Input impedance

[polyestr]

Hello all. I am wondering if there is a way to measure the input impedance of my Apogee AD-16x. I ask because I want to record some synths, and am wondering if it is necessary to pass their signals through a DI Box / preamp chain. I want to be able to objectively determine this, since I'm not completely clear what an impedance mismatch sounds like. Specifically, I want to record a Korg MS-20, Jomox XBase-09, and a Roland TB-303. Particularly concerning is the multitimbral XBase, which would require 3 DI/preamp pairs. Any guidance would be appreciated, thanks.

[RodC]

The input inpedance is prob so high that you shouldnt have to worry. Looks like you are wanting to connect fairly modern equipment (solid state) to a high impedance input.

Where you really run into issues is trying to take something like a guitar that wants to see a real high impedance. If you plug this into something much lower, say 600 ohms, this will cause an issue. If the pickup is connected to somthing much lower than 500K it will start to tax its output and the sound will suffer.

You cant plug a guitar into a line level device directly, but not due to an impedance mismatch, but due to the lack of gain. A modern line level will have a high impedance, but will lack the gain needed.

[emrr]

email them and ask. i'm sure it's published

[RodC]

email them and ask. i'm sure it's published

Thats what I thought, I checked out their spec pdf, dont list a thing for impedance.

[Andy Peters]

Hello all. I am wondering if there is a way to measure the input impedance of my Apogee AD-16x. I ask because I want to record some synths, and am wondering if it is necessary to pass their signals through a DI Box / preamp chain. I want to be able to objectively determine this, since I'm not completely clear what an impedance mismatch sounds like. Specifically, I want to record a Korg MS-20, Jomox XBase-09, and a Roland TB-303. Particularly concerning is the multitimbral XBase, which would require 3 DI/preamp pairs. Any guidance would be appreciated, thanks.

It's got balanced line inputs. Figure the input impedance is no less than 10k and probably not more than 25k.

The output impedance of your synths is probably low enough not to matter.

-a

[polyestr]

Thanks for the suggestions. I have no issue with running these guys straight in and addressing any gain deficiencies digitally. Still, determining input impedance would be a useful thing to know how to do, if it didn't require disassembly.

[brianroth]

Quick and dirty way to check (approx.) input impedance, specifically for an unbalanced input:

1. Locate a potentiometer with a value *higher* than what you guesstimate it will be. I'd start with a 50K pot for newer gear.

2. Wire the pot in series with the "hot" audio lead between a signal generator (test signal, or in the case of a synth, the output of the synth with it progremmed to produce a sine-wave-ish sort of steady signal).

3. With the test tone running, adjust the pot so that the signal seen by the gear the generator is driving drops by 6 dB.

4. Disconnect pot and measure the resistance of the pot in that setting with an Ohm meter.

Balanced inputs are trickier since the input stage design may throw off this simple test method.

Bri

[Mradyfist]

Quick and dirty way to check (approx.) input impedance, specifically for an unbalanced input:

1. Locate a potentiometer with a value *higher* than what you guesstimate it will be. I'd start with a 50K pot for newer gear.

2. Wire the pot in series with the "hot" audio lead between a signal generator (test signal, or in the case of a synth, the output of the synth with it progremmed to produce a sine-wave-ish sort of steady signal).

3. With the test tone running, adjust the pot so that the signal seen by the gear the generator is driving drops by 6 dB.

4. Disconnect pot and measure the resistance of the pot in that setting with an Ohm meter.

Balanced inputs are trickier since the input stage design may throw off this simple test method.

Bri

That is actually really interesting. As a guy who's good with computers but not analog electronics, it's nice to have a step-by-step instruction for something sort of mysterious like testing impedance. I was very surprised at how difficult it was to get a proper answer out of Google regarding a similar question.

[brianroth]

"Seat of the pants" test method.

Bri

[polyestr]

Brian - I love it. Thank you for the valuable insight. It's too bad a similar technique doesn't exist for balanced inputs (as I am faced with), but still, the method you describe is really, really good to know.

I wonder if I may sneak a follow-up in here, by asking if anyone can describe the audible characteristics of an impedance mismatch. As RodC suggested, in practice, such situations can involve two problems: the impedance mismatch, and gain deficiency. I can't hear past the gain deficiency, and haven't any clue what I should even be listening for with regards to the impedance mismatch.

This only bugs me because there are situations where impedances are both unpublished and not easily measurable...

It would be great if a device existed, which could be inserted into a signal chain, which would indicate an impedance mismatch above a given threshold, and thus, need for a DI/preamp at that point in the chain.

[The Scum]

A couple of additional tips for this:

- If you're testing with a 2-channel oscilloscope, the common input range switching goes in multiples of 1, 2, and 5 ( IE: .1, .2, .5, 1, 2 5, 10, 20, 50 V per screen division). Put one channel on the generator-side of the pot, set it to a 2X range. Take the other probe and hook it to the other side of the pot, near the input being measured, and set it for a 1X scale. Set the DC offset so the two traces line up. Then apply signal and adjust the pot - when the two traces line up, you've got your 6dB attenuation.

- You might also measure at several frequencies - sometimes impedance isn't entirely linear over the audio range, particularly at the extremes (say, below 110 Hz and above 10KHz). But the specs are often stated as impedance at 1KHz...

[RodC]

Check this out:

http://tinyurl.com/el9bh

[brianroth]

You can TRY the "diddle the pot" method with balanced inputs, but the results may not be accurate with electronically balanced ins or outs, depending on the circuit design. The pot trick WILL work when transformer I/Os are involved.

Hmmmm...maybe a 2-gang pot might work (one section in series with the hot audio lead, the other with the cold) when testing electronically balanced circuitry. The resultinginput impedance would be 2x the measured value of one of the pot sections. But, I haven't tried this, and may be overlooking something.

Bri

[brianroth]

As discussed in the Shure article that Rod linked, modern systems use a "bridging" topology, with a low source impedance driving a comparatively high load Z. This maximizes the transfer of voltage. So, impedance mismatches are a relatively uncommon problem.

One area where a problem can arise is using some gear to drive into a true 600 Ohm input device, like a vintage compressor, etc. Some new gear can't supply enough current into 600 Ohms, and/or has wimpy output coupling capacitors. The latter will cause a droop in the LF response.

Bri

[Andy Peters]

I wonder if I may sneak a follow-up in here, by asking if anyone can describe the audible characteristics of an impedance mismatch. As RodC suggested, in practice, such situations can involve two problems: the impedance mismatch, and gain deficiency. I can't hear past the gain deficiency, and haven't any clue what I should even be listening for with regards to the impedance mismatch.

With modern gear, you don't have to worry about impedance mismatches. By definition, bridging requires an impedance mismatch.

To expand on the notion of bridging impedance (drive is very low impedance, input is high impedance): yep, the idea is to maximize voltage transfer. Practically, what this means is that the driver ouput impedance and the receiver input impedance form a voltage divider. The receiver's input voltage is the divider output. A voltage divider is an attenuator so there's always some voltage loss. However, practically speaking, a set-up with 50 ohm output impedance and 10k ohm input impedance, the loss is 0.04 dB, which is insignificant.

A nice feature of bridging is that one output can drive several inputs with minimal effect on levels.

One thing to watch out for is output drive capability. Sometimes your gear will say something like "capable of driving 600-ohm loads." Some budget gear (specifically, the output drivers) can't do that; some op-amps are specified into 2k loads and lower-impedance loads reduce the maximum output voltage and THD goes up. Even in this case, though, if you tend to run outputs around the nominal +4 dBu and you don't need to drive to the +20 dBu max, you'll probably still be fine.

-a