High efficiency multi mode analogue LED driver

royfellows

Well-known member
I do seem to have hit a nerve with you Roy, certainly NOT my intention. Regards low battery voltage, why do you think nearly ALL reasonable kit shuts the circuits down at about 3v? No need to drain to 2.7v, and doing so can further complicate charging if the charging circuit does not attempt to kickstart the protection board into waking up again. Surprisingly common! I have also seen some boards not kick in until below the 2.7v level, which again, is not ideal!
Also, there really is no capacity between 2.7 and 3v. I suppose you could argue you do not want to stop at any voltage with caving lamps, but then again I suppose that is what backups are for!
I forgot to add that the low voltage cutoff could be micro-controller considerations but most these days have quite a wide operating range with low voltage modes down to 1.8v and below.
Not sure what you are referring to with regard to 5A. Did I state something to advocate such a current?
Its OK mate

Yes, the devil is in the data sheets.
Some Li Ion cells have higher recommended cutt off, generally earlier ones. Protection boards bought off the shelf may have a cut off set to a "catch all", to match to a specific array of cells may require some tuning so as to get the max burn time out of the battery by setting cut off to the recommended. Or more importantly, these last dregs before manufacturer recommended cut off often supports a diminishing output, we call the "slow death". Note "manufacturer recommended", I seem to be repeating myself here.
As far as accuracy of cut off is concerned, its all about testing components before using them.
One can wonder into the realm of "opinion" on this, with me the manufacturers data sheet is good enough.

I note the "no capacity between" commment above, I have graphs that I have prepared reflecting test results that indicate otherwise, but note my comment about slow death. Obviously a decreasing output will extend the burn time.
 

royfellows

Well-known member
and therein lies a tail as the water rat said to the , Oh never mind.


When I was young I had a Jag XK150 to which I fitted a new clutch. When I came to refill the gearbox I looked at the Jaguar manual and noted the recommended SAE 30 oil. I considered that this was a bit too thin and filled with SAE 90 gear oil.
The gearbox was noticeably quieter, I had got that bit right, but the overdrive didn't work. Mmm.
 

royfellows

Well-known member
Just a heads up that the project is back on the rails.
I have assembled some LED modules and just tried one with a redesigned controller. Basically, it works.
I do have more confidence in this one as I am using the driver IC that I originally wanted but stuck on back order until now. These are being used in the way they are intended whch I think gives one a head start.

I am now struggling with getting the values of shunt resistors working through Schottky diodes, not to mention that RSense value of the driver module affects the lower outputs as well. I am therfore reduced to the most basic form of exerimentation, that of "trial and error", plus the complex mental algorithm of guessometry.
 

royfellows

Well-known member
Some reasonably encouraging bench test results using the new driver chip. There are 4 brightness modes, this shows the lower two which are the most important. As commented before, the lower mode efficiency of most off the shelf buck drivers are poor.
 

Attachments

  • Efficiency V Voltage.jpg
    Efficiency V Voltage.jpg
    45.9 KB · Views: 119

royfellows

Well-known member
My apologies for reopening an old thread but is really is a continuation of the original project. As can be seen from my last two postings I did get a bench model up and running with encouraging results, however managed to trash the chips. The re-order arrived about June last year but I was into a big project at home here that had priority, and then got sidetracked into yet something else.

Anyway, I picked up on the project early in this year, completed a working and reliable lamp and have been testing and kicking it about until I finally assembled a few to offer for sale. Main incentive was non availability of parts for the old design, but what has been achieved is much improved lamp.

I think that I had better explain at this point what it is all about.

There is a basic issue in producing a lamp with a high maximum output but with efficient lower output levels. Additional to this is the problem of producing a lamp that retains its output on any brightness level, but which also will not suddenly die on you when the battery goes down. In effect you are trying to have your cake and eat it. Obviously, higher efficiency means longer burn times; this combined with sustainable output means the lamp is suitable for long weekend activities away from recharging facilities. Go back underground the next day and every mode bar Turbo as bright as the day before.

All of the above has been achieved. Four useful output levels plus an automatic power down to a slow death that lasts for hours. There are Standard (300 lumens), Standard Plus (450 lumens), High 1250 lumens, and Turbo (3400 lumens). On low battery the lamp will start to flash down to a lower level, this will increase in frequency until it stays on the lower (slow death) level. This is about 20% of Standard to start, slowly dying down to a glimmer over an extended period. On practical test, Standard burned for 19 hours and then powered down to the SD mode were I lost interest after 5 hours and turned it off.

There is also an issue with some lamps whereby they flicker at low output levels. The X12_Ultra is flicker free at all output levels.

I will be able to easily produce a custom lamp for anyone who wants lower outputs for say tighter calcited cave systems where the 300 lumens base setting would be too much light. This would be easily achieved by simply increasing the vale of the RSense current sense resistor. If effect, this will reduce all the outputs and can be made to order. An example is RSense = 0.05 ohms will give about 180 lumens Standard and Turbo of about 2600 lumens. Normal fitting is 0.04 ohms.

I have made a fundamental change to my original design. The low battery power down was originally designed to shut down at low input voltage to the lowest main driver output level; however I have had to increase lowest brightness level as the lowest reliable output relates to the highest output as set by set by the current sense resistor. I also redesigned so as to achieve a better ‘Kelvin’ connection relative to the RSense.

The change I made is that on the new design the lamp will now give warning flashes which increase in frequency until it closes down on to the aforementioned “Slow Death Mode” whereby the lamp will continue to burn from a level of about 20%Standard down to a glimmer. This is over a timeframe of several hours; I tested it and lost interest after 5 hours. If on a higher mode the battery may recover on power down somewhat, if this happens it will return to Standard Mode until battery voltage falls again below the 9.5V (adjustable) threshold.

I consider the driver and its analogue controller as a unit which I have called the F_Tech driver for reference purposes. The new lamp and any derivatives I call the “Ultra” range.

The lamp itself, the Ultra12, is similar in appearance to my old X8, the two vertical reflectors with a bare emitter each side, this suggests an 8V input, but its not, it’s a 12v model. The two bare emitters are parallel wired and then serial wired to the 2 serial emitters behind the reflectors. There are also indicator LEDs mounted each side of the reflector, described later. The placement of the bare emitters gives a better external appearance as well as sharing the power and the subsequent heat generated between both sides of the lamp. This one will never burn out a flood, the weak point on many lamps.

The indicator LEDs will show a green indicator on battery connection which will on power down slowly go out to be replaced by a flashing red.

The lamp is actuated by a double throw momentary toggle switch mounted on the same side of the lamp as the cable gland. As a flick switch its extremely user friendly. Flicking down is ON/OFF, up is mode cycle through the 4 modes. Flicking up with lamp off enables the 4 blue LED battery meter, it works with it on but advised to use it lamp off.

The new lamp has virtually nothing sourced from the Far East; all electronic and major components are from major western manufacturers. All the electronics are my own design, bar the main driver circuit which is based on the manufacturer’s data sheet ‘typical application’ schematic.

I have a couple to sell, but not holding my breath. Lamp sales over the last few years has fallen to a mere handful each year, but I get orders for new batteries and of course lamp overalls for ones seen a lot of use

So there you have it all, the project was not a failure but a success, and a good one.
X12U_efficiency_V input_voltage_comparison.jpg
X12U_output_currents-_V_input voltage_comparison.jpg

.
 
Top