[Elecraft] [K3] Audio circuit distortion

[Scott Manthe]

On 7/23/13 11:39 AM, Don Putnick wrote:
> How much of the audio quality described in the recent threads can be attributed to the audio circuit distortion, as reported by Rob Sherwood at this year's Dayton Contest Academy? The K3 audio section starts around 41:40 into the video.
>
> http://www.youtube.com/watch?v=SOf2OOGeGi8&feature=c4-overview-vl&list=PLRSwUN4qr1Loc58Ip6IB7d43IXkjMat8H
>
> Don NA6Z
> K3, KPA500, KAT500
>
Since that issue was addressed in 2008, I'm not sure they're related. 
Here is one of Wayne's comments from that thread:
> Rob helped us identify the K3's audio output distortion issue, which we subsequently cleaned up with a hardware change (five years ago) and new firmware (about three years ago).
>
> If you have an older K3, see:
>
>     http://www.elecraft.com/K3/mods/K3_AF_Stage_Upgrade_Instructions_Rev_B.pdf
>
> This mod, long since incorporated into production, adds a large RF choke in series with the AF amplifier stage's DC supply. As Rob's plots show, this dramatically knocked down the IMD products; most are now down 70-80 dB. I have pretty good ears, and once we made this change, I couldn't hear any difference between the internal amplifier and external powered speakers.
>
> Rob is correct that it would be possible to further reduce low-level distortion products. However, many of these fall outside the hearing range of the average user because they're related to an image of the 12-kHz CODEC sampling rate. What's left within hearing range could be reduced with a change in the AF amp IC, though there are diminishing returns and pragmatic considerations. The stereo audio amplifier IC we use is capable of driving two 4-ohm speakers at up to a few watts apiece, and it generates no RFI because it is a classic analog device running class AB. We could get higher drive power using a beefier analog device with much higher quiescent current, or by using a class-D or higher switching-style device. The former would increase the radio's current drain substantially, while the latter would require additional shielding and decoupling.
>
> As always, our goal is to continuously improve the K3 in a way that is applicable to units in the field. If we find a practical way to achieve a further significant improvement in the audio channel, we'll certainly make it available.
>
> 73,
> Wayne
> N6KR