The program/algorithm is called ALCE2 by Roberto Bigano and is a Photoshop Plugin - that photo is just an example they show and just to be obvious, maybe too obvious. But the effect can be dialled in to be a lot more subtle than that example, from 0% to 100%.
It is the same with the Darbee processor - you can set it between 0 and a max of 100. I haven't seen it yet but I have spoken to an Oppo distributor who has seen it with the '103 Darbee and he said the best effect without exxageration is when setting it to 25-30. If you set it to zero, it does nothing at all.
But some people will want to really wind it up, but I suspect you and I will treat it the same way we add a little salt to our food and if you add too much it will be worse than too little.
I will be getting a '105 Darbee during the 2nd part of February. The ship is due to arrive in Melbourne in the middle of the month.
Cheers, Joe
I can only say one thing: USE THEM !!!
You won't regret it - and they are nothing like the old ceramics - and yes I agree, I wouldn't use those either.
Cheers, Joe
You are right about the size, this is one instance it does matter.
These caps actually do two things at the same time.
Let's say you have a regulator, most of them usually struggle to have sufficient bandwidth, so at HF their impedance goes up. Then you also has centimeters or even inches of track to the power input on the IC or DAC. Those tracks will have some inductance, not a lot, but important at HF and VHF.
Put those stacked film caps as reasonable close to the target, the the output impedance plus any latent inductance (the track) will give it a 2nd order filter effect that attenuates noise that originates from the power supply.
At the same time, ground noise coming out of digital (switching) chips, especially DACs - and you have a very Low Z ground path for back-EMF coming out of the DAC. These things are like Black Holes to noise coming from two directions.
But you are right, the effect, both ways, is maximised by using a LOT of capacitance.
It also works very well with analog circuits, but the biggest pay-off is on the DAC.
Modifying SMPS or even converting them to linear, makes less difference than this little device does when used properly. This is an elegant solution to noise management and just shows if you throw money at something is not always that smart. Anyway, that is my opinion.
Cheers, Joe
.
I understand where you are coming from - but use these where there is no ripple current and I use a minimum of double voltage rating and they are fine. I have literally used thousands of them, strictly according to those conditions, and failures has been very rare indeed.
Cheers, Joe
Pile 'em high
Here's my latest build - there are 20 1206 22uFs (4wide, 5high) sitting on top the AD8017 in SOIC8. In the background is the inductor (two ferrite beads) feeding the power in. Missing are the lytics which go in parallel to maintain the 'crowbar' effect down to the bottom end of the audio band.
Here's my latest build - there are 20 1206 22uFs (4wide, 5high) sitting on top the AD8017 in SOIC8. In the background is the inductor (two ferrite beads) feeding the power in. Missing are the lytics which go in parallel to maintain the 'crowbar' effect down to the bottom end of the audio band.
Attachments
Yeah you don't have to set full pop. The ideal i find largely depends on the panel/projector and the source quality. Either way, once setup right, if you turn it off, the picture looks "broken" without it...
Which caps X5R?
X7R MLCCs are used on high reliability products all the time with no problems, I would suspect either they are using the cheapest caps or some other problem.
This is a clever and recommended way to make large ceramics capacity. Unfortunately after one solder together all these caps, there is not possible to desolder it again. Well, it is, but the rate of destroyed components it may increase quite much... If one will then want to experiment something, then it should build up another one...
I will suggest that when you solder together two or many such caps, only place a kind of spacer in between the caps, and make a soldering bridge over to connect/solder the caps together. Then remove the spacer. So it will be more easy to remove the soldering bridge and separate the caps. As spacer one can use a piece of thick paper, or something which will not melt under soldering process. Something around 1 mm it may be right enough...
So, it may look more like the Sears tower, and it will be less Manhattan style...
Last edited:
Adding spacers degrades the performance though - more resistance and inductance. I'm not worried about re-using individual caps but have re-used blocks of them. They're not very expensive these ones - a reel of 2000 costs around $35. But I do hanker after the top of the range ones (100uF) which go for six times that
Adding spacers degrades the performance though - more resistance and inductance. I'm not worried about re-using individual caps but have re-used blocks of them. They're not very expensive these ones - a reel of 2000 costs around $35. But I do hanker after the top of the range ones (100uF) which go for six times that
When you solder so the caps, there is almost nothing left as total inductance in the packet.
A half mm or 1 mm in between is also nothing as resistance or inductance as well, related to all the construction...
But up to you, of course.
I agree that one mm it may be huge distance inductance wise in Ghz or many hundreds Mhz range, but is not the case in this application here...
At least it may be half mm, or so, but anyway this is not so important.
for any decoupling application inductance is the killer and stray parasitic inductance can make your decoupling totally useless. The determining factor for high speed digital design is the signal rise time, and the knee frequency derived from this. Look up all the documentation on decoupling and you will see that minimising inductance is more critical that ultimate capacitance value...
On the Oppa I believe there are 50MHz clocks so yes 1mm is critical in decoupling applications, and on a cap stack like this would render the stack pointless due to parasitic inductance....
Remember for digital decoupling parasitic inductance must be minimised.
On the Oppa I believe there are 50MHz clocks so yes 1mm is critical in decoupling applications, and on a cap stack like this would render the stack pointless due to parasitic inductance....
Remember for digital decoupling parasitic inductance must be minimised.
Coris, in movies, as in music, our world is quite imperfect: The light is not right, the focus is wrong, the dynamic range is restricted, there is too much contrast, too much reverb, the cameraman is hangover, the music recording/mixing engineer is stoned, etc., etc., etc.
So, what do we do? We take charge ourselves and we modified to our own taste. We correct some mistakes made, we make our moving pictures and music recordings more attractive, more lively, less dead and less badly fabricated.
We trust ourselves more than we trust others.
We live in an imperfect world ...
You have right!
I do not know exactly what the circuit shown by abraxalito does, but I suppose it works in analogue domain, and maybe DAC post processing. Here it may not be about clock frequencies, but a HF noise into few Mhz. If is something which process clock of 54Mhz or so, then I agree with you that parasitic inductance of the decoupling caps it may be important.