*BLF LT1 Lantern Project) (updated Nov,17,2020)

Good point.
But what about the possibility of an advanced version after the BLF Lantern is finished ?

There is plans for a deluxe version at a later date.

I am in for one. please add me to the list.

Definitely works for me. I like having the charging function, although I probably won’t ever use it. But, the power bank function is completely useless to me. I’d rather have light than to power anything else. So whatever power is in the lantern needs to stay there in my case. I know others are more hard core in their camping, hiking, caving, etc. and would find having a power bank built-in to be quite useful as it’s one less thing to carry.

For a future model we will look at the power bank function. for now the built in charging will be a great feature to have on long camping trips or off grid use.

Agreed, the charging function is much more useful to me.

“Via in pads” is generally a no-no. Google it. Unless you really know what you are doing, how to tent the solder mask etc.

They potentially suck out the screen-printed solder resulting in a weak joint.

Then there is ROHS lead-free solder which doesn’t like this either.

Then the idea is that the underside provides good gold plated vias for the pogo pins to contact. Which they may not if the via is plugged with solder.

Lots of things to go wrong here, for no good reason. Separate the vias from the pads is my suggestion, I see plenty of area to do so.

Edit: agree on powerbank delete decision. I certainly don’t need or want it.

PS: I don’t know what charging chip you have chosen, but please be sure that it can cope with a small solar panel and dump every bit of available energy into the cells, no matter how weak the light, and how variable, night, day, clouds, shadows etc. Not all can, some only work in certain conditions and trip out otherwise.

This is nothing like charging up e.g a mobile phone from a mains adapter. Totally different situation.

Edit: I mean accepting current from any solar panel, from a standard nominal 12V panel (actually a lot higher), to one of the the popular portable USB output (5V ? delivered wastefully) devices.

TBH I think that making it only a USB input is a waste. I’d rather see a full-featured input via a rugged co-ax socket, that can accept almost any supply, from a 12V solar panel, to a re-purposed laptop 19V adapter, to a 12V or 24V automotive connection, to a USB passive or active adapter of any type.

That was my original plan & the reason why i want to limit the charge rate to no more than 1 amp-draw from the USB line.
I know some have argued & cried to want it to have a 2+ amp “fast charge” but that’s not needed or practical for this lantern to be a true off-grid capable light. I am still wanting it to have the maximum 1-amp charge rate for many reasons. One is that a 10-watt small solar panel can be able to charge the lantern without suffering to much voltage “sag” with the heaver amp draw of a 2+ amp charger. on the V1 Prototype i used a single TP4056 charger module, set at 1-amp. On the 10 watt small panel it still charges the lantern with four 18650, 2500mah cells from 3.4 volts to full charge in 1 sunny day of light. (My test it took 7 hours to fully charge) then using the lantern for 5 ~ 6 hours average per night for 3 nights on medium & high intermittently with no problem or worry of it going dead. My tests also showed that just 4 hours average per day of sunlight on the same charger charged it back up to full after 4 to 5 hours of use per night, meaning the lantern can run every night for years as an off grid light source using small solar panels.
I did a test with two TP4056 chargers charging four cells ( in a 2-bank box) and the 10-watt charger with the 2-amp load, would loose so much voltage from load-sag the charger would continuously turn on & off, and the batteries did not charge up at all after a full day of sun. ( same goes with charging from some of the smaller 12-volt adapter USB power plugs, they can’t sustain much past 1.2 amps before they overheat, cause voltage sag, or burn out. ( i burned out a couple of them during load testing.
To try to design it to be a “multi-voltage” charging is just adding unnecessary costs again, (another reason why were dropping the power-bank option for this first lantern model.) The team has set the design now to get this rolling, as its been to long already discussing different options and delayed this project way to long.

@DBSAR, I have a TP5100.

It’s quite a bit more efficient than the TP4056 since it’s a buck converter, and doesn’t heat up much, if at all, at a 1A charge rate.

Since it’s also more efficient, that means you don’t need as much current from the charger to achieve the same charging rate, meaning you can slightly lower current, which will lower the voltage drop of the solar panel and the cable, making it easier on the panel and more efficient :slight_smile:

TLDR: Let’s go with the TP5100 if we can.

interesting, i have not tested that module before.
(though it will be Lexel who is designing the driver to determine what he can incorporate on his driver prototype for that built-in charging circuit.

Outside of this project, I’m working on adding support for attiny1634 and/or attiny841, which both have more pins than the attiny85. This should allow Lexel’s dual-linear-FET driver to work, and also increases the ability to control a powerbank circuit.

But there’s no point holding back the lantern project until that’s done. It makes more sense to do the lantern now, and then maybe make a lantern V2 later with more features.

About tiny1634 vs tiny841, the latter is generally a better chip… newer, at least, and more flexible in how it’s used. The main drawback is it has only 8 KiB of ROM, while the tiny1634 has 16 KiB. And this matters because, um, Anduril already uses almost 8 KiB. I filled up the ROM with things like candle mode. This isn’t an issue for the lantern because it doesn’t need thermal regulation, so that’s a big chunk of code I can leave out. But on other lights, the extra ROM should be pretty helpful.

That would definitely be a good set up for a future deluxe-version of the lantern later on after we get the lower-cost standard one into production for now.

Lexel has not chimed in for a bit on this thread, but I am guessing no power bank means no charge capability either. There is one chip that would perform the charging and power bank function, that is the 20 pin part U4. Unless the ATTiny 85 as some functions we don’t know about. There are no other ICs on the layout, except the AMC7135s. It’s hard for me to imagine he would half populate this section of the board, as it just doesn’t make sense. But I’ve already been wrong before today . . :blush:

Speaking of U4, the lands for the leads do not look symmetrical around the thermal pad. Something funny is going on here, either with the screen shot posted or with the actual design.

You must have missed ToyKeeper’s post above. The power bank function needs a better MCU with more pins. It will be possible in the future, but right now we have no firmware for those chips needed, so we stick with charging only, and power bank in the next Deluxe model. :wink:

Ish.

At minimum, the power bank needs something to turn the circuit on and off. I was hoping someone would come up with a clever way to sense whether a USB plug was inserted, and enable the chip only when something was plugged in, so it could run completely on its own. But that hasn’t happened, so the main MCU would need to enable/disable it instead.

One thought was to enable the powerbank whenever the main LEDs were on, so to charge a phone you’d put the lantern in moon mode or battcheck or something, and then plug in the phone. But this would also reduce efficiency, especially on the lowest modes, and it could be a little weird that simply plugging in a device isn’t enough to actually make the powerbank function work.

For full powerbank control, the MCU needs to actually communicate with the powerbank chip over I2C or something, and that’s definitely going to require more pins and more ROM. But this isn’t strictly necessary, if we just let the powerbank use all its default settings. For that, it only needs an on/off switch of some sort.

There’s also the option of adding a second MCU, which would do nothing except watch the USB port and control the power bank. It’s probably a good option, since it would make all that completely independent. Simple and robust. But that’s not happening in this version.

I did suggest a chip that could do the charge and powerbank without MCU connection, and can cope with flakey intermittent solar input, and has a pedigree, and is available, and inexpensive.

But I can’t remember. It’s here in the thread somewhere. Just look up my posts, It’ll be there.

If I can be bovvered maybe I’ll find it.

Yeah, it was recent. Post #3088

I admitted I might be wrong, but the design I see above looks like it is power bank and charging capable with a single IC, which is U4. I do recall that the current firmware hosted on the ATTiny85 has limitations in controlling all that could be done in this driver. I don’t see a driver that is providing just a charging function. I am pretty certain that the design is the same as the one DEL generated a year or more ago, and was provided to Lexel.

But to be certain we need Lexel to clarify this IMO.

I know we discussed dropping the power bank for this model in the team chat, but i was sure we all agreed to keep the charging feature, which is a key feature needed in this type of light as a very usable feature for long-term use & off-grid use. (will hope he chimes in on that soon.)