I know that the impedances should be matches so the input impedance of the device is around 10 times bigger than the output signal that goes to this input. But is there a limit?

If I have a signal coming out of the device and its 1k ohm then I'd load it into something with impedance around 10k (line level). But what would happen if I loaded it into input that has around 1 milion ohms (instrument level)? Will the signal be attenuated so I can hear it because it's too weak? Where's the boundary that shouldn't be crossed? Or is there any?

  • 1
    I think you're worrying about it all too much. I've been doing this over 40 years & basically if the plug fits the socket, it'll do.
    – Tetsujin
    Jul 19 '19 at 17:37
  • @Tetsujin Agreed, with the caveat that 1/4” instrument cables should not be used for speaker connections, even though the plug will fit. Well, there’s also that 1/4” instrument outlets generally won’t work in 1/4” line level inputs. Really 1/4” cables are the confusing part. Jul 19 '19 at 17:41

There’s a difference between matched impedances and bridged impedances. With matched impedance, they are matched. Meaning both output and input impedances would be about the same, like 8 Ohms for a speaker connection. Matched impedances provide the best power transfer.

Bridging impedances, where the input of the next stage is about ten times or more than the output impedances, provide the best voltage transfer. They also generally support longer distances.

I think one limitation on the input impedance is the next stage is that after the signal gets past the input, you have something you want to do with it, and you don’t want to have to change the impedance overly much before wharvwe processing you want to do.

Also, the greater the difference in impedance, the lower the power transfer, so at some point you really start to weaken the overall signal level. And higher value resistors are more expensive.

  • Ok, I got it. Since I'd first send microphone signal to mic preamp to add gain (which has low impedance line output, around 60 ohms) and then to a high impedance guitar pedal (around 500k input impedance), will the signal suffer greatly, worsening noise to signal ratio?
    – Toby
    Jul 20 '19 at 8:41
  • Good answer, but in what sense do bridged impedances support longer distances? And, the cost of higher value resistors is hardly going to matter. What does matter is that they incure more Johnson-Nyquist noise. Sep 11 '20 at 8:51

High impedance inputs tend to be also comparatively high voltage inputs. That's what makes them reasonably noise tolerant. Feeding them with low level signals will force you to crank up your gain with a resulting less than favorable signal/noise ratio. For that reason, using a DI for transforming a low level low impedance signal to a higher level (and, given a passive DI, higher impedance) signal may make quite a bit of sense. One typical use case would be for matching a dynamic microphone (balanced, low level, low impedance) to an instrument amp input (unbalanced, higher level, high impedance).

  • @user62068 Hi! to match impedances I'd use a reamping device, right? The DI you mentioned translates high impedance signal to low impedance and that's the opposite of what I want to do. Am I wrong here?
    – Toby
    Jul 20 '19 at 8:37

If a high impedance source is connected to a lower impedance input the signal will be attenuated, more so at the higher frequencies. When a low impedance source is connected to a higher impedance input the signal is weak, but not distorted. The best power transfer occurs when the impedance matches and the best voltage transfer occurs when the input is up to ten times high impedance than the source. The greater the difference in impedance the more the signal will be affected.

I was a loudspeaker (cabinets and drivers) designer in the 1970s.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.