Will upload a video shortly on this.
About 7kHz whine from speakers (verified with scope.) Horrible noise, very audible. Also, it will not always turn on - it shows digital glitch art (random mess) on the screen when it does not. Picture is fine if it turns on, even if it makes noise. And, once set warms up hideous noise goes away.
I checked the amp board (since it has a separate one) and the noise is present on both input & output. Ruled that out for now then. So, I checked the power supply. There's some ripple at about 420Hz, but it's only about 100mVp-p on the 24V rail so I disregarded that.
I then went to check main board. I tried freeze spraying a few things and once I hit an LM2576T switching regulator IC223 the noise got worse...hmm...
Checked the output, 1.77V DC on the DMM so seemed OK (rated 1.8V on schematic), maybe slightly low but nothing to write home about. So used my scope on it and I find there's about 300mVp-p noise riding on it at 30kHz. In addition to this there's a lower frequency harmonic at 7kHz being modulated into this - the same frequency as our speaker noise. Freeze spraying it alters this noise and the tone heard from the speakers.
No wonder the CPU crashes sometimes! It's getting a supply voltage from 1.6V to 1.9V at 30kHz, it's amazing it works at all!
Replaced capacitor C905 and the problem was fixed, both noise and glitch art, it is near IC223 and is the input capacitor for the converter. It had ESR of 34.4 ohms!
Moral of the story is that just because a switching converter has a good output voltage doesn't mean it's working!
About 7kHz whine from speakers (verified with scope.) Horrible noise, very audible. Also, it will not always turn on - it shows digital glitch art (random mess) on the screen when it does not. Picture is fine if it turns on, even if it makes noise. And, once set warms up hideous noise goes away.
I checked the amp board (since it has a separate one) and the noise is present on both input & output. Ruled that out for now then. So, I checked the power supply. There's some ripple at about 420Hz, but it's only about 100mVp-p on the 24V rail so I disregarded that.
I then went to check main board. I tried freeze spraying a few things and once I hit an LM2576T switching regulator IC223 the noise got worse...hmm...
Checked the output, 1.77V DC on the DMM so seemed OK (rated 1.8V on schematic), maybe slightly low but nothing to write home about. So used my scope on it and I find there's about 300mVp-p noise riding on it at 30kHz. In addition to this there's a lower frequency harmonic at 7kHz being modulated into this - the same frequency as our speaker noise. Freeze spraying it alters this noise and the tone heard from the speakers.
No wonder the CPU crashes sometimes! It's getting a supply voltage from 1.6V to 1.9V at 30kHz, it's amazing it works at all!
Replaced capacitor C905 and the problem was fixed, both noise and glitch art, it is near IC223 and is the input capacitor for the converter. It had ESR of 34.4 ohms!
Moral of the story is that just because a switching converter has a good output voltage doesn't mean it's working!
Comment