>2.1.1 is insane. They can't make up their mind whether to use 0r >resistors or beads, so they are still guessing.
>Splitting grounds is usually a bad idea. Except when it's a terrible >idea.
>Fig 5.3 looks so weird to me that I assume PDF rendering errors. Does >anyone else see huge black triangles?
>2.1.2 suggests sequencing or possible latchup problems. Be careful >here; get more info maybe.
>John
I get the same triangles, silly idiots for letting something out like that in a ref. design. I don't think I need all the extra bits, just 1 vid in and an AVR or 8051 to control it, so I was just going to have a vdd and gnd copper flood on the inner layers, get some cheap boards done to test.
>>2.1.1 is insane. They can't make up their mind whether to use 0r >>resistors or beads, so they are still guessing.
>>Splitting grounds is usually a bad idea. Except when it's a terrible >>idea.
>>Fig 5.3 looks so weird to me that I assume PDF rendering errors. Does >>anyone else see huge black triangles?
>>2.1.2 suggests sequencing or possible latchup problems. Be careful >>here; get more info maybe.
>>John >I get the same triangles, silly idiots for letting something out like >that in a ref. design. >I don't think I need all the extra bits, just 1 vid in and an AVR or >8051 to control it, so I was just going to have a vdd and gnd copper >flood on the inner layers, get some cheap boards done to test.
>martin
If you're going to do a test board, start with one solid ground plane, with power planes below/above as needed (likely only one power plane is necessary here) preferably with thin dielectric between the planes. Handle any signal integrity/noise problems locally.
It's amazing how many really silly reference design/eval boards there are.
Martin Riddle wrote: > "Martin Griffith" <mart_in_med...@yah00.es> wrote in message > | Just been looking for ideas > | http://www.gennum.com/video/pdf/32469DOC.pdf > | and I havent done any multi layer boards, is fig 5.3 on P24 correct?
John Larkin wrote: > Splitting grounds is usually a bad idea. Except when it's a terrible > idea.
Except when it's audio of course and you don'y want digital return currents you can hear in the audio ground. In which case pkysically separated but carefully linked planes are the way to go in order to direct the currents where you want them to go rather than randomly anywhere.
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote: >Fig 5.3 looks so weird to me that I assume PDF rendering errors. Does >anyone else see huge black triangles?
Yes, with both the Foxit and PDF-XChange viewers so it's probably in the original.
>> Splitting grounds is usually a bad idea. Except when it's a terrible >> idea.
>Except when it's audio of course and you don'y want digital return >currents you can hear in the audio ground. In which case pkysically >separated but carefully linked planes are the way to go in order to >direct the currents where you want them to go rather than randomly >anywhere.
>Graham
Ha.. I remember a student who said he could make a small audio desk with the equalisers using MF10's. Never saw him again....I wonder why?
Martin Griffith wrote: > Eeyore wrote: > >John Larkin wrote:
> >> Splitting grounds is usually a bad idea. Except when it's a terrible > >> idea.
> >Except when it's audio of course and you don'y want digital return > >currents you can hear in the audio ground. In which case pkysically > >separated but carefully linked planes are the way to go in order to > >direct the currents where you want them to go rather than randomly > >anywhere.
> Ha.. I remember a student who said he could make a small audio desk > with the equalisers using MF10's. > Never saw him again....I wonder why?
Was that the NatSemi ? switched capacitor filter ?
Joerg wrote: > Martin Griffith wrote: > > Just been looking for ideas > > http://www.gennum.com/video/pdf/32469DOC.pdf > > and I havent done any multi layer boards, is fig 5.3 on P24 correct?
> Weird. But I'd be a lot more concerned about figure 5.2. Splitting > grounds is rarely a good idea. Might also blow the EMC cert.
