Board bring-up and testing

amcewen commented 2 years ago

We have three boards soldered up: A, B, and C.

Results from some initial testing...

amcewen commented 2 years ago

Test conditions:

Boards with a 2-pin header soldered onto J4 to allow us to stick a meter inline to measure current into the battery
My 22V/0.3A solar panel
One of @Tactile-electronics's stepper motors driven by a drill to simulate wind power
Bench power supply for more controlled variable power options

Board A

Seemed to charge via USB, but was putting ~230mA into the battery and the charging chip got exceedingly hot
Didn't work with wind or solar
Not long after the wind and solar tests, we noticed the diodes D5 and D6 seeming to allow current in both directions

Board B

Diodes checked and okay
Solar test done outside, but with no charge to the battery
Charging LED (D4) comes on
Bench testing showed no charging LED below ~4.1V, but also no charge even at higher voltages
After running the board C tests, I tried switching back to this board with the bench supply
Then at 5.5V I started getting a tiny bit of current into the battery: ~0.12mA
At 8.5V charge into the battery was generally around 70mA, but sometimes ramped up as high as 120mA
At 9V, after leaving it running at 8.5V for a couple of minutes, it started hitting the current limit on the bench power supply, and dialling that up I got to ~580mA going into the battery. At that point the charging chip gets warm, but nowhere near as hot as Board A, according to @huffeec. The Done LED is lit, and the charging one is off, but it's still happily pushing 440mA into the battery...

Board C

Diodes checked and okay
No current draw from the bench supply until the voltage got to 8.5V
At 8.5V and 72mA the charge into the battery was 80-90mA
Above 9.2V and 93mA, charge into the battery was 103mA and the Done LED (D3) came on but wasn't solid (shimmered rather than flashed)
At 16V and 90mA, charge into the battery was 130mA and the Done LED was sometimes out, sometimes shimmering
At 21V and 84mA, charge into the battery was 133mA and the Done LED was out.

Next tests?

Cut one of the boards in half to test the solar without the wind capacitors, etc. in play
Build a data-logger (I have some INA219 current/voltage sensors we could use) so we can do some longer-running experiments
Run a wind turbine with the data-logger, with varying load resistors rather than the charging circuit, to better understand the range of voltages and currents that those steppers can generate
Debug board A to get it to the same place as boards B and C. Might just need new diodes?
Testing a depleted battery, rather than one which was 3/4 full would also be useful
Try a longer-running solar test. Maybe it was still charging the capacitors first with our short outside testing?

I think some of the varying behaviour is down to the charging chip trying to manage/monitor the battery. Hopefully some data-logging tests would give a better picture of that (particularly if we log performance on a bench supply).

amcewen commented 2 years ago

It is possible to push 1.2A into the battery, at ~9V from the bench supply. At that point the chip is getting rather hot, and probably the mosfet is hotter. Think it'd be worth putting the FLIR camera on it if we repeat that experiment, and see just how hot it gets :grin:

amcewen commented 2 years ago

Did a basic long-term test yesterday with the solar panel connected indoors but in full August sun in a South-facing window

11:00 - 0.9V across the solar panel connection; 3.85V in the battery
14:30 - no obvious changes :confused: I did disconnect the solar panel and check its open-circuit voltage, which was 20V
17:30 - still no noticeable change. The output voltage from the charger (the screw terminals next to the battery connector) did have a voltage that was fluctuating across a range of values, so I think it was at times coming from the battery and at other times coming from the solar panel. Not sure exactly what that tells us, other than that the charger circuit is doing something

amcewen commented 1 year ago

I chopped board C in half to see if it was the caps on the rectified circuit which were causing the problem with it not charging from solar. Then when setting up a test I noticed that the solar panel had the polarity switched :facepalm: Correcting that resulted in the charging LED coming on as expected.

From the test, initially in bright South-facing sunlight:

10:00 - 3.8V. Solar panel was providing ~8V in circuit and ~110mA
13:00 - 3.83V. Had stopped charging as the panel was no longer in direct sunlight. However, when I checked the voltage across the panel it was ~4V, and checking it started it charging again. I didn't check the current then.

Tactile-electronics commented 1 year ago

Hi Adrian Thanks for the update - trying to get my sleep deprived brain around it.. does it mean that the board might be fine and it was the panel all along? Or were the capacitors causing a problem?

Thanks Laura

On Fri, 21 Oct 2022 at 10:21, Adrian McEwen @.***> wrote:

I chopped board C in half to see if it was the caps on the rectified circuit which were causing the problem with it not charging from solar. Then when setting up a test I noticed that the solar panel had the polarity switched 🤦 Correcting that resulted in the charging LED coming on as expected.

From the test, initially in bright South-facing sunlight:

10:00 - 3.8V. Solar panel was providing ~8V in circuit and ~110mA

13:00 - 3.83V. Had stopped charging as the panel was no longer in direct sunlight. However, when I checked the voltage across the panel it was ~4V, and checking it started it charging again. I didn't check the current then.

— Reply to this email directly, view it on GitHub https://github.com/Tactile-electronics/windmill-electronics/issues/12#issuecomment-1286692451, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACCA5M3N47ELBJT2NO3UFW3WEJN7XANCNFSM55PHLQWQ . You are receiving this because you were mentioned.Message ID: @.***>

amcewen commented 1 year ago

Good question. Just done another very quick test, and it looks like the capacitors do affect things (and not in a good way). Taking it outside (on an overcast day) the one without the wind circuit and caps lit up the charging LED immediately, but I didn't manage to get it to light at all on a board with the wind circuit.

We should try removing the caps sometime, given the charger is expecting to cope with variable power from a solar panel it might also do okay with variable power from wind power. And I'll give the one with wind circuit another go sometime when it's really sunny, just out of interest.

Tactile-electronics commented 1 year ago

Ahh ok thanks - great the solar part works :) If it ends up being separate boards for wind/solar then that’s not a problem.

On Fri, 21 Oct 2022 at 13:32, Adrian McEwen @.***> wrote:

Good question. Just done another very quick test, and it looks like the capacitors do affect things (and not in a good way). Taking it outside (on an overcast day) the one without the wind circuit and caps lit up the charging LED immediately, but I didn't manage to get it to light at all on a board with the wind circuit.

We should try removing the caps sometime, given the charger is expecting to cope with variable power from a solar panel it might also do okay with variable power from wind power. And I'll give the one with wind circuit another go sometime when it's really sunny, just out of interest.

— Reply to this email directly, view it on GitHub https://github.com/Tactile-electronics/windmill-electronics/issues/12#issuecomment-1286900854, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACCA5M6FMKZSPKYH26WEN6LWEKEPFANCNFSM55PHLQWQ . You are receiving this because you were mentioned.Message ID: @.***>

amcewen commented 1 year ago

Need to charge my My Bike's Got LED battery today, so thought I'd test the USB charging.

The USB connector on board C doesn't seem to be soldered on properly—I'm not getting a voltage on the input connections. Have switched to one of the other boards instead.

10:15 - battery at 3.66V
13:15 - battery at 3.9V. The chip gets fairly toasty while it's running, but it's doing a good job of charging the battery (a big 13400mAh Li-Ion one)

Battery fully charged okay, and more quickly than the other charger I've got, which takes an age.

amcewen commented 1 year ago

Tried the solar panel with the wind charging circuit again this morning.

The polarity is correct but I didn't get the charging LED from the solar. Plugging it into USB lights up the charging LED, so the circuit is working okay, and plugging the solar panel into the charging circuit that doesn't have the wind charging board attached also lights the charging LED, so there's enough solar energy.

I tried leaving it for a few minutes, but that didn't seem to make any difference. I think it'd be worth trying it with the wind circuit attached, but no capacitors on thatt—if that works with solar then we could see how well the wind charging circuit runs without the caps as maybe they aren't needed here (given that there's a charging circuit that handles variable input power and a battery to "smooth out" any variations...)

amcewen commented 1 year ago

I've been using board A as a USB charger for my My Bike's Got LED battery for a few months now, and it seems to work just fine. As noted above, the chip gets fairly warm but has been working fine. We've used this circuit on v2 of the My Bike's Got LED boards and have included more copper around the chip to work as a heatsink. If that works better we can replicate it here.

One issue with the charger is that the "done" LED never comes on. The "charging" LED stays lit regardless. I've spun that out as #13 to see if we can work out what's causing that.

Tactile-electronics / windmill-electronics