I’ve seen threads on this before but still am not clear about it. I am ordering convoy S2+ And they offer choice of how many 7135 resistors. So having 4 instead of 8 chips in theory makes it more efficient running at the same brightness level? I don’t use turbo modes hardly since it overheats quickly and would be using it at lower levels mainly.
That’s almost correct, your maximum output will be somewhat lower, but it will be more efficiënt.
More number of 7145 regulators = Higher Brightness/Lumen output with reduced runtime
and vice versa
I just don’t quite understand why it would be less efficient with more regulators and the same brightness say 1amp. Does the current run through all the resistors regardless of what output it’s at creating more loss at lower levels? I guess you always have a trade off, higher max lumens but less efficiency.
It does sound confusing, I agree. What you’re missing happens to be the main thing that makes it less efficient to use more chips for the same output level - that is PWM. LEDs are more efficient at lower power settings and less efficient the more power you pump through them. What PWM does is reduce the amount of time that power is applied to the LED (s) but not the actual voltage or current that is used during that time. So, if you use a 50% mode, what you’re actually getting is a series of equal intervals of 100% and 0% alternating back and forth. So with PWM you’re basically getting the same efficiency at 50% as you would at the 100% mode. You can expect runtime to reflect that. Simply, If you got one hour at 100, then you should get two hours at 50 because the LED is only on for one hour out of that two hour period.
Now, if you reduce the number of 7135 chips on the driver, then you are reducing the actual amount of voltage and current going to the LED (s) and thereby putting them into a more efficient operating level. To see exactly how much more efficient it will be, look at the datasheets, or just hop on over to the CREE PCT and compare the efficiency of any LED at higher and lower power levels. For example, the XP-L HD V6 at 3A current level uses 10.38W and gives you an efficiency of 120.5 lm/W. That efficiency doesn’t go up when you use PWM to make a “lower” level! It is simply turned on-and-off rapidly at that level to make the average light output lower! If you really take half the current away by reducing the regulator chips to 1.5A output current, you’re then using only 4.76W at the LED (half of 10.38 would be 5.19) so your efficiency goes up to 154.8 lm/W!
With the above example, the efficiency of the LED goes up by ~28.5% when you go from using a 3A driver at half power to using a 1.5A driver at full power! And because the voltage needed by the LED goes down, and the lm/W goes up, you’re actually getting more light AND more run-time at the same time just by making that change!
this picture tell you all you need to know. www.budgetlightforum.com/comment/976067#comment-976067
I use 3x but i may go down to 2x. i used 1x for a while too but on biscotti it wasn't good with mode spacing. mode 2 and 3 looked the same.
i use it on 18350 and it doesn't get too hot with 4x though.
Thanks for that detailed explanation that’s really what I wanted to know. I love the knowledgeable people here.
I purposely got the S2+ with 4x as my first 18650 light. I wanted the longer runtimes and for it to not get too hot on high. I have many others now but it’s still a great budget light and my first choice to hand to a wife, kid, friend, etc… (I also like the Sofirn SF36 for that, when throw is better.)
I did a lot of reading on the S2+ (before ordering something else, but that’s another story), and the consensus seems to be that, while it won’t damage itself if left running, the 8x is only good for short durations in the highest mode, unless one has heat-resistant fingers. And, as nicely explaned above, running the 8x in lower modes is less efficient than the drivers with fewer chips.
The 6x is better in that regard, but subjectively speaking, some feel the it still gets too hot for extended or non-vigilant usage.
That makes the 4x the one that best balanced, and most versatile, and even then, some feel that the difference in output vs the 3x isn’t significant enough to discourage going even lower.
However, I still wonder how a 5x might perform, but that would require a self-mod to neutralize one of the chips on a 6x model, and may not make much of a difference. The soldered attachment of the 8x/6x board to pill also makes it slightly harder to remove.
You can just remove / desolder (you can even break them off) one or more chips to get the nr you like.
Something else to consider is that if you don’t need the turbo mode, a lower number of 7135 chips will mean you can still have access to all of the modes, but they will be spread over a more suitable range of levels.
I've been using this regulated MOSFET driver in all of my recent and not so recent builds already. Kick ass unit: nice mode number and spacing, last mode memory, hidden double click strobe and 3V slow pulse ramping on/off low battery voltage warning. Output power levels can be tuned via sense resistor mod (not yet tested LoL, if sense resistor is bridged modes should be lost). Really powerful AON7520 onboard.
HKJ ordered one a few weeks ago, hope for him to review it.
What advantage does it have over the stock driver? Is it more efficient ?
Thanks for the head up, I just ordered this one too.
Does this mean that non-PWM drivers such as the MTN17-DDm (while not using the 7135) will have higher efficiency running at lower levels compared to 7135-based drivers with the same max current?
Sorry to say, the MTN17-DDm is still a PWM driver. It isn’t the 7135’s that make the driver less efficient, it is the PWM itself. Or rather, it is the case of using PWM to lower the “power” level that makes it less efficient. The MTN17-DDm is actually a 2-channel driver. It is capable of PWM independently on each channel. It has a single 7135 chip on one channel and a FET on the other channel. It is actually less efficient than a multi-chip 7135 driver, because the FET will allow ALL the voltage and current available from the cell(s) to go through the circuit to the LED (s). Putting PWM on that channel makes it VERY inefficient! The single 7135 helps that by taking over for the lower modes, where a more aggressive PWM would be needed to get the FET to approximate low power levels to the LED. The 7135 regulates it down to 350mA, and then PWM can be used to get it even lower as needed. If you had a cell capable of giving you 35A through the driver, then the FET would have to be PWM’ed down to 10% ON and 90% OFF to approximate the 350mA level of a single 7135 chip. Using the 7135 then, also gives you a fuller range of “levels” from 350mA down to the lowest level the LED can use. Because this is true, the more intermediate channels you have available, the more 100% levels you can use for maximum efficiency at each of those levels. Some members here have maximized that by making 3-channel and even 4-channel drivers (maybe even 5-channel?) so that PWM is used rarely, or never at all in their driver designs. There is also led4power, who has made all his drivers using a FET operating in linear mode, so that PWM is never needed and every level is as efficient as possible.
Thank you for your correction, David! It seems that I still have a lot to learn about drivers…
So in short, does constant current drivers make a difference when using those lower power levels, compared to PWM drivers?
Yes, as far as LEDs go, constant current will always be more efficient than PWM at any given “comparable” output level. This is because the LEDs operate more efficiently at lower levels, and PWM doesn’t actually produce lower levels. The 7135 chips regulate to 350mA each, and can be added together (or not) to produce different constant current levels as needed. That’s what some of the multi-channel designs make use of. Almost all 7135-based drivers resort to some PWM for modes below 350mA though.