Strings
There are many ways to approach synthesizing a string ensemble. You can use PWM waveforms and/or a sawtooth. With my Kawai K5, I used a resynthesized version of the "Resonant String" waveform from the Kawai K3.
Since this is an ensemble sound, we need more than one oscillator/voice/part, whatever nomenclature your synth uses to describe one waveform voice. The more the better, but you need at least two.
For this example, we'll use the sawtooth waveform for all of the voices. You may want to experiment with other types of waveforms, especially PWM.
First, layer two or more sawtooth waveforms together. Detune each voice so that you get a nice beating effect between the voices, but not so much so that they sound out of tune with each other. My approach to detuning voices is to detune one sharp slightly and the other flat slightly. That way they are both centered around being in tune, if you see my meaning.
Once you have them detuned, adjust the amplitude envelope so that you have a nice, slow attack, a slow decay, and the sustain level that should be somewhere between 60% and 75% of the overall amplitude. The release should be set so that the sound doesn't immediately cut off when you release the keys. Not too much, though, or it will make the sound muddy (depending on how you play it).
Run this through a low pass filter. Turn the filter's cutoff down somewhat. How far down you set it really depends on the sound you're going for. If you want dark strings, you'll set the cutoff point lower than you would if you want a bright strings sound (obviously). Next, raise the resonance slightly. You just want it to barely accent the cutoff point of the filter.
What you should have at this point is nice, albeit nondescript, pad sound. Perfectly respectable. And you may even want to save the patch to another location at this point to use when you need a soft, quiet pad.
Now, let's really bring this to life. First, turn all of the voices off except for one. Modulate this voice with vibrato. Use your ears to determine how much vibrato to use. What I've found is that in isolation, you may set the vibrato too high. It sounds fine for one voice but may be too much when mixed with the other voices. That's ok. You can readjust it later.
Set the vibrato rate to around 5 Hertz. Also, set the LFO creating the vibrato so that it is delayed somewhat, or better yet, fades in somewhere between an eighth to one-fourth of a second.
What you should have is a nice violin type sound. While we're here, use a pitch envelope or "auto-bend" to start this voice of sharp. Set it so that the envelope brings the voice in to pitch quickly. Use your ears to tell you how much modulation to use and how fast to set the envelope. Since the amplitude envelope has a slow attack, the results of the pitch envelope may not be very evident. That's ok.
There are a couple of reasons to use the pitch envelope: One is for realism. When a bow strikes the string and the violinist pulls the bow across the string, the string is being displaced. It then quickly snaps back into position. Frequency wise, the string starts out sharp and quickly goes back to pitch after it stablizes, just like a plucked string.
The second reason to use a pitch envelope, at least on a digital synth like the K5, is that it will offset the phase relationship of the voices. Since at least one of the voices is starting sharp before going back to being in tune, its phase will be at a different point than the other voices from the beginning of the sound. This makes a difference and it is one of the reasons why analogs sound so nice. Using a pitch envelope in this way on a digital synth is one way of approaching that phenomena. Modulate the pitch envelope with velocity, and it's even better.
Time to bring in the other voices. For each voice duplicate the vibrato settings of the first voice. Then slightly change them so that they are not exactly the same. For example, if the first voice's vibrato rate was 5 Hertz, the second voice's vibrato rate could be set to 5.125 Hertz. The different settings is important to acheive an ensemble effect. Do the same for the pitch envelope.
This should bring us to the point of having a nice string ensemble sound. Be sure to experiment. The audio example I've posted to the files section uses four voices, all with different vibrato settings, to create a nice string ensemble sound.