signal voltage across them, hence no distortion to speak of... For this electrolytics are not bad if your design can tolerate the leakage current.
Where caps become a real design issue is in filters where by definition there is a meaningful amount of signal voltage across the cap, fortunately in electronics these caps can usually be made small enough that C0G, Mica or film parts are practical (Sometimes even polystyrene if you can find them).
One place those bus caps might actually be useful in audio is in PASSIVE crossover networks, but that approach is IMHO flawed in this age of chap amplification stages, better to do the crossover electronically.
Regards, Dan.
Where caps become a real design issue is in filters where by definition there is a meaningful amount of signal voltage across the cap, fortunately in electronics these caps can usually be made small enough that C0G, Mica or film parts are practical (Sometimes even polystyrene if you can find them).
One place those bus caps might actually be useful in audio is in PASSIVE crossover networks, but that approach is IMHO flawed in this age of chap amplification stages, better to do the crossover electronically.
Regards, Dan.
Out of curiosity, which school is that
It appears that way so I will live with (neutral).
Perhaps not so strange if one consider that looking back over a 40 year period, its a blend of 70's equipment still working but would benefit from cap replacement, to products failing within a few years due to failing el-caps (Capacitor plague 1999-2007).
I actually hate BS and audiophoolery related stuff. All I want is good components that is accurate and durable. But I do not mind visually appealing constructions. If I bellong to a school then foremost it would be "that which is visually appealing". Cost is not accounted for here because that is not an argument for or against, it might be a consequence due to certain choices but is not a rule. I totally understand that if you can get away with a more restricted budget and get the job done, this can certainly be very appealing to many, the issue I have is if it becomes a rule that components must be dirt cheap. On the other hand, there are components that was developed for a specific job and using it in another situation is misusage or meaningless because you can't possibly benefit from the "extra" engineering that went into it, then that is disqualified as well. In the end, its a balance between performance and cost, this has always been the case. Looking back at the start of this thread, we saw that circuit distortion is not only related to schematic and board-layout but also component choice. I would say that depending on distortion target - 0.1 vs 0.001% THD or sound signature - this can and often dramatically impact what you choose.
Regarding cost, funny story, but as the price dramatically increase, so does the impression of reviewers or customers. Take Duelund products, they look nice and are visually appealing, but the cost is often not only ridiculous but rather insane. "It has to be good because its expensive", compared to "it is well engineered and perform extraordinary". There is a difference, one is subjective and the other is objective.
But this is how things are. There are different schools, different goals and "rules" for life and people pick sides, be that objectively or subjectively. In the end, if one is happy with the choices, then that is the only thing that counts...that and that the machine or apparatus does what it is suppose to.
I could never use polystyrene due to one reason. They can drift and settle at a new value simply due to soldering.... lol. Something I learned by listening to and watching videos from Mr Carlson's Lab. So yeah, I am scared of does caps
I think its fare to say that I gently lean towards that side, however. Functionality or performance must always come first. If it looks nice but is not doing what its suppose to or do it poorly, I could not possibly still argue that it was a good product because of how it looks or make me feel. What my point is: If you spend hours upon hours engineering something, in the end, if you also spend some on visual appeal, I would not complain. But visual is subjective and to some, functionality is all they need and you know what, sometimes that gets my heart jumping a beat as well.
Lets take an example. Vitus audio, they visually look okay, but opening the lid and looking into their monoblock amplifier and you'll find a big EI core ... That is a big no no, and I think Vitus Audio pricing is a reflection of narcissism but that is for another day... LMAO.
Its not a cut and dry situation or that I can be only one or the other. The technical aspect as well as the visual or "feel" aspect go hand in hand in my world.
The usual problem is that equipment is designed to do one thing (which may or may not involve high fidelity sound reproduction) and then claims are made for another thing (which often involves 'better' sound).Oneminde said:
Anyway, the point we are making is that you cannot necessarily improve a circuit by simply changing to 'better quality' components. You first have to carefully reverse-engineer the circuit (i.e. do the designer's calculations backwards) to gain understanding, then see what he got wrong, then put it right. This is not possible for someone who appears to be struggling to understand where ripple comes from.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.