Power missing

la3rkla3rk
in Problems and Issues
My KiwiSDR has stopped working, just a blink on the blue LED. Checking the troubleshooting guide the following appears;
5V IN - ok, reads 5.17 volt
5VE - ok, reads 5.17 volt
5V (from BBG) - reads 0.67 volt, apparently a problem somewhere, should read same as 5VE. According to troubleshoot guide this is power back from BBG.

The BBG seems to start normally when KiwiSDR is plugged away.

Where is my problem? On the BBG or on the KiwiSDR itself?

Any good advice is welcome.

73 de Olaf - LA3RK

Comments

  • n6gnn6gn
    From your description, it sounds to me like the SDR PCB has a very low resistance path to ground and is pulling too much current such that the BBG sourcing it goes into current-limit. When the SDR is unplugged from the BBG what resistance do you measure on the SDR from 5V input to ground? If you find that to be too low, perhaps something in the few ohms region or less, you might try connecting a current source, a current-limited power supply set to 5V and 1A perhaps, and then use a sensitive voltmeter to follow voltage drop across PCB traces to discover where the current is going. That can lead you to the offending part of the SDR so that perhaps you can replace a shorted component. But if it turns out that the low resistance is inside a fine-pitch IC, you may have a challenge to replace such a part.
    At least, I think this would be my approach.

    Glenn n6gn
  • jksjks
    edited September 2020
    This doesn't make a whole lot of sense given how power distribution on the Kiwi and Beagle works. If you are measuring 0.67V on the "5V" pins of the Kiwi (P9-7,8) then the Beagle should not be working either. Are you really getting "double flashing" of the left-most of the group of 4 LEDs that are together? Both when the Kiwi board is attached and when it is not?

    I've included the BBG schematic below. 5V comes from your power supply, through the Kiwi board, to the "5VE" pins P9-5,6. On page 10 of the BBG schematic this shows connected to VDD_5V. Page 2 shows the Beagle power management chip (PMIC). Note how VDD_5V at top left goes to the PMIC as the "AC" source input. When the voltage on this rail is detected to be "good" it is automatically switched to the output directly across which goes to SYS_5V. Note further down on the left of the PMIC how SYS_5V is the input for ALL the subsequent regulators of the Beagle (with one exception) and generates VDD_MPU, VDD_CORE, ...

    Back on page 10 SYS_5V is connected to P9-7,8 which is "5V" on the Kiwi. So the Kiwi 5V is switched by the PMIC like the rest of the Beagle. If you are reading 0.67V on the Kiwi 5V then it should not also be present on SYS_5V unless there is a bad connection on P9-7,8 or burned out SYS_5V trace on the Beagle. When the Kiwi board is detached do you measure 5V on P9-7,8? (connect your multimeter probe to a piece of wire and carefully insert the other end into P9-7 or 8 until there is contact)

    It would be interesting to know the resistance of 5V to ground on the Kiwi board which detected from the Beagle to see if there is a short. Has there been any thunderstorms/lightning in your area recently?

    PDF preview. Click on icon below to download.
    image

    Attachments:
    https://forum.kiwisdr.com/uploads/Uploader/3d/ed6f67f5d4353c714ce3211c8eb9c9.pdf
  • la3rkla3rk
    Thanks for tips and feedback. Yes has had a nearby lightning strike, but no other equipment out and kiwi was not connected to antenna, but connected to power and ethernet. Has done the following;

    1 Resistance 5V on sdr is basically high, shows a low for some sec, cap charging, seems ok.
    2 Running BBG from usb power. There is no voltage at the P9-7,8. It seems that usb 5v also goes to pmic ac pin via a 0r resistor r168

    Will try to power the sdr directly and check current draw.

    The BBG seems to work fine, boots, blinks and responds to ssh calls.

    I may have a black beagleboard that can be used as a reference check.

    Olaf LA3RK
  • jksjks
    edited September 2020
    If there is no 5V on P9-7,8 then, most likely, a portion of the SYS_5V trace from PMIC to P9 has burned up. You might be able to spot it on the top or bottom of the PCB but it might also be on an internal layer.

    Will be interested to hear about the Kiwi standalone current draw. But the PMIC protects SYS_5V from excessive current draw in addition to voltage out-of-spec. Unless there was a catastrophic failure of one of the Kiwi regulators causing a short at the input? But then I would think the PMIC would have opened before the trace could have physically burned up. But that is too much speculation.

    Perhaps a BBG manufacturing defect? How do the P9-7,8 solder joints look? Might be worth reheating to see if that makes any difference.

    The USB_DC signal via is of course how the BBG is normally powered from the micro-USB connector when used standalone. Note that R168 is marked "DNP" (do not populate) and so doesn't really connect to VDD_5V but rather the pin 12 (USB) of the PMIC.
  • la3rkla3rk
    Solder joints on BBG looks fine. The BBG boots ok both on usb power and by powering the 5V pin P9-5,6. However there is no voltage out on P9-7,8. I am inclined to conclude that something has gone wrong on the BBG. Can not see any burnt traces or other visual signs.
    Resistance on KiwiSDR on the 5V power pin from the BBG is high, no apparent short on the KiwiSDR.
    Seems to be time to order an replacement of the BBG. Luckily not a costly item, only a matter of time to source. Not something found in local shops.
  • jksjks
    edited September 2020
    Very strange. As far as I can tell there are no components on the BBG between P9-7,8 and SYS_5V on the PMIC. It should just be a trace.

    If you're adventurous you could try adding a jumper wire. From the schematic: pin 2 of the U4 TL5209 3V3B regulator is the most easily solderable source of SYS_5V. See attached image. I think this would work.

    image

    Attachments:
    https://forum.kiwisdr.com/uploads/Uploader/f8/9e73c063ebc6ada5e47144962d5455.jpg
  • la3rkla3rk
    Yes could be done, fairly easy mod. However the connection between pin7 on the 3v3 regulator and p9-7,8 seems ok, zero resistance. 5V only on the 5VE pin on the kiwisdr indicates 250 mA current draw, ie not excessive and maybe ok as there is no control signals or other to start the kiwi with only the 5VE present.

    Has connected the kiwisdr to a black beagleboard, voltages seems fairly correct, but I did not load that beagle with the correct firmware as it was taken from another item.

    A spare green beagleboard has been ordered and, if that does not work, the kiwisdr may have a problem.
  • jksjks
    I'm sorry, what does "pin7 on the 3v3 regulator" refer to? Do you mean pin7 of BBG U2 the PMIC? (TPS65217C). I doubt that because the PMIC is a BGA package. Did you mean pin2 of BBG U4 the 3V3B regulator I mentioned previously?
  • la3rkla3rk
    Yes, refers to your picture and the 3v3b regulator. Thought you referred to pin 7
  • jksjks
    It's pin 2 (of 8), but it's pin 7 if you look at it upside-down. You said there was no 5V on P9-7,8 when sticking a wire into the header connector. But since you measure a connection between U2-pin2 and P9-7,8 does that mean, when the BBG is operating by itself, you measure 5V at U2-pin2 and the solder pin side of P9-7,8? I.e. is this a failure inside the P9 connector itself?

    If there is no 5V on U2-pin2 then the BBG shouldn't work at all as U2 provides the vdd_3v3b rail which goes to the Ethernet PHY and eMMC (flash filesystem). It also connects back to the Kiwi board as the main source of 3.3V on P9-3,4
  • la3rkla3rk
    Lets see, powering the BBG from USB gives;
    P9-3/4 = 3.3V -> OK
    P9-5/6 = 0V Not a surprise -> OK
    P9-7/8 = 4.92V -> OK
    Blue power led between USB host and ethernet connector are lit constantly and the status leds are blinking. As I read the schematic the USB power port is connected to separate input on the PMIC ie the USB_DC
    And BBG works normally, boot messages can be read from serial port. Not surprisingly it complains about missing sdr as BBG runs kiwisdr firmware.

    Powering BBG on the 5VE pin, P9-5/6 (as I understand the schematic this is the pin with power from the KiwiSDR) gives;
    P9-3/4 = 0V -> not OK
    P9-5/6 = 5.17V (power in) -> OK
    P9-7/8 = 1.16V -> not OK
    The blue led between the USB host connector and the ethernet connector lits up for 2-3 secs and goes out. P9-5/6 is connected to the AC pin on the PMIC chip. I need a magnifier and a steady hand to measure voltage on these pins.

    My conclusion, something seems to be wrong on the BBG and therefore a replacement has been ordered. Hopefully this will sort out the issue. It will take some time to arrive and in the meantime the KiwiSDR is "dead".
  • la3rkla3rk
    Just for info. Kiwi seems to be up and running again. Got a replacement and checked again. Seems my 5V power supply is not accepted anymore by the Beagles. Voltage is measured to 5.17V and the Beagles shut down after 5-10 seconds. Maybe the voltage is unstable.

    The Beagles started ok when running only from usb, but stopped when powered via the Ve pin, ie the same route as when powered by the Kiwisdr.

    Anyway time to change to a known good analog supply.

    Thanks for advices from those contributing.

    73 de Olaf - LA3RK
Sign In or Register to comment.