[ppl]

aos i wanted to as previously sed incoperate a regulator plus capicitence multiplier however this has got put on the back burner. for reasons stated above installing one master regulator on the PPA board and allow wall, wart operation without sound quality reduction is an atractive Idea like your P-Dac. but adding a preregulator is apperently not an option.

As Tangent sed the PPA dose not include every option i would like however it is as close as i can persuade the team with.

Dave the cap multiplier should Be after the REG IMHO. not the sunt reg after that may be why you got less than expected out of your circuit.

 

[morsel]

With regards to what PPL just said, I remind everyone that we need to do more testing before we make final decisions regarding the power supply.

I feel it is important to test and compare the performance of the various options before us instead of automatically using the most complex options. If the PPA gets too large, complex, and expensive it may lose its practicality and appeal. On the other hand, we don't want to compromise sound quality. Kurt and I will be doing extensive power supply tests. We will let you know what we learn, but don't hold your breath as we have a lot on our plates.

 

[aos]

The way I "test" them is to place oscilloscope probe at the buffer / opamp power pins and let the amp blast bass heavy music at high volume into some headphones. You will be able to see wobbling of rails to the rhythm of the music. Additionally you can see if supply is clean (i.e. low noise / ripple) when the signal is turned off. Needless to say, if you achieve visible reduction in this contamination, you'll also get quite an audible improvement to the sound. PPA has a lot of circuitry to help eliminate this effect already and the rest of the task is up to the power supply.

 

[Tomo]

Hey,

I don't get it.

Aren't we using CAPACITIVE MULTIPLIER because we need more effecitve capacitance out from small capacitor? Why don't we use BlackGate, ELNA, Muse, OSCON, etc? They aren't so big and from my experience we buy them no matter the cost anyway.

On top of that you are using rail splitters. If you go this much, the amp is NOT portable. Build a REAL PSU. Why not? Battery operation allows really small but good PSU. That is the virtue of using batteries and you are giving that way just because of some convenience.

I understood when you told me that it will take too much space to put in any opamp bypass capacitance at the pins. This is because META was supposed to be a "portable" amp. Now you add AC plugs and abandon the virtues of battery operations. And all for the sake of "compromises." Aren't you going a little too far? How many compromises are you willing to make?

There are too many compromises for DIY amp. You are going away from what you originally intended and what you "are" good at by making more compromises.

Tomo

 

[yejun]

Quote:

Originally posted by Tomo Hey, I don't get it. Aren't we using CAPACITIVE MULTIPLIER because we need more effecitve capacitance out from small capacitor? Why don't we use BlackGate, ELNA, Muse, OSCON, etc? They aren't so big and from my experience we buy them no matter the cost anyway. Tomo

the capacitive multiplier can be as high as 1F, it is impossible to use a real 1F capacitor in a portable headamp.

 

[aos]

Designing a PSU for something like this is pretty tough. You want to use it with batteries, with AC and also likely to charge batteries while AC is connected. Plus the generation of +/- with rail splitters, etc.

I don't think you can get constructive feedback or indeed even have all the issues resolved until you define what is that you really want. Do you want a common on-board part and separate external part for AC only - e.g. achieving one-stage regulation for already clean batteries, and two-stage for rippley AC power - or nothing on board (so when running on batteries, it runs straight from them) and everything power-regulated externally? For the external part, it makes no sense to compromise - once it's out of the box it should be the best it can be, minding the intended use of course (it doesn't need to be better than what the device powered by it deserves, but in designing it space is generaly of no concern). I do not see point in resisting having a regulator in an external supply. Them being noisy is not much of an issue when what they regulate is in a different enclosure (so there'll be noise thanks to long wires anyway). Them taking space is not an issue (external = don't care about space, within reason of course). Them being hard to solder is not an issue either (plenty of 78xx and LTxx). Them being expensive and hard to get is also not a problem.

Anyhow, what I'm saying is that you cannot consider PSU separate from all the usage pattern issues. You can't really say oh let's discuss AC power, and we'll leave battery power and battery charging out of the discussion, we already have that sorted out or will leave it for later. Because even if you do have it sorted out, people out of the loop will not be able to see the whole picture so they won't be able to contribute (or contributions would be worthless). That's a lesson I learned from my own experiences.

 

[Tomo]

aos -

You put it better than I. (I tend to get offensive) Compromises should be made carefully only after you know what you really want.

yejun -

Bypass capacitance is all about "where" they are not "how much" you have. 1 Farad wouldn't do anything if you place them poorly.

And, I think they are planning on making "external" version. This means they can have as much space as they need. Space isn't much of issues. ...

T

 

[Squalish]

Quote:

Originally posted by Tomo aos - You put it better than I.

Me too.

 

[morsel]

Tomo, I am not sure what you are talking about. You are making some strong statements about capacitance multipliers that seem to suggest you are not up on them.

Aos, people are free to use wall warts. This power supply is for those who want something a bit better, either for the PPA or for other amps. I get it that you are really big on regulators. Forgive me if I want to try it with the capacitance multiplier first. Like I said, if it doesn't work out, we can still go with regulators. I think we can get better dynamics from a capacitance multiplier supply than from a classic voltage regulator supply. We won't know for sure until we try it.

Kurt and I have two prototypes so we can do A-B listening and measurement tests. Step by step we will be able to see what works better and what doesn't matter.

I think it is perfectly reasonable to discuss the AC power supply on its own merit, separate from the battery charging circuit. I also believe we have a good idea of what we want, we just haven't tested everything yet, but we are working on it. Even if 90% of what is said here is useless, there is still some good input, and we will keep it in mind.

 

[Tomo]

Morsel -

No, I am not. I don't see why it would offer more dynamics like you "declare."

Anyway, I commented because you declare that you are going official with it. Apparently not.

I think Jung's is better though. You should read Jung's four articles on power supply. However, remember that you are powering OPAMPS and NOT Discrete Class-A or AB stage. You must design an PSU which make sense. Cap Multi doesn't because it is freaking beefy. You are trying to fish a tiny gold fish with a 1 mile long dredge.

Some people telling you it is good isn't enough. You must know why you pick a design. In my opinion, this amp start to sound like a daisy chain of expensive devices and esoteric circuits which do not have clear cut advantages as a whole.

Tomo

P.S. It is like over-stuffing your stomach with turkey on thanksgiving. Just because turkey taste good doesn't mean you can stuff yourself more for more pleasure. It taste good up to certain point and the rest just cause nausia.

 

[eric343]

Whoa, calm down, everyone...

Tomo, I believe the capacitance multiplier was in ppl's original portable amp design as published, and this is after all the PPA...

Keep in mind, everyone, that this is an EXTERNAL supply, so if you don't like it then you can use whatever you want.

 

[morsel]

Tomo, no amount of promises to test everything thoroughly before committing to a design will appease you. You seem to hate our ideas and think we don't know what we want or what we are doing. Thank you for sharing.

BTW, it will be weeks before Kurt and I are done with this phase of testing.

 

[aos]

Morsel, I wasn't suggesting to get rid of the capacitor multiplier, just to use a regulator in front of it. You will need say 120dB of ripple rejection and I don't think you'll get it from a capacitor multiplier. I may be wrong though, as I haven't evaluated them, though I have seen that just increasing capacitors beyond certain value is not very effective at reducing ripple (the rate of reduction compared to increase in capacity is pretty small).

 

[morsel]

I know, aos, I didn't think you were suggesting we get rid of it. I get it that you don't think it will be enough. I don't know how else to say it: We are going to try it without the regulator first. If it is not good enough then we will reconsider the regulator. We are not going to guess, we are actually going to build and test it.

Various people have implied that a capacitance multiplier is equivalent to a big capacitor. That is not the case. If it were, we would not be interested in using it. As I already said, it is a relative regulator, as opposed to an absolute regulator.

We also have lots of PPA amp board testing to do first, including using crappy wall warts and low frequency square wave signals to see how the amp stands up to adverse conditions. We have not forgotten about requests for onboard regulators. Rome was not built in a day. Give us a few weeks to get back to you.

I am off to see X2: X-Men United. Woohoo!

 

[Tomo]

Morsel and Erix,

I am in doubt because all the current must come from somewhere. We are talking about transient reponses, so it is not entirely coming from transformer. Something must supply it. No transistors for that matter would supply that. It must come from the capacitors and everything else must lag behind them.

I don't hate the idea. I don't like the fact that you "declare" cap multi is excellent without much documentations on your part. I want to know why do you think capacitance multiplier is good. Currently I don't see it. More capacitance? I don't hear any difference after I slap in 2200uF. (I hope you realize few thousand uF is really big!)

What I meant by "You do not seem to know what you want" is that your focus is changing. First, your focus was to design the best "portable" amplifier. Now it is "portable" amplifier that plugs into the wall. You should see that portability is not much of issue anymore and your amp is now "non-portable" amplifier. Which means the compromises you made in the name of portability are irrevant now.

Tomo