All about LFOs

[vendettav]

ok so.. you know I've never really done anything with LFOs i know what they are... but can't realy see using them... well may be with LDR/LED thingy but other than that. how would you go with implementing an LFO into something?
sorry if it's a stupid question but i've never really done anything in the LFO area

[R.G.]

That's actually a good question, and a good intro to something I keep intending to write up.

Every LFO has to match what the modulator it drives needs. And modulators vary, hugely. So you have to make your LFO output match the DC level, signal size, and voltage or current nature your modulator needs. IF you don't, it either won't work at all, or won't work well.

Modulators are all different from one another. A JFET used as a variable resistance has to have its gate at a negative DC level from its source, and the size of the signal at the gate needs to be less than its Vgsoff. A MOSFET needs its gate to be positive with respect to its source by typically a few volts, and it also has a reverse diode across it that complicates the signal level. A bipolar transistor can be used as a modulator for small signal (under perhaps 50mV) but it also has a DC offset needed on its base, and generally prefers to work with what amounts to a variable current into its base. An OTA needs a no-fooling current drive into its bias pin, and this current drive needs to be DC offset and voltage compliant enough to set the pin at about one to two diode drops. An LED-LDR needs a voltage offset of the LED on voltage, and then the current needs to be varied, not the voltage. An LED plus resistor fakes this by the current being the difference between the drive voltage and LED on voltage, and the current then being determined by ohm's law on that voltage difference.

They're all different, and to do a good LFO modulator, you simply have to have the LFO drive match the modulator's needs. No substitute for knowing these exists.

[vendettav]

thanx R.G that was a very informative read. I now got the intro a very good intro. what would you recommend as further read/reference?

also interested in Transistors as variable resistance. I'll try and research but if anybody can contribute for this thread I'd be thankful

[R.G.]

thanx R.G that was a very informative read. I now got the intro a very good intro. what would you recommend as further read/reference?

also interested in Transistors as variable resistance. I'll try and research but if anybody can contribute for this thread I'd be thankful

The way it was put to me in class was that real resistors have a straight line of R = V/I. curved V/I lines represent changing resistances to any sufficiently small signal. A tiny signal can only "see" a small section of a curved V/I line, so it acts like the resistance it sees is the slope of the V/I curve at its point. For instance, diodes look like high resistances for tiny signals around bias points of 0V up to maybe 0.4V, at which time the tiny signals "see" a decreasing incremental resistance. When the voltage hits above the conduction knee, the incremental resistance goes down to maybe 10-20 ohms. So you can change a diode's resistance by changing the voltage/current in it. This is how diode modulators work, and also some voltage controlled synth filters.

BJT based variable resistors work by varying the conduction resistance of the transistor in the region below about 0.4V from collector emitter. JFETs as variable resistors do much the same, but their variable resistance region goes up to a volt or so before the non-ideal nature of the changes make its distortion impossible to ignore. Bipolars work somewhat better as variable resistors and signal switches when used in the inverted mode, swapping C for E.

Remember - for diodes and bipolars, the region of usable variable resistance is quite small. You may have to divide a guitar signal down for modulation, then amplify it back up, living with the noise issues that causes. There are good technical reasons that these applications are hard to find data on. They are hard to get right and have limitations big enough to make designers hunt for other solutions.

[vendettav]

hmmm so what i get is that transistors as variable resistance aren't quite practical and aren't of much use?

[R.G.]

hmmm so what i get is that transistors as variable resistance aren't quite practical and aren't of much use?

Well, they have a quite limited signal range, and special care has to be taken to get rid of the DC offset and control feedthrough, which modulates the DC offset as well.  It's more like there are less complicated and fussy ways to do it, which is why it's generally passed over in favor of other techniques, unless the application requires a modulator that's so cheap that the problems can be worked around or ignored. The two examples, the EH Pulsar and Vox repeat percussion come to mind as examples of this, too.

[vendettav]

i was actually looking into pulsar yesterday. hmm makes sense actually. how about other, better solutions. again so far i know of optoisolators only