It looks like you're new here. If you want to get involved, click one of these buttons!
Hopefully not, but if so it would be good to have the 'S-meter OV...' setting on the /status page
Related question: I have three KiwiSDR's fed from a 4-way mini-circuits splitter, two with BBG's and one with a BBAI processor board. I've noticed that the adc_ov count climbs on the BBAI Kiwi (41 in a day or two) while the two with BBG's stays at zero, or climbs only a few counts over several days. All are running v 1.566 firmware. Dropping the input gain by an additional 3db reduced the count frequency on the AI, and the other two stayed at or around zero adc_ov
Is there any code differences or processor capabilities of the AI version that could account for this difference in ov sensitivity, or is it more likely that component tolerances on the Kiwi that happens to have an AI board attached are a factor? S-meter readings appear identical (as far as I can see) between boards.
does the OV count follow the wd.conf?
This adc_ov count is being displayed on the mykiwi:8073/status page, so wd not afactor.
Are the supplies to the BBG and AI similar?
How is the AI powered, what port, cable etc.
All linear. I measured the 5V and 5Ve pins on the kiwi header while in use 4.36, 4.46 vs 4.32, 4.27 BBG and BBAI respectively. Looked at the schematic and the A/D is set to have a 1 volt input range and is supplied by 3.3 volts. So I don't think the supply voltage variance at the header pins is a factor. Probably just component variances are the reason for different overflow levels.
I don't know that detail but the lower frequency overloads can (I assume) be affected by the complete path.
Noise can be injected by the power supply and that does depend on the type of antenna in use and how that is grounded so my thought was mainly about the AI supply would be providing less filtering and possibly allowing greater LF currents.
The AI I think is meant to be powered through the BeagleBone, not via the Kiwi filtering components, so it is not quite the same beast.
Are you sure the RF levels are all the same? Switch the feeds between the the Kiwis to see if the OVs follow the RF cables.
I adjust the 5 V PSU while measuring the 5V on the USB jack of the BBG. The AI draws much more power, so there will be a lot more drop between the barrel connector and the AI board, which is I think why John recommends the AI be powered through its USB-C port.
The other reason to power through the bbai is that the filter inductor on the Kiwisdr board runs way over spec when powering the bbai. It will eventually fail.