Log in
Register
Home
Forums
New posts
Search forums
What's new
New posts
New profile posts
Latest activity
News
Members
Current visitors
New profile posts
Search profile posts
Features
Log in
Register
Search
Search titles only
By:
Search titles only
By:
New posts
Search forums
Menu
Install the app
Install
Reply to thread
Home
Forums
Pro Audio
Test Drive
Powersoft K3, K20, M50Q (Part II)
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser
.
Message
<blockquote data-quote="Langston Holland" data-source="post: 48917" data-attributes="member: 171"><p>Re: Powersoft Digam K3, K20, M50Q (Part 2b)</p><p></p><p><strong>Switching Noise on the Output:</strong></p><p></p><p>It seemed reasonable to have a look at the ultrasonic switching noise emitted by these amplifiers that gave me such fits during my initial measurements with and without 4Ω resistive loads. These high frequencies (typically >/= 10x the audio passband) are responsible for the measurement error in test gear designed for the lower audio frequencies as mentioned earlier. In contrast, the scope(1) is down 3dB at 60MHz and laughs at the slew rate required to measure 1/4MHz switching frequencies. The scope was connected at the end of a 12' 14 AWG cable connected to the amp. This cable is either inserted into a 4Ω load resistor or not. The amps that show a marked reduction in switching voltage with the 4Ω load attached have a relatively high source impedance at those frequencies, which is a good thing. These amps may be able to use a loudspeaker load to effectively finish the job of squelching the noise.(2)</p><p></p><p>The time scale (X axis) is constant for all amplifiers. The voltage scale (Y axis) is 200mV per division for all but the M50Q which apparently uses less filtering on its output. More noise does not necessarily indicate a poor design from an audio point of view, but many introduce EMI approaching the AM radio band. It may be best in terms of raw efficiency for the manufacturer not to add any filtering on a switch mode amp's output, in other ways the ideal is to allow absolutely nothing on the output other than a larger version of what is seen on its input. Engineering draws a line somewhere between these two points to satisfy competing goals, most of which are beyond my knowledge to comment on.</p><p></p><p>A final observation is that many switch mode amps employ a gating feature that eliminates the switching noise when no signal is present after a short period of time. The reason for this apparently is to meet some RF friendly specification when the amp isn't doing anything. These measurements were made without the gating engaged. BTW, it's wise to keep your RF racks away from racks with Class D amps.</p><p></p><p><u>K20, K3, M50Q</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/SwitchingPowersoft.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>FP14000, IT8000, PL380</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/SwitchingOther.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><strong>Transfer Functions:</strong></p><p></p><p>These were made at 8Ω (blue trace), 4Ω (red trace) and 2Ω (pink trace). Note that for convenience I used a single pair of a 12' 14 AWG quad cable from the amps to the junction of the load resistor. That formed a slight voltage divider with the 12' amp cable to load resistor. When I combined all four wires of the quad cable to drop its effective wire gauge to 11, you can see a 0.1dB improvement. Another slight improvement could be made by measuring directly at the amps' output terminals, but that would be a pain and wouldn't highlight the point that loudspeaker cabling forms a passive attenuator on the output of your amp, particularly into low impedance loads. More on this later.</p><p></p><p><u>14 AWG vs. 11 AWG Amp Cabling</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/11vs14AWG.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>K20 Transfer Functions</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/K20_TF.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>K3 Transfer Functions</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/K3_TF.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>M50Q Transfer Functions</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/M50Q_TF.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>FP14000 Transfer Functions</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/FP14k_TF.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>IT8000 Transfer Functions</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/IT8k_TF.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>PL380 Transfer Functions</u>:</p><p></p><p><img src="http://soundscapesweb.com/files/PSW/AmpTests4/PL380_TF.png" alt="" class="fr-fic fr-dii fr-draggable " style="" /></p><p></p><p><u>Footnotes</u>:</p><ul> <li data-xf-list-type="ul">Fluke 192C ScopeMeter.</li> <li data-xf-list-type="ul">Then again, maybe not. Most loudspeaker systems do not have a 4Ω impedance into the MHz region, but probably something more on the order of 600Ω or so.</li> </ul></blockquote><p></p>
[QUOTE="Langston Holland, post: 48917, member: 171"] Re: Powersoft Digam K3, K20, M50Q (Part 2b) [B]Switching Noise on the Output:[/B] It seemed reasonable to have a look at the ultrasonic switching noise emitted by these amplifiers that gave me such fits during my initial measurements with and without 4Ω resistive loads. These high frequencies (typically >/= 10x the audio passband) are responsible for the measurement error in test gear designed for the lower audio frequencies as mentioned earlier. In contrast, the scope(1) is down 3dB at 60MHz and laughs at the slew rate required to measure 1/4MHz switching frequencies. The scope was connected at the end of a 12' 14 AWG cable connected to the amp. This cable is either inserted into a 4Ω load resistor or not. The amps that show a marked reduction in switching voltage with the 4Ω load attached have a relatively high source impedance at those frequencies, which is a good thing. These amps may be able to use a loudspeaker load to effectively finish the job of squelching the noise.(2) The time scale (X axis) is constant for all amplifiers. The voltage scale (Y axis) is 200mV per division for all but the M50Q which apparently uses less filtering on its output. More noise does not necessarily indicate a poor design from an audio point of view, but many introduce EMI approaching the AM radio band. It may be best in terms of raw efficiency for the manufacturer not to add any filtering on a switch mode amp's output, in other ways the ideal is to allow absolutely nothing on the output other than a larger version of what is seen on its input. Engineering draws a line somewhere between these two points to satisfy competing goals, most of which are beyond my knowledge to comment on. A final observation is that many switch mode amps employ a gating feature that eliminates the switching noise when no signal is present after a short period of time. The reason for this apparently is to meet some RF friendly specification when the amp isn't doing anything. These measurements were made without the gating engaged. BTW, it's wise to keep your RF racks away from racks with Class D amps. [U]K20, K3, M50Q[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/SwitchingPowersoft.png[/IMG] [U]FP14000, IT8000, PL380[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/SwitchingOther.png[/IMG] [B]Transfer Functions:[/B] These were made at 8Ω (blue trace), 4Ω (red trace) and 2Ω (pink trace). Note that for convenience I used a single pair of a 12' 14 AWG quad cable from the amps to the junction of the load resistor. That formed a slight voltage divider with the 12' amp cable to load resistor. When I combined all four wires of the quad cable to drop its effective wire gauge to 11, you can see a 0.1dB improvement. Another slight improvement could be made by measuring directly at the amps' output terminals, but that would be a pain and wouldn't highlight the point that loudspeaker cabling forms a passive attenuator on the output of your amp, particularly into low impedance loads. More on this later. [U]14 AWG vs. 11 AWG Amp Cabling[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/11vs14AWG.png[/IMG] [U]K20 Transfer Functions[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/K20_TF.png[/IMG] [U]K3 Transfer Functions[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/K3_TF.png[/IMG] [U]M50Q Transfer Functions[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/M50Q_TF.png[/IMG] [U]FP14000 Transfer Functions[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/FP14k_TF.png[/IMG] [U]IT8000 Transfer Functions[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/IT8k_TF.png[/IMG] [U]PL380 Transfer Functions[/U]: [IMG]http://soundscapesweb.com/files/PSW/AmpTests4/PL380_TF.png[/IMG] [U]Footnotes[/U]: [LIST=1|INDENT=1] [*]Fluke 192C ScopeMeter. [*]Then again, maybe not. Most loudspeaker systems do not have a 4Ω impedance into the MHz region, but probably something more on the order of 600Ω or so. [/LIST] [/QUOTE]
Insert quotes…
Verification
Post reply
Home
Forums
Pro Audio
Test Drive
Powersoft K3, K20, M50Q (Part II)
Top
Bottom
Sign-up
or
log in
to join the discussion today!