Possible problem with BattBorg

Submitted by hardrain2@comca... on Mon, 05/30/2016 - 22:13

Forums:

Hi,

I've been using a Pi 3 for the last couple of months and a Pi 2 for the previous to that, 9 months. With either board the Red Pi power light flashes form time to time and a lot more lately. I have added new stuff LedBorg, XloBorg, Headlights, but the flashing started before this. I've seen it flash as many as several time in a second or two and dozens of times over say an hour. Now I searched the Web to try and understand the operation of the Red Pi light and all I can find is that it should never flash.

OK, so recently and several times before in the past, I've measured the Pi voltage on the Pi board. The 5 volts, start at 4.95 volts and in 5 seconds or so drops to 4.81 – 4.82 and an occasional flash to 4.75 volts. I believe the spec for the 5 volts is +- 5% . The the voltages I am reading are very marginal according to the spec. So this has been going on for at least 6 months maybe more (just not paying attention back then). I also measured the 3.3 volts and it measure a steady 3.1 volts. So I believe that every time I see the Red light go out, the voltage is to low?

Now I am also using the UltraBorg which is powered by a separate 4 AA battery pack. I have removed the 5V link jumper as described in your documentation.

I use two main power sources, 10 AA Duracell Alkaline non-rechargeable Batteries and a 12 volt 3 amp power brick for testing, that plugs into the AC outlet. Both of those sources produce the same voltage readings.

So what do you think?

Thanks

piborg's picture

piborg

Tue, 05/31/2016 - 09:10

We have seen this before, usually when the Raspberry Pi is doing work which requires a large amount of power quickly (like 3D rendering).
To date we have not seen it cause a reset unless the BattBorg is genuinely overloaded.

Could you tell us:

  1. What voltage do you see from the BattBorg when it is not powering anything?
  2. If you can measure it, what current is being used?
    Let us know where you measured as well :)
  3. Do you still see this flashing afer the Raspberry Pi has booted up if you let it sit there doing nothing?
  4. Are you seeing any other problems (such as resets), or just the power LED flashing?

hardrain2@comca...

Wed, 06/01/2016 - 03:32

Hi,

My Pi 3 Configuration is using an Adafruit Perma-Proto HAT with No EEPROM. This hats is Configured with some resistors for my 6 Head/Rear Led lights, an Led for my 5V PS indicator for the UltraBorg and another resistor connected to pin 40 for the OS shutdown pushbutton. Now on top of the Adafruit hat is your TriBorg+ hat with an LedBorg on one side and your xLoBorg on the other side.

Now this leaves the middle row of pins to which I have connected the UltraBorg cables (2 three pin cables to pins 1,3,5 and 2,4,6) which are daisy chained from the Battborg. So, to reiterate, BatBorg to UltraBorg to TriBorg+ hat resting on top of Adafruit Hat attached to Pi 3 board.

So the voltage/amp are taken from pins 2 and 6 (From your Triborg+) for voltage and in series (the meter) with pin 2. Pin 4 was removed and not connected to the Pi so all current would pass through Pin 2.

Now you asked for the voltage readings when the Pi is not powering anything. This would difficult because it would require removing the hats and cabling and wiring, and from experience, stuff happens when you do that kind of thing. So I can only give you everything connected to the Pi with the OS running, but with no Borg software running.

Testing

Turn on, boot Pi, red light went out several second, green light flashing, red light back but flashing a lot and alternating with green light. Current .30 -.50 Amps

Made remote connection to Pi using TightVNC, Red light out, (same as boot) then flashing alternately with green light . 0.3 amps. Red light continuously flashing now, seems to happen a lot with Network activity.

With PiBorg script running, Motors turning, LedBorg lights, Headlights, and XloBorg just plugged in .8 Amps. Red light no longer appears. Stopped Motor, now .40 Amps with Motors stopped and lights out. Camera still streaming.

So highest amperage was 800 milli amps.

Shutdown OS, .08 Amps, Red light Back. Turned off. Will now go back and measure voltage

Power on, Booted up, Red light in and out, settled down Red Light back on steady. 4.88 Volts.

Remote connection, Red light blinking Voltage 4.75 for a couple seconds, then stable @ 4.88 again. Red light blinking some.

Start Borg script, Red light blinking a lot, then less blinking, then just a little blinking 4.87 volts

Now start motors, lights, everything. Ops! voltage cables knocked off, no camera image. Shutdown.

Ok, start again, booting, red light flashing. Red Stable now, 4.88 Volts, Camera Back (reseated cable connector).

Make remote connections, Red Flashing, 4.75, 4.76 … 4.88 V, then eventually stable.

Start Borg script, no motors or light, Red light blinking a lot, 4.82 Vols. Code is just looping through Manual mode, waiting for command to move. Red light blinking.

Red light still blinking. Bring up Browser make Borg connection, Red light blinking away. Start Roaming command, motors running, Red light mostly out, 4.77 Volts solid. Turn on all lights 4.76 Volts.

Stop motors, turn off lights, 4.83 Volts. Red light blinking.

OK done with testing.

Note:All network connection were Wi-Fi. Connected to PiBorg using remote terminal (TightVNC) on Linux Mint 17.3 system. Browser testing is done with Firefox 46.0.1. Also, the UltraBorg Link jumper had been removed when I first installed it. All 12 Volt power came from 12 volt 3 amp power brick that plugs into the AC outlet.

So bottom line, This problem has been getting worse over the last 6 months or so. Voltage is fluctuating between 4.75 to 4.82 volts a lot (some from previous testing) with everything running. The current peeked at 800 milliamps (lower then I expected). Network connections put a hit on the Voltage levels, blinking and 4.75 Volts. Since the 800 milliamps is with the 1.5 amps spec of the BattBorg, I'm thinking the problem may be with the voltage regulation. The flashing RED is getting much worse in the last few weeks, this week, with it going out for several seconds at a time now.

piborg's picture

piborg

Wed, 06/01/2016 - 10:36

As far as I can tell there are two likely candidates here:

  1. Something attached is dragging the power down at brief intervals
    The multimeter will not be able to see these fluctuations
  2. The BattBorg is not producing a high enough voltage, giving marginal results

Can you disconnect just the BattBorg and measure the voltage output when it is on its own?

hardrain2@comca...

Thu, 06/02/2016 - 23:49

In and effort to isolate the current problem, Red light Flashing, low voltage levels, I'm looking at different things now. I'm trying to reduce the number of functional component as best I can (basically as you requested, run the Pi with nothing connected).

1. I have not mentioned this, but I recently added 2 new distance sensors to the UltraBorg. This is the first time I have had 4 sensors running at the same time. The sensor connected to position 3 is not working correctly (keeps blinking on and off with incorrect values) I have checked this many different ways and its bad. I've swapped the sensor with a different one, I swapped the cable, I swapped a working sensor and its cable to position 3 and took the sensor and cable form position3 and moved to a good sensor position. The problem persist with the sensor plunged in position 3. So for this test I removed the cable and sensor form position 3.

Test complete. I'm just leaving this at checking the Blinking. The lights blinked when booting and blinked when remote connection initiated. Didn't have the meter hooked this time.

2. Remove the LedBorg and xLoBorg and booted up and made remote connection. Same results as previous testing Red Light flashing and voltage got a little lower to 4.71 Volts.

3. Ran the same test as 2 but did not turned on UltraBorg 5 Volts PS. Same results, voltage jumping around 4.70-1 more often. Wish I had a scope to look at this.

4. Disconnect All UltraBorg Power Cables. Cables coming from PicoBorg connected to nothing show 5.03 Volts. I think this is effectively the BattBorg output.

5. OK, so I connected the PicoBorg Cables to the Pi pins 2,4,6 and 1,3, 5 directly to the TriBorg+ pins and booted. Wow! Voltage @ 4.95-4.99 stable at 4.96Volts. Red Light not blinking even after remote connect.

6. OK so at this point it looks like either a device on the UltraBorg, the UltraBorg it self or Bad cable connection between PicoBorg and UltraBorg. Connected the UltraBorg back in and removed the cable to the second new sensor. Booted OK with no blink. Made remote connection and a blink or two. Voltage was low as 4.88 but stabilized at 4.93. Note: 2 sensor are still attached. Note: Voltages lower now since connecting to UltraBorg.

7. Attempted to run my Pi script and it complained about the xLoBorg init. So shutdown and re-installed xLoBorg on the TriBorg+. Booted up, Voltage as low as 4.83 briefly but became stable at 4.93. Remote connection became stable at 4.93, no blinking. Start Web Program, Stable 4.91V. No Motors. Brought up Browsers and made network connection to Pi, still good @ 4.93V. Start Motors, ran OK 4.89-4.90V. Move Camera servos, worked, stable 4.90-4.91Volts. Shutdown.

8. OK test to see if the problem is the sensor or the connector slot. Reinstalled second new sensor in slot 2. Booted, one blink, Volt 4.88 to Stable 4.93. Expecting this to fail. Made remote connection, 1 blink and 4.88 – stable 4.93. Ran Web program OK but sensors are bouncing all over the place.

9. Add back first sensor slot 3. One blink on Boot 4.85 then stable 4.93. All 4 sensors back, jumping around all over the place. Sensors Blinking None then 2000-3000, then none…..

10. Time to put back the LedBorg. Booted, remote connection, ran Web program. With no Motors running, 4.88 -4.90 Volts.

11. So every thing connected now but voltage is lower then it should be. Let go back and remove all sensor now (before I had just removed 2 trying to check out a specific sensor position 3. Booted 4.88 became stable at 4.93-4.94. Made Network connection 4.88-4.93-4. Shutdown. Well forgot to turn on UltraBorg PS. After shutdown 4.99V. Turned on UltraBorg PS and rebooted. Became Stable @ 4.93V Shutdown.

12. Remove the two servos. Rebooted, Voltage Briefly to 4.99 ...them 4.85….Stable 4.93. Voltage not as good as Test 5. Shutdown. So all servos and sensors are remove now. Voltage decrease when UltraBorg attached with no devices. Voltage drops as low as 4.85. This may be a problem.

13.Add back sensor to position 3. Booted Voltage briefly started at 4.99 ….then bounced 4.85 eventually back to 4.93. Remote connection Stable 4.93, Start script, sensor much better than before.

14. Shutdown, add another sensor. Reboot 4.85 ...to Stable 4.93-4. Ran Web program 2 sensors, a little shaky but maybe OK. Voltage down to 4.90-1.

15. Shutdown and third sensor. Again initial power on starts with 4.99 Volts and goes down to 4.93. After shutting down OS Voltage at 4.99 again. Remote connect, start program. Sensors reasonable, V4.90-1.

16. Shutdown, add 4 th sensor. Reboot Voltage 4.99 >4.93. Remote Connect, Start program. Reboot 4.99 > 4.86-7 >4.93. Remote connect, Start Web program. Wow! All 4 sensor stable now, how?. Run motor 4.88-9 and LedBorg lights, headlights still around 4.88-9.

17. Shutdown, Reboot, Add back 2 camera servos. Reboot, 4.93 Volts. Program Running 4.90, Motors (4) running 4.84-4.90

Conclusion:

1. The Red light requires a minimum voltage to stay on. Don't know what that voltage is but voltages of 4.85 Volts or higher keep the RED light ON. Almost no blinks after test 5.

2. Disconnecting the PicoBorg cables from UltraBorg in Test 4 show 5.03 Volts!. Reconnecting the cables to the Pi (Triborg) resulted in Voltage @ 4.95-4.99 Test 5. Then reconnecting to the UltraBorg with no device Test 12 resulted in Voltages as low as 4.85 and later stable at 4.93. I think the voltage increase shown in Step 4/5 may have simple been a bad connection to the UltraBorg from the PicoBorg or perhaps a bad connection between the wires on the UltraBorg to the TriBorg.

3. Now there may still be a problem with the UltraBorg as connecting to it from PicoBorg reduced the voltage in all subsequent tests ranging from 4.86-7 >4.93 Test 16. The final voltage with everything running is 4.84-4.90 Test17. Seems like they should be higher like ~4.95 Volts

Final Thought – Red light not blinking, Sensors stable, Voltage marginal, UltraBorg suspect

Thanks for all your Patients

piborg's picture

piborg

Fri, 06/03/2016 - 12:17

Well it is good to hear you have made some progress :)

It sounds like the cables connecting between the boards are a likely culprit here.
The total length is possibly adding enough resistance to cause significant voltage drops.

If you can it might be worth making some cables for connecting the 6-pin headers.
Thicker and shorter wires will have less resistance and may reduce the voltage drops you are seeing.

Subscribe to Comments for "Possible problem with BattBorg"