[sdiy] Module power - regulated or filtered (passively)?

[rsdio at audiobanshee.com]

Lately, I’ve been noticing that the manufacturer data sheets are actually bothering to show an image of the layout to show exactly how close the filtering caps should be. These diagrams seem to all be SMD, and they favor caps with pads that are the size of the pins on the chip, so that signals from neighboring pins can easily get past the cap that’s placed as close as possible to the supply pin. I notice that the “ground” side of the cap is placed in an almost random direction, probably to facilitate routing. They’re probably assuming a solid supply plane on the PCB so that a via can tie the other side of the cap to a low-impedance path to the other supply pin of the chip.

Brian


On Jan 6, 2020, at 8:53 AM, ColinMuirDorward <colindorward at gmail.com> wrote:
> Maybe this is a good place to ask: When people recommend getting those 100nf power filtering caps *near* the chip, how near are we talking? I usually work to make sure they are the closest thing to the chip, but maybe this is overkill. Any rules of thumb here?
> 
> As an aside, re regulating the power down for a 3340 chip, I'd just like to comment that when you drop it to 10v, instead of 12v, I noticed that it tidies up the output levels and becomes a bit easier to derive your 5v p-p outputs.
> 
> Thanks,
> Colin
> 
> On Mon, Jan 6, 2020 at 8:03 AM Florian Anwander <fanwander at mnet-online.de> wrote:
>> Am 06.01.20 um 10:16 schrieb ShedSynth:
>> > One issue I struggled with: the regulated supply has some resistance so a few milliamps drawn by the flashing LEDs causes fluctuations of a few millivolts on the +12V rail,
>> If I remember right, then the solution is not to drive the LED directly, 
>> but to switch a bypass around the LED with an transistor, so the current 
>> is drawn always, but it is running either through the LED or through the 
>> transistor. You may have a look at the schematic of the Roland SH-101, 
>> how they do it there.
>> 
>> FLorian