Removing unused power supply parts from SID boards

[Altitude]

So are there any potential pitfalls or problems with powering SID boards from a common power rail like the Core's in the optimized PSU schematic?  I have to use a different PSU anyway (would TK's OPL3 PSU work here?) since I need the +/-12 V.  Would the procedure be the same as for the cores? (Meaning pull the big cap and rectifier and then just short the regulator pins?).

Also, for adding the 5V to the psu: is it better to branch everything after the rectifier, use 3 rectifiers, or the simply run the 5V regulator off of the +12 regulators output?

thnks

raph

[moogah]

From a post of mine at SDIY asking pretty much the same question about the additional 5v.  I can't answer specificaly if the SID will be ok with not having local regulation, but from a theory perspective (at least without having a datasheet handy) it will be ok.

>I'm working on PSU schematic for my 808 clone today. <snip>

>

>"Here is a simple idea that works really well to isolate noise

>generating sections of a circuit from sections that need clean power -

>it uses just 3 diodes and 1 extra capacitor. In the common leg of a 3

>terminal regulator you add a diode that raises the regulator's output

>voltage by the drop across the diode. You then fit 2 identical diodes

>to the output from the regulator to drop the voltage back down to

>where it would have been. Separate smoothing capacitors are then

>fitted to the 2 separated outputs."

Not really that practical. it will prevent positive spikes over 5 volts

going back to the other circuit by some extent, but a) power diodes are

slow, so it wouldn't be great, and more importantly b) it is very unlikely

such spikes would be generated.

The main noise issue is where one circuit suddenly draws more power than the

other, in which case a spike in the downward direction will be generated.

This is what decoupling capacitors are there for.

>Seems logical enough, but I still have a few questions.  First can the

>7805 really handle +15DC at Vi?

It is the heat dissipation that is the real issue here. If you can get rid

of the heat, yes. Possible ways to reduce what the regulator has to drop are -

1) a power resistor before it. Low ohms. I don't like doing this, as I've

seen it create a lot of noise when current demand gets too high, if too high

a value resistor is chosen. It can also cause a slow power rise, resulting

some weird lockups!

2) string a few poser diodes in series before the input. Each will drop the

voltage by about 0.6 volts. As such, the diodes will be dissipating part of

the waste heat instead of the regulator.

>Will I get better isolation by giving the digial PSU it's own

>rectifier, or will I be fine using the same rectification for both?

That depends entirely on the load. For a few CMOS chips, the regulator may

as well come directly off the same rectifier as the +15 volts. Maybe add a

low ohm resistor (1 ohm?) between the rail and the +5 regulator, and give

the +5V rail another smoothing capacitor after this resistor.

>Ahh, I misread that.. I get it now.  I've got 15 logic IC's in the

>design, so I would think an additional smoothing cap before the 7805

>would be a good idea, 100nF should be enough, no?

100uF and 100nF.

>And then another

>100nF on the output of the regulator to help with those downward

>spikes?

and another 100nF PER CHIP as close to each chip as physically possible

(i.e. shortest track length).

>What about dropping the voltage at the 7805 with a zener and a resistor?

Same issues as just using a resistor. Add a transistor as an emitter

follower, and a 10uF across the zener and you have a basic regulator, which

would be better.

Other possibilities - use two 7805, each sharing half the load, or put a 9

volt regulator before the 5 volt regulator (a little odd, but dissipation

will be shared).

[Altitude]

interesting. the psu i'm working on has a power resistor on a 12V line after the smoothing caps before the 7805 which should drop it (according to alex who is also working on one) to about 7 VDC, which is more than manageable.  I dont know how clean the 5V has to be since it is driving the digital side only

[moogah]

From what I know the important factor is not letting the digial section affect the rest of the PSU, as digital circuits have a lovely noise immunity to them.

Is there a schematic posted somewhere?  Perhaps I could learn a few things here too.

[Altitude]

http://www.midibox.org/forum/index.php?topic=6374.msg48464#new

[stryd_one]

From what I know the important factor is not letting the digial section affect the rest of the PSU, as digital circuits have a lovely noise immunity to them.

Kinda sorta... But the important part, is the other side of that coin....

Digital circuitry is not "immune" to noise - however it is good at "ignoring" noise, provided that the noise is delivered in small enough amounts. Give it enough noise, and you'll have major problems.... Noise in an analog circuit means that the sound might be fuzzy (or all kinds of artifacts)... Noise in a digital circuit means that the system won't be able to tell 0's fom 1's any more. That's bad (duh).

But you'd have a very hard time making a PSU that noisey.

The main reason for keeping the digital section of the PSU separate from the analog, is crosstalk. You don't want 100010011100101001000101 leaking into your analog audio, because it will sound fuzzy and hissy and scratchy and yuk.

I'm at work eating lunch right now, so let me know if you want the extended version of that, and I'll write it when I get home.

[moogah]

What you've got there is pretty much identical to what I'm using for the 808 clone  ;D

With a bit of luck I'll find that the parts I've got backordered at mouser will be here soon and then I can give a more expirienced opinion on the design.  Ken Stone uses something very similar for his modular supply too.