US Amateur / Ham Radio operator Jon, KD9DAL has discovered a bug which affects digital performance of nearly all amateur / ham radio equipment that uses USB connections to a computer.

Devices include but not limited too

All Icoms with built in USB Audio
All Kenwoods with built in USB Audio
All Yaesus with built in USB Audio, as well as the SCU-17 Interface
All Signalink USB’s

Jon has reported a with the USB audio chipset used in many ham radio sound interfaces that occurs in windows vista and later. The affected chipset is the TI PCM2900 series PCM2904 and below, pre-C revisions.

All Icoms with built in USB Audio
All Kenwoods with built in USB Audio
All Yaesus with built in USB Audio, as well as the SCU-17 Interface
All Signalink USB’s

If using one of this devices on Windows Vista or later, your performance is reduced due to this bug even if you think the device working fine.

This is confirmed by Texas Instruments, who never recommended this chip set for use on windows Vista and later in the first place, and also tested with more than 50 hams who gained massive receive performance after the fix.

The bug is very odd and has some quirks that bring it back very often even after the fix so I made a detailed video linked below demonstrating the bug, the fix, and showing real measurements of the affected audio on a scope and several meters, and also demonstrating how to properly level your audio in the analog domain and in windows for maximum dynamic range and minimal noise after you fix the bug.

The bug is that this line of TI chipsets identify themselves to Windows Vista and later as microphone devices via the Input Terminal String, even though they are being used as Line In devices. This makes windows add 30db to 50db of gain digitally to the input, clipping your incoming signal. A lot of people I’ve worked with have “worked around” this bug without knowing it by lowering the level slider in windows recording properties panel to very low levels, or by turning their Signalink RX knob nearly all the way down, or a combination of both.

This is very bad, it reduces the dynamic range of your incoming signal pushing it into the noise floor, and most importantly, does not solve the already clipped and distorted waveform due to the unnecessarily added gain. Nothing will return your incoming signal to the pure undistorted sine wave besides the “fix” outlined below, even if you think it’s working well enough now. Trust me, we’ve verified with 50+ people on different setups and radios that performance on modes like JT will double.

For the fix and video demonstrating the problem - http://k3rrr.com/signalink-and-other-usb-digital-interfaces-huge-bug-fix-for-amateur-radio-digital-modes/