Quite often you need to split or copy a signal to send to more than one destination. This is commonly done with a multiple, where you plug one source in, and then plug in additional patch cables to go off to multiple destinations. An active or buffered multiple is one that includes a buffer circuit between the input and output, making sure the signal does not lose its strength or integrity by being split too many times, and that no funny business happening on one of the outputs affects any of the other connections. Some modules do have good buffering built into their outputs, and can drive multiple modules without issue. But if you try to use a passive mult to connect to, say, three oscillators, and you realize the tracking isn’t very good (they quickly go out of tune as you go up and down the scale), then you need a buffered mult instead.
Why is this an issue? When one voltage is set to multiple destinations, there can be a drop in voltage if the input resistance or impedance of those destinations are not high enough, or if the output impedance is not low enough. A typical module output impedance is around 1k (1000 ohms); a typical input impedance is around 100k. This ratio means the input would bleed off about 1% of the voltage, which isn’t too bad. But if you have two outputs, they become a divider network that lowers the effective input impedance: for example, two 100k inputs look like a single 50k input. Now the ratio gets up to 2%, which is twice as noticeable. More inputs = more loss = a bigger droop in voltage. That’s why you want to put a buffer in between the input and output: to shield the output from these losses. Ideally, the perfect active multiple would have infinite input impedance and zero output impedance.
All that said, as hinted above some manufacturers (such as Make Noise) make a point of designing their outputs to drive multiple inputs without voltage droop (and therefore, without the need for an active mult); let’s hope we see that trend become more widespread.« Back to Glossary Index