Something wrong below 0.5V in DC

[Macross]

I just purchased the DSO Shell and installed every thing as described in the manual. Coming to the calibration process, I couldn't made it as something went wrong. Switching to DC, I could see the test signal (0.1V) upto 0.5V scale when move to 0.2V scale the signal become flat. Therefore I couldn't calibrate the unit at 50mV scale as the manual said. I tried switch the test signal to 3.3V and you could see the result as in the attached pictures. Moving from 0.5V scale to 0.2V but the signal didn't enlarged as it should. Anyway, I could see the test signal (0.1V) all the way to 50mV if I switch to AC instead. Any help please.

[qu1ck]

Try V/Pos realignment: switch input coupling to ground and long press sensitivity button. Then try the calibration process again.
I had very similar symptoms and found that I missed this step. After I performed it per instructions, everything worked.

[Macross]

@quick, thank you for your help. I did try that. Setting the coupling to gnd and long press sensitivity button, I could see a line center of the grid. When switch back to DC, unlucky me, it couldn't solve my issue. I am preparing videos. Hope it could help fixing my issue easier.

[Macross]

Some Videos
https://www.youtube.com/watch?v=Pzo0r4VVIUQ
https://www.youtube.com/watch?v=Q2SjAiwbBlE
https://www.youtube.com/watch?v=Vykdzxi6kE0

[qu1ck]

I'm pretty sure that what you observe is the normal operation. My unit behaves the same way. On your first video you show that waveform is displayed properly on AC coupling for 0.1v test signal and high sensitivity settings. That's how I adjusted the compensating cap, on AC coupling, not DC.
Since the 0.1v signal has 3.3v DC offset I think you are supposed to use AC coupling. Otherwise on high sensitivity and DC coupling the waveform won't fit on the screen. The on screen readings are off when the waveform doesn't fit the screen and that can be improved in software (also the trace should disappear instead of being flat in the middle of the screen) but I don't think that is a malfunction per se.
I'll let jye1 confirm my theory.
Also hopefully firmware can be corrected to not show fake values in that situation.

[KSV]

Have the same situation, see my posts at the neighborhood topic:
http://www.jyetech.com/forum/viewtopic. ... 2&start=10

It looks like we have the high absolute voltage at 0.1V curve generation: 1150 mV.
This doesn't fit at screen at DC mode with 50mV sensivity. When we use AC, absolute value is dropped by capacitor (immidiately after BNC input) and only variable path is left. Then we can see the curve.

Hope for the final clarification of this moment by JYE, with the possible ways to solve.

[Macross]

@quick, @KSV thank you very much for your explanation.

[jye1]

We can confirm that the 0.1V test signal contains a +3.3V offset due to a bug in firmware. This will be corrected in the up coming firmware release. For the time being please use AC coupling for the adjustment of C3. We are sorry for the mistake.

[qu1ck]

jye1
That is great, thanks!
Can we also fix the issue with wrong values displayed when the input doesn't fit on screen? Preferably the trace should disappear as well.

[KSV]

We can confirm that the 0.1V test signal contains a +3.3V offset due to a bug in firmware. This will be corrected in the up coming firmware release. For the time being please use AC coupling for the adjustment of C3. We are sorry for the mistake.

jye1, thank you for the information! Good news that the oscilloscope part is good, problem with the test signal is not so important. Mostly needed for initial tuning.
Please, notify, after the firmware fix release. Hope we can fix this on existing board.