On power up, test signal is on and gives noise

[MatNieuw]

I´m using boards rev E and software version 062. I noticed noise on the input after switch on. It turns out the test signal is on at 3.3 level. If you press ADJ for 3 seconds, you can change it to 0.1 V. I presume that if you press ADJ a second time for 3 seconds, the test signal is gone completely.
Anyway, after power-up I don´t think the test signal should be on: it should always be off.
Suggestion for a next major overhaul: put the coupling switch next to the BNC connector, then the test signal terminal. Or move the test signal terminal to the bottom, near the power switch. Both of these of course require new side panels.

[MatNieuw]

If have looked into the noise sources a little further. I measured with a unused BNC connector on the input, to screen it somewhat. Causes of noise on my device, boards rev E and software version 62, iIn order of severity (the worst first):
1. Test signal is on at high level on power up. Use 1 ms/div to see it. Set test signal to 0.1 V to fix this. It would be better if the test terminal was at the power switch side, I think.
2. Pick up from the environment. It picked up mains noise. This disappeared completely when I put aluminium foil below the device and made it in contact with the BNC connector shell. What it needs is an internal shielding, there's quite good silver-based paint for that (EMI shielding) but very expensive.
3. Power supply. I compared a very clean linear lab power supply with an external MT3608 up-converter and USB power bank. The MT3608 is slightly worse, but not much. It probably can be fixed by extra filtering (coil + small C), the scope has a 10 uF C internally. The extra filtering could be inside the scope.

[eysi]

isn't an always-on test signal also bad for the performance of the STM32?