Hi sworder84,
Not familiar with the project, so I will ask two questions. It appears as though you have multiple filter capacitors in parallel. Are you going for the maximum capacitance? If so, there would be your first error. The larger a capacitor becomes, the higher the inductance and worse the high frequency response is (think peak current demands). What you do need is a reasonable amount of capacitance for the power you are delivering. It isn't nearly as much as popular opinion is.
Secondly, have you looked at the specifications for each? I'm not going to take the time to do so since you have narrowed it down to two types. You have some reason for this and I suspect it isn't about capacitor performance. What do you think are important factors in capacitor selection?
I would say you should rethink your project. Also, without a schematic and some idea of bias current levels plus total power capability there are no metrics to go on. The other thing to consider is that most listening is done aroud 1 watt RMS or less. I'm different, at this momenet my systems are running around 10~15 watts RMS - and I am talking right now at this moment into a 4 ohm load. It's pretty loud. My capacitance values in these amplifiers is 10,000 uF per channel with supply voltages of about 71 VDC. That is a ton of power! One thing these amplifiers do not need is higher capacitance. 18,000 uF would be extremely healthy for both channels of any high powered amplifier.
The quality of the capacitor is most important and both of those are high quality brands. There are differences between models. Specifying high capacitance does tie the design into compromises that may not be the best for your application. It's about engineering, not blindly going for the highest amount of capacitance as so many people mistakenly believe. There are other practical considerations due to the rectification process as well. Think rectifier conduction angle, and fast recovery types cause even more problems. Hmm, looks like popular consensus might be incorrect here!
-Chris
Not familiar with the project, so I will ask two questions. It appears as though you have multiple filter capacitors in parallel. Are you going for the maximum capacitance? If so, there would be your first error. The larger a capacitor becomes, the higher the inductance and worse the high frequency response is (think peak current demands). What you do need is a reasonable amount of capacitance for the power you are delivering. It isn't nearly as much as popular opinion is.
Secondly, have you looked at the specifications for each? I'm not going to take the time to do so since you have narrowed it down to two types. You have some reason for this and I suspect it isn't about capacitor performance. What do you think are important factors in capacitor selection?
I would say you should rethink your project. Also, without a schematic and some idea of bias current levels plus total power capability there are no metrics to go on. The other thing to consider is that most listening is done aroud 1 watt RMS or less. I'm different, at this momenet my systems are running around 10~15 watts RMS - and I am talking right now at this moment into a 4 ohm load. It's pretty loud. My capacitance values in these amplifiers is 10,000 uF per channel with supply voltages of about 71 VDC. That is a ton of power! One thing these amplifiers do not need is higher capacitance. 18,000 uF would be extremely healthy for both channels of any high powered amplifier.
The quality of the capacitor is most important and both of those are high quality brands. There are differences between models. Specifying high capacitance does tie the design into compromises that may not be the best for your application. It's about engineering, not blindly going for the highest amount of capacitance as so many people mistakenly believe. There are other practical considerations due to the rectification process as well. Think rectifier conduction angle, and fast recovery types cause even more problems. Hmm, looks like popular consensus might be incorrect here!
-Chris
Yes, I need the maximum capacitance with 35mm diameter. They said that's correct. I aslo saw 10 and 15000uf in sale.
I am choose the electrolytes by quality, brand, prices. Sometimes needs the low esr capacitors. Datasheets of theses two capacitors a very same but I don't know Cornell - never use them and recently learned. 380lx positioned by for audio like Nichicon Muses, is that correct?
I can give you a schematic if it helps.
I am choose the electrolytes by quality, brand, prices. Sometimes needs the low esr capacitors. Datasheets of theses two capacitors a very same but I don't know Cornell - never use them and recently learned. 380lx positioned by for audio like Nichicon Muses, is that correct?
I can give you a schematic if it helps.
The inductance of large capacitors will be compensated by small capacitors near the transistors. This question is removed.
I want to know about these two different capacitors 380lx and Nichicon LS - which one is better for sound?
A cap is an electrical specification. It’s required todo X and it either does it or not.
Audio marketing is an easy way to make a product without changing the specification.
Now if you asked an audio cap manufacturer what the specification that differentiates from a standard cap - that’s lab backed, I think you’d lose 95% of the products as most will be unable to provide the audio specs.
It sounds better - yes but why?
Some things, such as not being polarised, works better with AC where the AC transitions zero..
Audio marketing is an easy way to make a product without changing the specification.
Now if you asked an audio cap manufacturer what the specification that differentiates from a standard cap - that’s lab backed, I think you’d lose 95% of the products as most will be unable to provide the audio specs.
It sounds better - yes but why?
Some things, such as not being polarised, works better with AC where the AC transitions zero..
Hi sworder84,
Given your viewpoint I simply cannot help you. The direction you're going in is simply wrong. Sorry, but that is true. My earnest suggestion would be to accept whatever the people who told you you need massive capacitance without question, whatever they like as far as capacitors. You didn't offer any information as I requested, so I have no information to attempt an intelligent choice. I refuse to take time to examine datasheets on behalf of anyone who is not listening to what I am trying to tell them. Besides, you probably wouldn't accpet my choices anyway.
-Chris
Incorrect.
Given your viewpoint I simply cannot help you. The direction you're going in is simply wrong. Sorry, but that is true. My earnest suggestion would be to accept whatever the people who told you you need massive capacitance without question, whatever they like as far as capacitors. You didn't offer any information as I requested, so I have no information to attempt an intelligent choice. I refuse to take time to examine datasheets on behalf of anyone who is not listening to what I am trying to tell them. Besides, you probably wouldn't accpet my choices anyway.
-Chris
I can only agree with Chris from my own experience.
Large capacitors, and maybe lots of them, are not a good choice for any amp, especially if you want it to sound good.
I did something like this 25 years ago, but I've learned.
There are various ways of calculating how many uF makes sense on the web, but don't be surprised if you only get 10,000uf as a total capacity, which is best divided into 3-4 capacitors, as CRCRC..... with 0.x - 0.0 x Ohm resistors are installed.
Alternatively, you could use a good dual voltage SMPS for audio. You would then only need a capacitor between 470 and 4700uF for each voltage (follow the recommendation of the SMPS manufacturer). I wouldn't be surprised if that beats your conventional power supply by X00,000uF and is cheaper.
And yet, someone who is intimately familiar with the technical aspects is doing their best to patiently - and repetitively - answer your questions, but you are ignoring and arguing with them because they aren't saying what you wanted to hear.Hi @anatech, I heard your opinion, I'll try to wait for what others have to say. I don't know much about the technical aspects. Honestly
Hal
Any option is very good. I use 22.000u/50v. Accuphase has 47.000uF in one capacitor and is very good. Have you 3 big capacitors in paralel on rail ?
I have not seen the design, have zero idea of the power output (into 8R) or the idling current. I don't think that anyone can give advice without this very basic knowledge. I have asked repeatedly for this information.
Honestly, most people don't understand much about circuit design and yet they do design things. Simulator drivers have no feel for reality and simulators are famous for lying to you. Many cmmercial designs bend to popular opinion in order to sell product. It has nothing to do with performance - but they will claim it does as a selling point.
I'm out until I see some actual useful information. Even then I may ignore this further.
-Chris
Honestly, most people don't understand much about circuit design and yet they do design things. Simulator drivers have no feel for reality and simulators are famous for lying to you. Many cmmercial designs bend to popular opinion in order to sell product. It has nothing to do with performance - but they will claim it does as a selling point.
I'm out until I see some actual useful information. Even then I may ignore this further.
-Chris
Thanks, that is an old design. Nothing remarkable and certainly will not benefit from huge amounts of capacitance at all.
So for max current peaks (that is what you are talking about - right?), you need a capacitor that can take and release energy quickly. That description is not for a huge capacitor. You would probably be clipping for this to be a factor anyway. Keep in mind that music has a peak to average level of between 10 and 15 dB, although you can find examples outside of this. That means that listening to a digital source at 10 watts average, your amplifier may well be clipping. Now, reconsider what you are worried about. You will only ever draw peak currents delivering a sine wave into a dummy load (most speakers will not survive that power level even though they are rated for it). So .... you're concerned about what exactly?
You didn't list the bias current, I'll assume betwen 50 and 100 mA. Any more is excessive. I guess you have three caps per rail for the high current supply then? The driver supply doesn't need very much at all. So for the high current supply, you could use 3,300 uF (that's 9,900 uF per rail) and have a ton of excess capacity. They would perform much better and greatly reduce hot switching current and hash from rectification. the inrush current would then also be at sane levels with an E-I transformer (better), a toroid requires a soft start and doesn't do much more than allow higher frequency line noise in, and conduct heavily with effective DC on your mains if there is any.
Smaller capacitors in parallel are only effective near the point of load (or use), and even smaller ones near the rectifiers to try and reduce switching noise. The best way to to avoid generting HF hash as much as possible to begin with.
I'm cetain the designer thinks this is superior to any amplifier that ever came before. You should be more concerned about PCB layout and wire lead dress as that makes a much bigger difference in performance. Think about what I just told you, and figure out your average current at a 10 watt power level.
-Chris
So for max current peaks (that is what you are talking about - right?), you need a capacitor that can take and release energy quickly. That description is not for a huge capacitor. You would probably be clipping for this to be a factor anyway. Keep in mind that music has a peak to average level of between 10 and 15 dB, although you can find examples outside of this. That means that listening to a digital source at 10 watts average, your amplifier may well be clipping. Now, reconsider what you are worried about. You will only ever draw peak currents delivering a sine wave into a dummy load (most speakers will not survive that power level even though they are rated for it). So .... you're concerned about what exactly?
You didn't list the bias current, I'll assume betwen 50 and 100 mA. Any more is excessive. I guess you have three caps per rail for the high current supply then? The driver supply doesn't need very much at all. So for the high current supply, you could use 3,300 uF (that's 9,900 uF per rail) and have a ton of excess capacity. They would perform much better and greatly reduce hot switching current and hash from rectification. the inrush current would then also be at sane levels with an E-I transformer (better), a toroid requires a soft start and doesn't do much more than allow higher frequency line noise in, and conduct heavily with effective DC on your mains if there is any.
Smaller capacitors in parallel are only effective near the point of load (or use), and even smaller ones near the rectifiers to try and reduce switching noise. The best way to to avoid generting HF hash as much as possible to begin with.
I'm cetain the designer thinks this is superior to any amplifier that ever came before. You should be more concerned about PCB layout and wire lead dress as that makes a much bigger difference in performance. Think about what I just told you, and figure out your average current at a 10 watt power level.
-Chris