I just wrote out a way better explanation of what I'm trying to do on reddit:
My Goal: Digital control of the "pots" (ie. the voltages being fed to pins 7 (and 6, in variation) on the 555 timers) in an Atari Punk Console using an Arduino.
Problem: APC runs on 9v DC, Arduino puts out 5v PWM.
(I can provide schematic of the APC circuit if you need it)
First solution attempt: run the APC at 5v directly from the Arduino's power leads.
-rigged up this way, the APC does not work or sound as it should.
-could be because of the voltage cut
-could be because of PWM(?)
-could be a bit of both
There is a possibility that further creative manipulation of the 555s, or something to smooth out the pulses (a cap? what size?) could fix this, but I don't have the knowledge to know, or to attempt a solution.
I did try a small cap when I had this breadboarded, and it just made the whole thing stop working. I'm also not that interested in pursuing this solution, because it would involve becoming an expert on 555s; At this point, I want to do focused study on Arduino, and complementarily learn more general electronics theory (circuit design, doing the math, etc) in order to enable better future Arduino projects.
...which is why right now what I need/want is to learn how I use a puny 5v PWM signal, over which I have complete control, to exercise dominance over a mighty, double-digit-volted river of electrons in some adjacent copper corridor. Cause I've got a lot of circuits that I wish to with this little chip. I want these mighty electrical Goliaths to bend to the will of me and my little Daviduino. You diggin me?
If, on the other hand, this is strictly a matter of its being PWM, and the APC can run on 5v just fine, then I'm definitely interested in getting my head wrapped around the answer to the question, "how do I smooth out the PWM so that a part which needs real DC can handle it?", because the theoretical and practical answers to that question are ones I really should have a solid grasp of at this point, but do not. Caps seem to act in contradictory ways to me, and the reason is that there is some basic thing about them that I have yet to grasp fully.
Anyways, I also tried the optocouplers that came with my arduino, but they seem to only work as switches. Not what I need at all, unless I was doing it wrong, which I probably was. I think my abstract theory is pretty good, but practical is just a bit higher than zilch.
What does seem to offer hope is the existence of digital potentiometers, but many of them seem to also be limited to 5v, which brings us right back to the beginning of our loop function... except there is this, which brings us to my next proposed solution:
http://uk.farnell.com/analog-devices/ad5293bruz-20/ic-dig-pot-1024pos-20k-14tssop/dp/1827274[1]
This digital pot is quite affordable, and can handle the 9v. But can it handle a 9v circuit and accept control from a puny 5v controller? I don't know how to read these technical pages to find out. Even bigger concern: can I solder this onto something that makes it usable to me? I'm sure it's a lot smaller than the picture on my screen, and I've never worked with Surface Mounted components. But it seems promising, if I can only find a brain with the knowledge I need and the willingness to give it to me.
Then it occurs to me that, perhaps, if the mountain will not come to Mohammed, perhaps Mohammed can still come to the mountain. Perhaps, maybe, I can amplify the outgoing voltage from an arduino pin, such that its 5v pulses are magically transmuted to 9v pulses, quickly enough to handle the ms-level switching speed I will need from these parts? I don't know, but I do know that the idea of using digital pots appeals to me greatly, if it is an easy way to translate from arduino to the wider world of exotic electronic voltages.
The last solution I have, for which I have already purchased the parts (cause if I wait around for someone to explain it I'll probably never get it done), is the use of a couple of TIP120 transistors to vary a separate 9v supply, which as I understand it, should both act exactly like a pot in theory, and solve the real DC vs. PWM problem, if that is actually a problem in this particular context (I still don't know).
I'm basing the last paragraph on a couple of tutorials I read where the TIP120 was used to dim 12v bulbs and vary the speed of 12v motors, with a big cap between the poles because of the brushes in the motor, as I recall. I'm assuming I don't need the big cap because I'm not powering anything with brushes... at least, I don't think I am. Anyways, that's what I'm sitting down to do right now, unless someone has posted some amazingly better idea in the meantime.