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

That’s what i believed one time, until i began to use USB monitors to test my powerbanks, panels, and various adapters. A 7 watt panel i tested dropped to 3.15 volts with a 1.5 amp load. The ports in my RV can sustain 5 volts up to 1 amp, then it drops as the amps increase past that, and while i can get close to 2 amps from the ports, the voltage sags to 4.3 volts at that amperage load. (I now use a similar model to this: USB 3-in-1 Voltage/Current/Capacity Meter
also the same thing with smaller powerbanks, the voltage sags badly when the load goes up, especially on those with two USB ports. Only the best, largest, and most powerful power bricks & booster i tested can keep the voltage above 4.9 volts with a large amp load.

That would be a good idea if its electronically possible, (and not to expensive) to have a dual charging rate design, (switchable somehow between a 1 amp trickle charging for use with lower power USB sources, then a higher rate 2+ Amp charging for use with a larger 120 volt USB power adapter. ( such as a tiny DIP-switch nex to the USB port on the lantern.

@Toykeeper and DBSAR, all of my USB power supplies can easily do 2A continuous without sagging down 5V, or even higher. You just need to have quality sources.

However, I still think we should stick to 1.5A charging using the TP5000 IC, as to support most USB chargers as you both said, and to support cheaper long cables.

I have had cheap cables from friends actually heating up and losing a lot of power just to thin wires.

cables definitely can cause voltage sag with small gauge wires. I never got the chance to really test out the TP5000 yet on any devices or charging.

The part currently being pursued is the BQ25895 made by Texas Instruments. I have been trading emails with an applications engineer at TI regarding using this part with less than stellar USB power source. Here is a bit of the conversation

ME: Can the Ilim pin configure the chip to limit the input current, and hence the current delivered to the batteries being charged? I think what I really want to do is limit the input current such that I don't load down a USB power source, and allow as much current as is available to charge the batteries. Likley want to limit the input current being supplied to VBUS to 1 or 1.5A. Is this a simple as selecting the correct resistor on the Ilim pin (pin 10).

TI: yes, the ILIM pin limits input current which is directly proportional to output current SYS and BAT. If you have no load SYS, then the relationship is simply efficiency=(VBAT*IBAT)/(VBUS*IBUS) where IBUS is clamped to ILIM resistor setting. The charger also has the VINDPM feature and ICO features which prevent the charger from collapsing its input source.

What is key is the last sentence from the TI engineer. I was reading about the VINDPM and ICO features last night. Basically the IC senses the input voltage and input current being drawn from the source, and limit the current into the batteries based on several variables. If the battery charging is demanding too much current for a given source, and causes the source to droop, the IC will back off on the demand to keep the source from collapsing due to excessive current. Seems like this will mitigate having to artificially limit the charge current into the batteries. The IC is also able to detect what sort of USB source it is, and base the current limit from the source based on that sources characteristics.

From the data sheet, The bq25890 contains a D+/D– based input source detection to set the input current limit automatically. The D+/D- detection includes standard USB BC1.2, non-standard adapter, and adjustable high voltage adapter detections. When input source is plugged-in, the device starts standard USB BC1.2 detections. The USB BC1.2 is capable to identify Standard Downstream Port (SDP), Charging Downstream Port (CDP), and Dedicated Charging Port (DCP). When the Data Contact Detection (DCD) timer of 500ms is expired, the non-standard adapter detection is applied to set the input current limit.

This is a very capable part which DEL selected. BQ25895 datasheet

it would be good if a charging circuit was able detect the current & voltage from a USB source to adjust the amp load/charging rate to sustain the needed voltage without to much sag, especially with solar panels which most have no regulation for the voltage sag when put under amp loads approaching their maximum output efficiency.

Fantastic idea sbslider, the BQ25895 seems to be exactly what we are looking for. The wide voltage range with automatic step up/down is perfect to deal with unstable power surces and would us even allow to add an additional 12V input for even faster charging in the car/ trailer or with larger solar panels.

I am interested x2

Put me down for 1 please sbslider.
Thanks

aslpg get number 890 on the interest list.

LGW gets number 891 on the interest list.

Yes, it looks like a very capable IC. DEL gets the credit for choosing it.

Add me for 1 more, for a total of 2.
Thanks.

Ooops, sorry I missed that, your second is 892.

The quote below is from the D4 thread were I was discussing a Anduril FW mode observed on my emisar D4

So I was telling my wife about the sunset mode on my D4,and her response was, “Is there a mode that does the reverse, so when it’s getting dark the light gets brighter?
Sounds like a great mode for a lantern.

Sunrise mode has been discussed, and there is even an early patch to implement it. It’s more tricky though, because the attiny timer is rather inaccurate. And the first patch doesn’t use sleep mode while waiting, so it uses quite a bit more power than necessary.

Let’s say the goal is to wait 8 hours and then turn on. The user could start two lanterns at the same time, and it could end up with one triggering at 7 hours and the other at 9. Or 7.5 and 8.5. Same code, same configuration, same battery voltage.

That’s the same reason why sunset mode isn’t exactly an hour. Except the timing seems to get less accurate while the MCU is asleep, and sunrise’s longer time scale amplifies the effect.

I had a feeling that the 64 minutes on my D4 may be 57 on yours. I think having a timer for 8 or programmable hours to light up could be a great alarm. I actually have an app on my phone that does this with light and sound, starting dim/faint, and slowly rising to wake me up. I like it alot.

I think what my wife meant was the sun it going down, so turn the lantern on in this new mode, much like sunrise, but the light would slowly get brigher. So when the ambient lighting is decreasing, the lantern light would increase. This would not have the same problem as the 8 or so hour timer, as it would start immediately, and a few minutes one way or the other would not really matter.

Practically speaking, most folks would just turn the light on and want it on. Me personally, I would definitely use it, likley more at home than not. I was just pleased my wife was thinking “light” thoughts. May convert another one yet. :wink:

How about no delay, make it like hitting the snooze button on the alarm, function being “wake me up gently” ramping up slowly from moonlight to full brightness over 5 to 10 minutes. If this were accessible with one button press after running sunset mode it might be a nice option.

:+1:

update:
2-channel MCPCB designed. Here is the basic design of the 2-channel MCPCB for the lantern for the 3000K ~ 5000K ( or 2700K ~ 5000K LED emitters) to work for the tint ramping. After some tests, i found that having the different color temp LED pads in-line with each other linear from the center post/tube, created the smoothest tint mixing beam pattern. (*Red ring is the white center tube that covers the wires and the center bolt) two other black holes are for screws to mount the MCPCB to the lantern head.
(open to conversation.)

Looks good! Have you got this made in PCB design software or just a MSPaint rendering? :smiley: