Booster IC cirquit for steady blade voltage

[sp00fman]

I found a nice way to get a constant brightness out of our main blade LEDs and get a better efficiency.
Runing a 1S lipo on my blades, when it is half way discharged some colors get really dim, while on a fully charged lipo some LEDs are seeing a sightly too high voltage, thus draining a lot more current than needed.


Reading a bit about running an Rx on 1S for unpowered-gliders someone noted he got a cheap booster circuit off ebay that works really well.
It supplies 5V up to 600ma from an input of 0,8 to 5V with 96% efficiency. The circuit should be adjustable from 2,5V to 5,5V output if you change the onboard voltage control resistors.

Now we can really play with the LED colors to get best performance. Say we want to run red and white LEDs. Red needs 2,2V, white 3,2V. if we series these together and run them on 5,4V we would not need any drop down resistor at all. Its like comparing a Linear regulator BEC to a switching BEC for efficiency.

For a blade running only blue/white/green that needs around 3,2V , we could just set output to that and still keep efficiency high.
With the usual resistor we would just burn up to 1V away on a 4,2V lipo. (75%eff) up to discharged lipo 0.5V 86% efficiency. The current that goes through the circuit of a discharged lipo is only half ,compared to when the lipo is at full charge. That is why it gets so dim.

The red color always seems a bit dim even at full power. If you want a full red blade you could set it to 4,4V and series the red LEDs together by soldering them directly together. That way you get double the brightness without much extra soldering work.

The cirquit:

About $1,50 a piece.
http://www.ebay.com/itm/400578426039...84.m1439.l2649


Info:

G1217 is an output, a constant frequency, synchronous boost DC / DC converters. This medium-power applications that target devices can be 0.6V to 5.0V input voltage range, and has a constant 1.4MHz switching frequency, thus allowing the use of tiny low cost capacitors and the height is only 2mm or Lower inductor. Output voltage range 2.5V to 5V regulated, internal synchronous 0.35Ω, 1A without the use of the power switch in the case of an external Schottky diode provides high efficiency.

G1217 new synchronous boost converter, it can be in a compact package for the designer to provide high output current. This number is G1217 device designed to support a variety of handheld portable applications, it makes a single alkaline battery provides up to 100mA of output, the output of dual cell alkaline battery up to 400mA, and a single lithium ion battery input can be realized up to 500mA output, while supporting the start as low as 0.85V voltage. Using a diode, the output voltage up to 5.5V

As you can see this board already has the schottky diode for efficiency. The below cirquit diagram does not show it but the pinout is correct.

We have to de-solder the USB connector. does not look that hard to do.
The PCB is strong, all components are very flat so it is a low profile. Perfect to heatshrink to the blade root on the opposing side of where the LiPo sits.

[kavic5150]

Great find. I'll have to check that out.

An easy way around the too dim red led problem is to get a few high powered ones from somewhere like digikey or mouser. I got red leds with 250ma or so of current draw where the normal ones draw only 20ma. I only use them on 550mm blades and bigger since they need 240mah lipos to get any decent run time. They are too bright to look at up close.

[sp00fman]

Alright, few bench tests of the booster circuit. It is complete GARBAGE.

Voltage is all over the place when loaded up 50mA. Adding input and output capacitor did fix that a bit, but voltage keeps jumping some if load changes. Thereby i cant get it anywhere near a decent efficiency.

I'm going to look into some other units to see if i can find one that works as advertised.

[setishock]

The three at the top lead to other sections with more options.

http://www.pololu.com/category/84/re...power-supplies

[sp00fman]

Cool, thanks! I already have three other cheap Chinese step-up regs on my way to test if there is one that is usable for us
I will also try this higher quality one.

Will do some testing when they are in and try to estimate their reliability and efficiency.

I ordered the adjustable Pololu U1V11A. Specs say it has a precision of 4%, meaning 0,2V voltage drift. That is a bit much but the input voltage we use is not shifting all that much (3,7-4,2V) and our load is constant, so i hope once connected to the LEDs we can set voltage and forget about it.

[sp00fman]

I got 4 different boosters and have done some bench tests.

From left to right (e-bay names):

600MA 0.9V-5V to 5V USB Step Up Module DC-DC Booster Module
DC-DC Converter Step Up Boost Module 2-5V to 5V 1200mA 1.2A for iphone
Mini DC Voltage Converter Step-up Boost Power Supply Module DIY 5V USB Charger
Pololu U1V11A


The first one is the garbage one that does not work properly

Second one works as advertised, cheapest option. 1S lipo with 1A load kept the voltage steady. Will need modifications to a resistor to change the output-voltage. Red-power active-LED onboard.

Third is comparable to the above, but has a dual-color led to indicate input voltage. When input voltage drops below 3,5V it turns from green to red.
This is a great feature for our blade application. cost around $3.

Last: Polulo. Kind of expensive, but i'm impressed. Has a potentiometer to set voltage on-board, so no need to modify anything. Does not have a LED but does have a power-off feature. Smallest fromfactor and only 1g weight.

It turns out i can use the power-off (called SHDN) pin to make a lipo protection. I put a 33K ohm resistor from ground to the SHDN pin. It will turn the circuit off when input drops below 3,2V, and the circuit will only start up again if the voltage is above 3,8V.

This is great in several ways:
-use the maximum available capacity from our lipo
-will prevent us starting to fly with a discharged lipo
-prevents lipo damage, the threshold is well above the minimum voltage of 2,8V where the cell takes damage.
When the cirquit is disabled it only draws 40uA. Say we guesstimate we still have 5mAh left at 3,2V,we still have 125 hours left to charge it up or disconnect the circuit from power

Efficiency on all the units is at least 80% (measured), and the Pololu claims up to 90% at our input/output voltage range. Can't confirm to the % as i lack a precise current meter that works around 1Amp.

Will post some more when i get along with the build. Am going with the Pololu as i bought them in a pair.