If it was a usb car charger and your car was off, it wasn’t probably not providing power? That’s been my experience at least, with car usb taps and cigarette lighter usb power.
My Wrangler has one switched 12v “lighter plug” and one “always on” plug. The Wisblock was plugged into the always on side. I am 100% certain it had power all day.
It seems to stay on for around 6-7 hours then shuts itself down regardless of power state.
An update:
I’d originally placed the Wisblock on a mountainside to operate as a router(is_router true) in a sizable mesh I’m building in rural, mountainous terrain. The farthest node-to-node reach is 11 miles and with the leaves coming in on the trees, that connection was getting spotty. The Wisblock was going to cut that distance in half or so.
It dropped off the mesh overnight and I retrieved it assuming the battery was dead(changed from is_router to is_low_power). I placed it in my vehicle at work powered by USB. It dropped off the mesh.
I brought it home and set ‘is_low_power’ to false yesterday evening. It has been on and running on battery since. BT is sleeping and I didn’t attach a screen as my original intent for this board was to be a router.
It seems that power management is bugged maybe?
Not sure, but my remote solar node deployments with the RAK boards have not had the router or low power settings on. I figured it wasn’t necessary to change the power behavior since the usage is reportedly so low.
Do you have the GPS module installed in your nodes? Curious as to how your solar panel cases perform since @geeksville mentioned the GPS rail being powered all the time. Had any issues with keeping power going?
I placed the Wisblock setup in a pretty difficult to get to spot on top a ridge and am hoping the panel can keep the 3000mAH battery alive until at least the next firmware update for Wisblocks. Haha
Mine is just the standard evaluation board with no screen or gps. I suspect that even with those rails being powered all of the time, if there’s no gps peripheral on it, it isn’t affecting the current draw in a significant way.
No screen, but I do have the GPS module installed.
Latest update in my Wisblock saga:
After staying on all night without issue, I put a full charge on the Wisblock and climbed a powerline cut through some gnarly underbrush to a ridgeline and placed the Wisblock there in a tree. Before I hiked out, it was on the mesh and connected to both nodes that it was to be the go-between for. A short time later, it cannot be reached at all even with direct line-of-site and only a few hundred yards away.
Guess I’m crawling through that thicket again in the AM. =(
I think I figured something out… and it explains why my solar deployment works and why yours drains quickly:
The solar power JST input just doesn’t work in the 1.2.30 firmware at all from what I can tell. I haven’t flashed it to the latest but I don’t think anything has changed in terms of power logic for the RAK boards. I supplied 5v to it (and yes I checked the polarity), and we get nothing
Switched to a micro USB breakout (like I used in my last solar setup with power manager board + 18650) and we get power:
Try yanking your battery JST while power is supplied via the solar JST. I bet your board will power off. I’m not sure if it’s solveable by some pin mapping in meshtastic device or if this is something just hardware implemented on the RAK 5005 base board that needs intervention somehow. I may do some reading in the datasheets tomorrow.
Makes total sense.
That could explain why, when placed in direct sunlight without a battery attached, the board seemed to be unable to boot and run on the panel alone. When I first received it, I noted that behavior in my short review but mistakenly thought the small solar panel was simply underpowered. I also noticed that when the battery was attached and charged and the panel went from shade to direct sunlight, it would reboot.
When I mentioned it ran all night without a problem, it was obviously inside out of direct sunlight and when I placed it on the ridge with a fully charged battery, it was early in the morning and overcast. Later in the day, the clouds cleared away and it was sunny.
@Muscles I was experiencing something similar to you when I was on USB. Issues with flashing and booting issues. I updated the bootloader to the latest version and that seemed to fix things: Quick Start Guide | RAKwireless Documentation Center
I’m testing one of the two RAK boards I’ve received today.
After successfully uploading the 1.2.39 firmware, I’ve noticed it wasn’t working anymore after a reboot, so I’ve updated the bootloader (thanks @KevinKW87 for the link!), re-uploaded the firmware and that apparently fixed the issue.
so is solar still down or does it work now?
I saw another user mention that his solar enclosures are working correctly. My experience has not been as good.
My Wisblock won’t stay running long enough to know if the solar can keep the 3000 mAH LiPo I have connected charged. It freezes after several hours both on charger and off. The green LED goes solid and the node is unresponsive. If you set “is_router true,” it shuts off and drops off the mesh entirely after only a few hours.
I have three of them and am excited to use them once development for them advances. The array of available sensors would make them a valuable part of any mesh once GPIO support advances.
I’ve got two of them running well for an extended time now (more than 2 weeks). Even with the small solar panel on the supplied enclosure they’ve never run out of battery (2650 mAh) so far but I am at 63 degrees north so there is plenty of sunlight at the moment. The battery level indicator isn’t working (it seemed to work right at the beginning) so I can’t tell if the system has much to go on.
One runs with GPS and the other without. On one of them I noticed a hang up similar to what you describe when I first tested it, but since updating the boot loader and updating to 1.2.38 or 1.2.39 problem hasn’t reoccurred. I run both in router mode.
I am very much looking forward to GPIO support and the possibility to connect to the internet / MQTT.
I had some issues initially too. Something in the following steps fixed my devices and they have been up for around 3-4 days continuously:
- Update the bootloader.
- Change the channel back to default settings.
- Set Broadcast Position Period to 400.
- Set Device Sleep Period to 280000.
- Hit physical reset button.
I also think I had a bad 18650 cell for whatever reason on one of my nodes. I changed that out with another one and everything has been working well since.
FWIW, I believe with the RAK builds, the traditional sleep state machine informed by is_low_power and some other sleep preferences are largely ignored in favor of keeping the board powered because even in full power the consumption is so low.
I charged the 3000 mAH battery that I am using with the Wisblock solar kit to a full charge separate of the board. I reinstalled the battery and left it sitting indoors away from enough light to activate the solar panel. I did this yesterday AM. So far, the node is running without freeze or shutdown for around 24 hours. This is not the norm for it when it’s on USB power or outdoors overnight.
Wonder if there is a bug in how the hardware transitions from sole battery power to solar/USB?
I am on the new bootloader and 1.2.39.
Did you check the polarity of your solar cable? The silkscreen is wrong on mine in terms of GND and positive
I didn’t. I thought maybe the RAKWireless solar enclosure would correspond with the RAKWireless Wisblock board. How crazy of me. I simply plugged it in. I’ll give it a look after while. Thanks!!
Edit: The solar panel is correctly indexed for polarity, ie red +, black -.
The connector is indexed to put the positive wire on the pin that has the arrow. I remember the battery pins had the arrow pointed at the negative. Is the solar plug the same?
Edit 2:
On the RAKWireless site, the polarity of the solar connector is correct.
I had similar issues with the RAK 4631 freezing up after a day or so, but updating the bootloader and re-uploading the firmware seemed to do the trick.
Does anybody have the battery percentage indicator working in the android app? Mine shows the “plugged in” icon even when running solely on battery, no solar connected. Currently running the latest firmware (1.2.39) There’s probably something I’m missing?
And has anyone range tested these boards? How do they compare to the T-beams? My initial testing suggests shorter range in comparison, but again I could have a faulty configuration.
Thanks to all the devs behind the scenes. Awesome awesome work on this project!