Problems charging LiPo with Sunny Buddy

Components & Accessories Power
1

Hi,

I’m having some issues with a sunny buddy board that I’m looking for some help with. The short version is that I can’t seem to get the board to charge my LiPo consistently. Occasionally the board will charge for 5 - 25 minutes or so before stopping, but the vast majority of the time it refuses to charge.

The parts I’m using:

  • Sunny Buddy PRT-12885

  • 2Ah LiPo battery RPT-13855

  • LiPo Fuel Gauge TOL-10617

  • 9W 6V solar panel PRT-13784 (this was not purchased from SFE, but I’m using this exact panel)

All of that is connected to an Artemis Redboard so that I can read the LiPo fuel gauge and also read the state of the charge and fault pins on the sunny buddy. (The Artemis RB itself is not currently being powered by the battery.)

I’ve adjusted the voltage at the SET pin on the sunny buddy to be 2.85V with the panel and no battery or load connected. I have not altered the default charge rate.

Symptoms:

When I connect the solar panel to the sunny buddy sometimes (very infrequently) it will begin charging (I can see the charge pin go low, and the voltage reported by the fuel gauge goes up). This will continue for anywhere between 5-25 or so minutes and then charging will stop (again verified with the charge pin and fuel gauge voltage). Throughout this the fault pin never goes low, so it doesn’t seem to be indicating any fault. At this stage there seems to be nothing I can do to get it charging again. I’ve gone so far as to remove all wiring, disconnect the battery, etc., but the behavior doesn’t change. Nothing seems to trigger it to go into charging. This is in full Colorado afternoon sun. The state of charge of the LiPo also seems to have no impact. I get the same results whether the battery is 65% SOC or 2%.

And I want to emphasize that it only starts charging every once and a while. I spent hours with this yesterday and was only able to get it to start charging 3 times after dozens of disconnect and reconnect iterations. It seems to be very random.

Troubleshooting steps I’ve done:

  • Verified the panel is outputting 7.15V in full sun with nothing connected to it

  • Verified the battery has no problems being charged with a different charger (used the SF adjustable charger PRT-14380)

  • Have connected various standard 6-7.5V wall chargers in place of the solar panel and get the same results: charging begins very infrequently and doesn’t last more than ~25 minutes at most when it does start

  • Have disconnected all other items so that only the solar panel and battery were connected to the sunny buddy and left it to charge in full sun for over an hour and noticed no increase in battery voltage

  • Verified I’m always plugging the solar panel into the barrel jack on the sunny buddy and not the artemis :slight_smile:

I’m not quite sure what I should check next… Any ideas?

Thanks!

Chris

2

Hi Chris,

I’m curious if there is a weird connection problem going on here since it is working intermittently. Do you have a multimeter handy to verify there are no loose connections on the board or in the rest of your circuit? Along those lines, can you take a few photos of your Sunny Buddy and the circuit you have it in? Please do your best to make sure they are clear and well-lit.

3

Hi Mark,

Thanks for the reply. I do have a DMM to verify things. Let me know if there are any specific measurements you’d like.

I’m having trouble attaching pics to this post so I’ve uploaded them here: https://subalpinedesign.com/sunnybuddy/

The sunny buddy is connected to the artemis redboard and proto shield via the twisted red/back wires that then connect to the LiPo fuel gauge JST connector. It is also connected via the 4-pin header with the white/yellow/green/red ribbon cable. The white and yellow wires are connected to the charge and fault pins on the SB, and connect back to GPIOs on the artemis. The red and green wires in the ribbon cable are not connected past the header on the proto shield. So the only ground between the sunny buddy and shield is via the ground fed to the fuel gauge.

The ground from the SB to the fuel gauge and rest of the shield and artemis RB combo is always connected. And the fuel gauge I2C pins are also connected to the artemis. The toggle switch on the shield controls VBATT out from the fuel gauge and into the 5V pin of the artemis RB. This way I can toggle the switch on to feed power in and have things run remotely off of the battery, or toggle it off to only be reading battery state via the fuel gauge and charge and fault IO pins (and power the artemis via USB).

The other pics show the sunny buddy disconnected from the rest of the circuit. I’ve done tests with the SB connected to the circuit and disconnected. I haven’t had success charging the battery even with the rest of the circuit disconnected (only the battery and solar panel were connected to the SB).

Let me know if there are any other pics that might help.

Again, thanks for the help with this. I really appreciate it.

Chris

4

Hi again Chris,

Thanks for taking those photos and providing information on how you have already tried to troubleshoot the issue. My guess is the LT3652 IC is the culprit here but to try and verify that, if you can probe the VBatt line for voltage (you can also open JP3 to test for current through the + pin on VBatt) and let me know what you see there both with and without a battery attached that would be very helpful. If you want to, you can also test the various pins on the LT3652, specifically VIN, VIN_REG and VFB. The LT3652 is oriented on the Sunny Buddy where pin 1 is at the top left of the IC if you are looking at the Sunny Buddy with the SparkFun logo oriented upright. There is a pinout diagram for the LT3652 in the [datasheet if you want to test those other pins.

Most likely, one or more of those pins are going to be putting out something out of whack. If not, my guess is the inductor is the culprit. Regardless, if you can test those and let me know what you find we can move forward with your case. Everything in your circuit checks out just fine so I think the board is just bad and if that ends up being the case, we would be happy to replace it for you.](https://cdn.sparkfun.com/datasheets/Prototyping/LT3652.pdf)

5

Hi Mark,

I was able to take some measurements over the weekend. In the process of taking those measurements and checking the LT3652 I now have some doubts about some of my original tests…

Like I said in my first post, when I first got started with this I took the sunny buddy and my solar panel outside to adjust the SET voltage to 2.85V in full sun. Once I started seeing the charging troubles I brought the setup inside to use a 6V regulated power supply in place of the solar panel for easier testing. I never re-adjusted the SET voltage when I changed from the panel to the normal PS.

Before starting these tests I checked battery voltage and what my power supply was outputting:

Starting battery voltage: 3.807V

Power supply voltage: 6.30V

So this first round of measurements was using the 6V supply, but without any changes to SET voltage. And all tests have been done without any connections to the artemis.

    Without Batt | With Batt:
VBATT   = 1.577V | 3.807V
VIN     = 6.299V | 6.30V
VIN_REG = 2.36V  | 2.36V
VFB     = 1.269V | 3.058V

After seeing these numbers and checking the datasheet I was suspicious about the SET voltage, so checked it and got 2.36V. So I reset the SET voltage with my power supply so SET was exactly 3.0V and remeasured everything.

    Without Batt | With Batt:
VBATT   = 3.917V | 3.865V
VIN     = 6.297V | 6.245V
VIN_REG = 3.00V  | 2.978V
VFB     = 3.147V | 3.108V

As you can see it is now charging the battery when the battery is connected. VBATT is increasing and the charge pin is low.

I monitored the charging and VBATT steadily increased until VBATT got to around 3.9V, and it didn’t increase much from there. The highest VBATT I’ve seen is 3.928 (which is really close to what I measured on VBATT without the battery being connected). It has been hovering at right around 3.9V on VBATT for close to 20 hours now. Throughout this time the charge pin has almost always been low for probably 95% of the time.

So it appears to be charging, but only up to a VBATT of ~3.90V?

Obviously I still need to go back to measuring with the solar panel, but I thought I’d pass on what I’ve found so far. What do you think?

Thanks again for all the help!!

Chris

6

Hi again Chris,

Thanks for performing all of those tests and providing the results. I’m not sure why it would be stopping at 3.90V on VBATT since the circuit should set the float voltage to ~4.2V (or very close to it depending on resistor tolerances). I’ll try and do some tests next week with a known good Sunny Buddy but I’m leaning toward something is wrong with your Sunny Buddy. Keep us updated on anything you find out and I’ll try and test this further after the holiday.

7

Hi Mark,

Hope you had a good Thanksgiving. Over the holiday I was able to drain the battery and start a new charge cycle from that baseline. The behavior was similar to before, where the battery continued to charge up to ~3.98V at which point it stopped charging and settled into a voltage right around 3.90V.

Since you mentioned the resistors in your last post I went ahead and measured the 5 resistors that are to the bottom-right of the LT3652 (above the ‘Sunny’ part of the Sunny Buddy silkscreen). The 2 on the left measured 220.6K (top) and 62K (bot), the middle measured 39.4K and the 2 on the right were 9.96K (top) and 9.93K (bot).

Let me know if there is anything else I can measure on this end.

Thanks,

Chris

8

Hi again Chris,

So I think I may have been mistaken in my previous response. After some testing here and talking to some other folks at SparkFun, I think that ~4V is about the best you’ll get with the Sunny Buddy unless you want to modify that float voltage circuit. A quick test with a known good here produced a max charge voltage of 3.97V. Unfortunately, the engineer who designed the board no longer works at SparkFun so I cannot say with 100% certainty, but I believe he designed it so that we could account for resistor tolerances that may push that voltage over 4.2V. Designing the circuit to set that voltage to exactly 4.2V could result in some unsafe behavior from the charge circuit if one or more of the resistors in that array were slightly providing more or less of their listed resistance.

Charging to ~4V should be just fine unless you need a full charge for any reason. If it is just a matter of monitoring the battery and reporting a “full charge” you can just add in a quick formula to turn whatever percentage the LiPo Fuel Gauge is reporting to equal 100%/full charge. I hope this helps explain the behavior you are seeing with the Sunny Buddy and lets you move on with your project. If you have any other concerns or issues with the Sunny Buddy or other SparkFun parts in your project, let me know and I would be happy to help.