Am I sailing too close to the wind?
I am looking at a new build, actually a couple, and would like to know from real world experience if the below combinations are too close to V limits to work reliably.
Triple xpg R5 driven at 1A from a bflex.
Has anyone tried this with an 11.1v lipo battery or do I really need to stick with my existing 14.8v Li-ion bricks?
double XPG r5 driven at 1A from a bFlex
Has anyone tried this with a 7.4v lipo battery?
I know in theory allowing 3.3v for each LED and an additional 1v for the bFlex both of the above should work but I would like conformation from someone that has it working, or correction for my "assumptions".
Thanks in advance
emu26, if you get lucky as I did for once ( thanks to Over The Hill) then you could end up with some XPGs with low vf. Mine have a vf @ 1A of only 3.09v each and that's when cold, it fell a touch lower when they warmed up and at first I couldn't beleive it so I checked them with three different meters and still got the same results.
So the short answer is if you get lucky then you can get away with it....just.
I run my twin XPG / bFlex light off 11.1v and have just got in from a ride, tomorrow night I'll use my 7.4v(Li-Ion) and let you know how it goes.
Last edited by yetibetty; 08-18-2010 at 09:28 PM.
Hey Troutie, I'm thinkin' about a triple, so I watched your vid, but I'm a bit thick so I didn't follow the plot. It looks like you start with a LED cold, and you're watching the current ramp up as the LED warms up (is that right)?
Originally Posted by hootsmon
Nope watch the power supply starts at 12.6 volts and as I turn it down you see the point when the bflex goes out of regulation and the light starst to dim .
so you sort of have an emergency bit of light at the end of the discharge .
Ah thanks, sorry for my dopey question.
first, thanks heaps for taking the time to post that, I really appreciate it.
As usual though I think it raises more questions than it answers, mainly because I am not sure if I am interpreting, or picking the change in output properly.
Is it usual for the bFlex set at 1A to only be supplying between .85 and .93A or does this discrepancy have something to do with the additional load of the meter?
Am I right in saying that it starts around 12.2v 0.85A, as it drops to around 11.2/11.1v the efficiency of the driver picks up putting out a max of .93A?
As the voltage drops to around 10.8 the led still seems to be as bright with a current of around .84A and the driver starts to loose regulation between here and around 10v where the led seems to dim noticeably? Sorry a little hard to tell as the ambient light picks up
assuming the above is a fair reading then it would seem that based on that, and what yeti has said then the triple looks feasible.
Loosing 3.3v out of the equation it would appear as though the double might also work, but go out of regulation slightly earlier in the discharge curve. Thoughts anyone?
Thanks again Chris
No you are reading it a bit wrong
The amps are what the light is drawing out of the battery not what the bflex is supplying to the LEDs
The meter is showing the input current to the driver, not the output current to the LEDs. They aren't the same. That's the magic of a buck regulator. You can calculate the output current based on the following formula which is valid while in regulation.
Originally Posted by emu26
Vin * Iin * efficiency = Vout * Iout
Let's take the 11.2v and 0.93A point as the example. If we assume an xpg Vf of 3.1v, then Vout = 9.3v
11.2 * 0.93 * 0.90 = 9.3 * Iout
Iout = 1.008
This is a good example of how the human eye can't perceive small changes in light output and therefore you can't pick out the point where it falls out of regulation by looking at the light. The best way to tell when the regulator drops out of regulation, is to watch the current shown on the meter. When a buck regulator is in regulation, input current will always rise as input voltage falls. That's because by definition, the regulator is trying to put out a constant amount of power. Therefore, it needs to take in a constant amount of power. Since power is P = V * I, I has to go up as V goes down to keep P constant.
Originally Posted by emu26
So, in the video posted, the current rises as the voltage drops until the point the voltage is 11.2/11.1 and the current is 0.93 amps. After that the current and voltage both drop and hence input power to the regulator starts dropping, and the output power is also dropping. That's where it's dropping out of regulation.
It's completely feasible, but it will technically be out of regulation when the battery voltage drops to 11.1 / 11.2 volts. But, as you can see in the video, that doesn't mean you'll be able to perceive the drop in light output. I think the more practical question is will a 3 cell pack give you the runtime you want. I guess you also have to decide if it's going to bug you knowing that it won't be in regulation the whole time.
Originally Posted by emu26
Hi emu, as the battery voltage drops from usage, eventually the regulator will drop out of regulation. Let's say the three XPG's have a voltage reguirement of 9.3 volts for a 1 amp draw. The regulator needs a volt or so for overhead. This means we need 10.3 volts (minimum) to stay in regulation. 3s Li-ions have a end of charge voltage of 9 volts. Your light will drop out of regulation (dim) before the battery is flat. 4s will have an end of charge voltage of 12 volts. The light will stay in regulation until the cutoff point. 18650 Li-ion cells have about 9W/Hr per cell. A three cell battery has 27W/Hr; a four cell battery has 36W/Hrs. Three XPG's will pull around 9 watts, regulator a watt or two more. Your light will draw about 11 watts from battery. So, with three cells, the regulator will drop out of regulation after a couple hours. With four cells, you should get 3 hours plus. James
I should add that you shouldn't take the 11.1 / 11.2 number to be the literal value of where your light will fall out of regulation. I was more trying to explain what is happening in that video. But in general, if you only have one cell per LED, the light will fall out of regulation at some point during the discharge cycle with a typical buck regulator.
The calculations above and the actual values for any particular light will depend on the actual Vf of the LEDs involved and the actual overhead value for the particular regulator. The Cree datasheet states 3.3v at 1A as the typical value for an XPG. In practice, they are usually that or a little lower. Of course in your case it would be helpful to get ones with a lower Vf. Also, don't forget about voltage drops in the wiring. This can be a significant factor when you start dealing with wires that are longer than a few inches. I suspect wiring voltage drops are playing some role in that video.
The bflex manual says the overhead for the bflex is 1.1v for 2 or more LEDs. I've never measured the actual dropout voltage of a bflex to see if that is accurate or if it varies much between individual regulators. Given the type of electronics involved with the bflex, I would expect it to be pretty consistent.
Also, Troutie didn't explicitly say the video was of XPGs. It would be helpful to know if that is the case.
The video camera itself also isn't being terribly helpful in this case. It's adjusting the exposure of the video automatically. So it's making everything brighter as the LED output reduces.
edit: made edit to clarify
Last edited by MtbMacgyver; 08-19-2010 at 12:03 PM.
Thanks for a good explaination MM
They were definatly XPG R5s
it was done this morning as I was having breakie so was a bit rushed and done on a whim as every thing was to hand so only took 2 minutes .
yes the joys of auto everything with cameras as I dont know if it has an exposure lock
Goes away to examine camera now
Ok I've just got in from my ride with my 2x XPG / bFlex light with my 7.4v Li-Ion instead of the usual 11.1v.
First I re-pogrammed the bFlex to shut off at 6v instead of the usual 9v (3v per cell)
The result isn't too bad with dimming that my eyes noticed only happening after around 6.5v and then pretty bad at 6.3v so not too far from the end of the bFlex cut off point.
But the big thing is the hour loss in run time. 3 hrs with 11.1v & 2 hrs with 7.4v.
Both batteries were tested on my balance charger at 2200 mah capacity.
It could be Ok for shorter rides, but for me it just felt wrong knowing that I didn't quite have enough volts and I like the peace of mind of having more volts than I need and the longer run times of the 11.1v battery.
Ha Found the manual exposure widget on the video camera
amazing what info is in the manuals
new vid up here
Ah, the old guard has come through again, thanks so much all for your replies and the effort in testing and videoing.
Yetti, I'm thinking something like this battery will take care of the lack of runtime. I will do the run time calcs though to see what difference there is between that and the smaller capacity 11.1v batteries as both are around similar size and weight.
Chris, thanks for the extra vid, but can you do something with the lighting / exposure, I can't see the readout towards the end of the video What you say about the current draw makes much more sense than my presumptions, I would have been surprised to see George duping us all this time.
MacGyver and OldMTB, thanks for the explainations, you've both confirmed what I already suspected about the theory of it, knowing that the lights would go out of regulation at some point in their discharge cycle.
Lots of food for thought and options to go away and weigh up.
Thanks again for the help