Topic: What happens to impedance when you split a signal, and what does that mean?

[taper420]

I'll use an easy example to help illustrate what I'm talking about:

CA cards > 9100 preamp > 1/8 inch split, 1 into minidisc recorder line-in, 1 into MT line in. Now I know I can do this and get a signal that can be recorded, but how is the split actually effecting my signal? I have a basic understanding of resistance and impedance, but this is a little confusing. I'd like someone to explain. What happens to my signal when I'm splitting it and how does this effect it audibly? Am I just losing a little volume, am I cutting certain frequencies? Am I potentially harming my equipment by giving a signal with the wrong impedance? Try to picture this problem not just limited to the MD and MT scenario, but on a bigger scale (more power, more splits)... I know transformers are the proper way to split, but I also know a cheap transformer will hurt your sound. In the case with the MD and MT, I wouldn't use one, but if I was splitting a mic signal (1 into pa, other into multitracker), then I would use a transformed split. So what's the deal?

[DSatz]

taper420, when you split a signal you are placing the two loads in parallel. The net load impedance will definitely be lower than either load impedance is by itself; if the two loads happen to be equal, then the total load will be half what either load is by itself. Looked at from the other end of Ohm's Law, twice as much current would be demanded from your preamp output.

Most preamps are designed to drive a high impedance--something greater than 10 kOhms, for example. And most line inputs of recorders are at least that high--20 to 50 kOhms, for example. As long as both loads are in that range and are primarily resistive (the usual case), there should be no problem, apart from a slight risk that at the highest signal levels, the distortion will be higher than it was with only a single load.

If you have a tone generator that you can connect in place of one of your mikes some time, you could use it to see at what signal level you get audible distortion out of the preamp, and then you could connect both loads to find out whether that maximum level becomes significantly lower. That might enable you to set your levels more wisely under the circumstances, and in any case you would know what the situation really is.

--best regards

[SparkE!]

The main difference will be to your frequency response.  There will not be enough current loading to significantly increase your distortion, although the increase would probably be measurable.

If you're capacitively coupled from your amp (most line level amps are) and the input impedance of your two devices are identical and resistive, then you essentially double the lower cutoff frequency.  I think that CA uses a 6.8 uF coupling cap on the output of the 9100 (correct me if I'm wrong, Chris).  If your input impedance is 10k looking into each of your recorders, then they present 5k of impedance when connected in parallel.  That will give you a lower cutoff frequency of less than 5 Hz which is probably at the lower end of your audible frequency range.  The truth is that most recorders also capacitively couple their inputs, so the real cutoff frequency may be significantly higher anyway, depending on what capacitance they use.

[taper420]

ty both.... I think I followed enough of that to get what I was looking for.... +t's

So in my given example, distortion would happen, but it would probably not be audible (last time I checked I couldn't hear below 20 hertz, let alone 5), and I'd probably need to boost the preamp output a little.
All things considered I should probably look for a way to split the signal with a high quality transformer. You think the A.R.T.'s are good enough?

I could do a passthrough from one to the other, but the whole point is to have a backup, and passing through would make that pointless.