Reply to thread

Message

<blockquote data-quote="Nick Hickman" data-source="post: 39307" data-attributes="member: 556">Re: Loudspeaker to Line Level PadHi Bennett,With two 30 ohm resistors, yes, but I&#039;m not sure how you&#039;d do it with one of each!If you put R1 and R2 in series across the amp and take the output from across R2, the attenuation is 20*log10(R2/(R1+R2)) assuming the amp output Z is zero and the load Z applied infinite.&nbsp; For an amplifier output and line-level input as load, the impedances are pretty close to zero and infinite respectively, but if they&#039;re not, you need to include them in the analysis.R1=570 &amp; R2=15 ohms (or R1=1160 &amp; R2=30 ohms) get you -32dB.It&#039;s not the 30 ohms (R2) that&#039;s the issue; it&#039;s R1.&nbsp; If the amp is rated at 1000W/4ohms, it&#039;ll dissipate about 7W into the 570 ohm resistor in the 570+15 ohm case.Excel is ideal for problems like this.&nbsp; <a href="http://100dB.com/misc/amppad.xls" target="_blank">For example</a> (you fill the yellow boxes).Ales is right is raise concerns about bridged outputs.&nbsp; If padding a line-level output into a line-level input, you can make the pad balanced by splitting R1 in half and putting half in each leg.&nbsp; But with a power amp output to a line-level input, things can easily get outside of the common-mode range of the input.&nbsp; A transformer after the pad is the easiest solution, though you then also have to consider the impedances the transformer wants to see hung across it to keep things linear.A 600 ohm load is really neither here nor there to a power amp, so I&#039;d be inclined to scale everything up so that the power in R1 diminishes.&nbsp; With a high attenuation pad, the output Z remains nicely low.I did have a plan to build a similar amp pad with switchable attenuation (and a pot to give something like a -20dB to -60dB range).&nbsp; Never got round to building it!Nick</blockquote>

[QUOTE="Nick Hickman, post: 39307, member: 556"] Re: Loudspeaker to Line Level Pad Hi Bennett, With [i]two[/i] 30 ohm resistors, yes, but I'm not sure how you'd do it with one of each! If you put R1 and R2 in series across the amp and take the output from across R2, the attenuation is 20*log10(R2/(R1+R2)) assuming the amp output Z is zero and the load Z applied infinite. For an amplifier output and line-level input as load, the impedances are pretty close to zero and infinite respectively, but if they're not, you need to include them in the analysis. R1=570 & R2=15 ohms (or R1=1160 & R2=30 ohms) get you -32dB. It's not the 30 ohms (R2) that's the issue; it's R1. If the amp is rated at 1000W/4ohms, it'll dissipate about 7W into the 570 ohm resistor in the 570+15 ohm case. Excel is ideal for problems like this. [url=http://100dB.com/misc/amppad.xls]For example[/url] (you fill the yellow boxes). Ales is right is raise concerns about bridged outputs. If padding a line-level output into a line-level input, you can make the pad balanced by splitting R1 in half and putting half in each leg. But with a power amp output to a line-level input, things can easily get outside of the common-mode range of the input. A transformer after the pad is the easiest solution, though you then also have to consider the impedances the transformer wants to see hung across it to keep things linear. A 600 ohm load is really neither here nor there to a power amp, so I'd be inclined to scale everything up so that the power in R1 diminishes. With a high attenuation pad, the output Z remains nicely low. I did have a plan to build a similar amp pad with switchable attenuation (and a pot to give something like a -20dB to -60dB range). Never got round to building it! Nick [/QUOTE]

Verification