Assistance requested for repair of KiwiSDR 1 [fixed]
A member of my local radio club gifted me his old KiwiSDR 1, which he reported that it was a little “deaf” when he took it out of service. It’s a Seeed Studio unit in a nice metal case, with serial number 35xx (exact number hidden to preserve security) and looks in good condition. However, my attempts to bring it to life have failed.
When I provided +5v from my bench PSU to the barrel connector on the fully assembled KiwiSDR there was 0mA current draw, and no led activity at all on the Beaglebone. There was zero activity on the beaglebone’s UART0 serial debug port. 5v was reaching the 5V_EXT edge connector to the beaglebone, but there was only 0v on the return 5V_INT and 3.3V pins.
I then ran some tests with the beaglebone with the KiwiSDR board unplugged. Power was supplied to the beaglebone via its usb socket and the usb cable “teed” so that +5v was from my bench PSU, a known good high current power source. The UART0 serial debug port showed that the beaglebone had successfully booted. I could log in via the UART0 serial port, and also ssh and ftp into it via the ethernet port. The kiwid app was running, but the web server on port 8073 didn’t respond, which is perhaps not surprising given that the KiwiSDR board was absent.
Finally I reconnected the KiwiSDR to the beaglebone, and powered it via the “teed” usb connection as in the previous test. The blue power led came on for a fraction of a second but there was zero activity on the UART0 serial debug port and 0v on the KiwiSDR’s 5V_INT and 3.3V pins.
My conclusion so far is that the beagleboard appears generally functional, and that there is something untoward with the KiwiSDR.
I would welcome suggestions on how to further diagnose the problem.
Rgds,
Dave
Comments
Hi! The fact that the Beagle board does not start with the KiwiSDR board connected indicates that there may be a short circuit on the KiwiSDR board via the 5V_INT circuit. Also check the 3.3V circuit for a short circuit. The resistance of the 5V_INT line should be tens (or hundreds) of kilohms per ground. The 5V_IN line supplies 4 power stabilizers. If the resistance along this line is low, one or more stabilizers are most likely out of order.
Thanks for the suggestion.
I've checked & there is no short circuit on ether the 3.3v or 5V_INT lines.
I briefly connected 3.3v to the 3.3v circuit & there's <10mA current drawn.
There's approx. 250mA when I briefly connect +5v to 5V_INT. The output of U701 is +1V, the output of U702 is 1.8v, the output of U104 is 3.3v, output of U404 is 3.3v - all as expected.
I can't see anything wrong here?
Cheers, Dave
Yes. That's how it should be. Current consumption and voltage are normal. I didn't quite understand, do you supply power to the Beaglebone board via the USB connector or via the barrel connector?
I've tried both ways to supply power.
Via the barrel connector was the first test I described in my original post (result: no current draw, no led activity, no UART0 activity).
Via the USB connector was the last test in my original post (result: no current draw, brief blink of blue power led, no UART0 activity).
Sorry for the stupid question, but are you sure you connected the KiwiSDR and Beaglebone board correctly? The main Beaglebone power controller must supply 5V_INT power to the KiwiSDR board and to the Beaglebone 5V_SYS board if the supply voltage is normal, the voltage rise is not too slow and the current consumption is not exceeded.
Hey, there's no such thing as a stupid question - especially for this weird problem.
Yes I'm 100% sure the two boards are connected correctly:
I've rechecked it with power provided via the barrel connector.
I've confirmed 5V gets to the 5VE pin on the connector to the Beaglebone. But there's no voltage on either the 5V pin or the 3.3V returns from the Beaglebone. 0mA is being drawn on the +5v plugged into the barrel connector. The blue led doesn't flash as I plug the barrel connector in & there's no output on the UART0 debug output. I plug the connector into the KiwiSDR rather than turn the PSU on to ensure there's a fast voltage rise.
I've also confirmed that my bench PSU is capable of 3A.
Well, for some reason, the TPS65217 on the Beaglebone board doesn't want to pass the 5-volt power on. I don't know why this might be happening. Try checking the other GPIO pins on the KiwiSDR for short circuits.
Have you read the power supply section of our documentation? http://kiwisdr.com/info/#id-power-undervolt
You can have a perfectly adequate power supply but a power cable that has too much voltage drop (IR^2 loss) such that there is a momentary voltage drop at power up (due to inrush current) causing the peak voltage to be less than 4.75V at some point. The power controller chip (PMIC) on the Beagle will see this and refuse to power up. This can happen even though you see a steady 5V at the 5VEXT pin on the header since it is a transient problem.
You really need a triggered scope monitoring the 5VEXT point to see this issue. Try a 2A phone charger with a really short cable just to see if you can get it to go.
It's also possible the T301 common mode choke is burned up causing voltage drop. See here: https://forum.kiwisdr.com/index.php?p=/discussion/2978/kiwi-1-how-to-fix-a-shorted-5v-input-cmc-t301/p1
I have good news and bad news.
The good news is that my KiwiSDR is now working! Many thanks to jks and studentkra for your guidance.
The bad news is that I've embarrassed myself! Despite reading heaps before posting here I completely missed the power supply section of the documentation. Oops. Also, having read that section I measured the resistance of the power cable I had recycled from another product - 1.5 ohms! That's hardly better than wet string! I replaced the cable with one with <0.2 ohms resistance and the KiwiSDR booted perfectly.
Thanks again for your help. I'm now on to reading ALL the documentation to learn how to set it up properly.
Congrats. I would say that's good news all around!