P60 drop-ins - wide voltage as bright as narrow voltage?

[DHart]

I read (but not sure of the veracity) that typical Q5 and R2 P60 drop-in lamp assemblies (such as Solarforce's R2-M lamp 4.2v - 8.4v) use buck converters to drop the voltage from two cells down to that of a single cell, so that you can drive the lamp with either one cell or two.

My question is, will a wide voltage lamp (like Solarforces 4.2v-8.4v or 4.2v-12v) give off the same brightness when driven by a single 3.7-4.2v rechargeable cell as a lamp which is specified for single-cell only (like Solarforce 3.7v-4.2v)?

Just wondering if there is benefit to choosing a single-cell-only lamp vs. a wide-voltage range lamp... if the lamp is going to be driven by a single cell only?

 

[Marduke]

If the lamp is designed for two or more cells, you will get better performance with a single cell using a lamp designed specifically for it.

The higher voltage modules are buck only, and will be in direct drive much of the time on a single cell. The single cell modules will boost the voltage and stay in regulation much longer.

 

[mdocod]

*some* of the narrower voltage range modules on the market have less over-head voltage requirement to run in regulation.

just as an example...

The MalkOff M60 runs at full output right down to ~3.8V, it has almost no over-head voltage requirement to run the driver at full output. When a single li-ion cell is used to run this module, it starts off at or near full regulated output, and then starts to dim at some point through the discharge when it drops out of regulation.

The 11836 (3.7-18V), LumensFactory 3.7-13V, and many other modules like these, don't reach full output until somewhere around or above 5V input. On a single 3.7V li-ion cell, they start off at about 50-70% of their maximum regulated output on a freshly charged cell, and steadily decline to ~25-35% of their maximum regulated output through the discharge. The resulting discharge curve is great for extended runtime where you don't need the absolute brightest possible light.

I have no idea about the 8.4V and 12V rated solarforce modules, maybe someone else with these modules and a multi-meter or light meter could give some useful information.

Buck regulated modules are all going to drop out of regulation during the run on a single li-ion cell, or start off below regulated output on a single li-ion cell.

On the whole, the narrow voltage band boost and boost/buck regulated modules (like 3.0-4.5V) designed for use on a single li-ion cell are going to run brighter than any of the buck regulated modules when compared on the same single li-ion cell. This is often desirable, however, these modules achieve maximum output and sacrifice some potential runtime. buck-only regulated modules will generally have more runtime on a single li-ion cell.

 

[kramer5150]

The 11836 (3.7-18V), LumensFactory 3.7-13V, and many other modules like these, don't reach full output until somewhere around or above 5V input. On a single 3.7V li-ion cell, they start off at about 50-70% of their maximum regulated output on a freshly charged cell, and steadily decline to ~25-35% of their maximum regulated output through the discharge. The resulting discharge curve is great for extended runtime where you don't need the absolute brightest possible light.

My findings exactly with DX11836, 14442, 6090, 17593, M60 and M60LL. All are optimal brightness above ~4.5-5 Volts (approx).

 

[mdocod]

....You can also add the M60 to this list.

Very interesting, I've been under the impression that it runs in regulation down to about 0.1V above the Vf of the LED.. Not true?

 

[kramer5150]

Very interesting, I've been under the impression that it runs in regulation down to about 0.1V above the Vf of the LED.. Not true?

:thinking:

the M60 and M60LL I borrowed from bigchellis were both slightly dimmer running off a 17670 and 16340... could have been placebo effect too, I had no way to "blindly" compare the differing battery setups. Could also have been higher than "normal" resistance in my 17670 cells, holding the modules back.

 

[DHart]

I did a little testing. All cells used were fresh off the charger.

First, two different Solarforce LC-1 R2-M 5-mode lamps, one a 3.7v and the other a 4.2-8.4v:

Solarforce 3.7v lamp = EV of 4.9 at 29" ceiling bounce

Solarforce 4.2v-8.4v lamp = EV of 4.9 at 29" ceiling bounce

So, no difference in output from a single cell.

Then I compared them for runtime, each with a single 18650 unprotected cell. They ran for 3 hours on high output setting until I shut them off when voltage dropped to about 3v. During the 3 hour period, the output of each lamp gradually decreased. (Apparently these Solarforce lamps are not regulated? I initially assumed they were.)

Output measured in EV as a direct beam measurement (non bounce) from the lamp at about 9'. Measuring intervals were not uniform (next time they will be!), but this will give an idea of how they dropped:

=========R2 3.7v==R2 8.4v
12:25:00 PM===8.1=====8.0
12:33:00 PM=== 8.0=====8.0
12:57:00 PM=== 7.9===== 7.9
01:07:00 PM=== 7.8===== 7.8
01:20:00 PM=== 7.7===== 7.8
01:33:00 PM=== 7.6===== 7.8
01:46:00 PM=== 7.5===== 7.7
01:54:00 PM=== 7.4===== 7.6
02:01:00 PM===7.4===== 7.6
02:11:00 PM=== 7.2===== 7.5
02:25:00 PM=== 7.1===== 7.4
02:46:00 PM=== 6.8===== 7.2
03:02:00 PM=== 6.4===== 6.7
03:14:00 PM=== 5.9===== 6
03:21:00 PM=== 5.6===== 5.5
03:29:00 PM=== 5.4===== 5.1

Interestingly, the output dropped a little quicker with the 3.7v version. Which may be more of a function of the two different cells rather than a difference of the two different lamps.

Also interesting, I compared ceiling bounce output on two identical SKU Solarforce lamps (LC-1 R2-M 5-mode 4.2v-8.4v) and got EV 4.9 with one and EV 4.0 with the other.... that's a considerable difference in output for what is supposed to be the exact same lamp!

I also metered two of my Malkoffs:

M60 w/2*16340 EV 5.2 (ceiling bounce)

M60 w/1*18650 EV 5.1 (ceiling bounce)
M30 w/1*18650 EV 5.4 (ceiling bounce)

M60 w/1*16340 EV 5.1 (ceiling bounce)
M30 w/1*16340 EV 5.3 (ceiling bounce)

So the M30 consistently had higher output than the M60 on a single cell. But, as we know from other's tests, the runtime on the M60 would be much longer on a single cell than that of the M30 with a single cell.

Where the M30 really trumps the M60 is when running with a single primary:

M60 w/1*CR123 EV 2.5 (ceiling bounce)
M30 w/1*CR123 EV 4.6 (ceiling bounce)

 

[Kestrel]

M60 w/2*16340 EV 5.2 (ceiling bounce)
M60 w/1*18650 EV 5.1 (ceiling bounce)

So assuming the M60 is ~240 lumens in the standard configuration, it is doing maybe ~235 lumens with the single 18650, but with a fantastic runtime & a graceful (and looooong...) decline in output when depleted. Very very interesting, an awesome configuration.:thumbsup:

 

[DHart]

Kestrel... yes, from the results of my tests, I see NO reason whatsoever to run an M60 with two 16340s if you can use an 18650 (or 17670) instead. I consider both the M60 and the M30 to be single-cell lamps from a practical usage standpoint, with the M60 offering far more utility at the cost of just a slight reduction of output as compared to the M30.

Plus, if you were stuck with primaries only, you could run an M60 in a 6P with two primaries or in a 3P with one primary. The M30 probably shouldn't be used in a 6P with two primaries... possible *poof*. The M30 does give much better output with a primary than an M60 does, however:

M60 w/1*CR123 EV 2.5 (ceiling bounce)
M30 w/1*CR123 EV 4.6 (ceiling bounce)

 

[Kestrel]

Can you monitor the ceiling-bounce outputs over a runtime, say every 15 minutes or so? I'd love to see a runtime 'curve' for:
M60 vs M30 w/ 1x18650
M60 vs M30 w/ 1x17670
M60 vs M30 w/ 1xRCR123
M60 w/ 2xRCR123 (helpful as a benchmark to convert your number to the 'official' M60 lumen output figure & regulated runtimes.)

Most interested in the 'semi-regulated' behavior for the M60 vs the standard behavior of the M30 for a single cell.

If you can get some data points, I'll set up the graphs.

 

[DHart]

Added info:

Where the M30 really trumps the M60 is when running with a single primary:

M60 w/1*CR123 EV 2.5 (ceiling bounce)
M30 w/1*CR123 EV 4.6 (ceiling bounce)

 

[DHart]

Can you monitor the ceiling-bounce outputs over a runtime, say every 15 minutes or so? I'd love to see a runtime 'curve' for:
M60 vs M30 w/ 1x18650
M60 vs M30 w/ 1x17670
M60 vs M30 w/ 1xRCR123
M60 w/ 2xRCR123 (helpful as a benchmark to convert your number to the 'official' M60 lumen output figure & regulated runtimes.)

Most interested in the 'semi-regulated' behavior for the M60 vs the standard behavior of the M30 for a single cell.

If you can get some data points, I'll set up the graphs.

I don't have the time to do this right now, but I will at some point, perhaps next week.

 

[Kestrel]

Where the M30 really trumps the M60 is when running with a single primary:
M60 w/1*CR123 EV 2.5 (ceiling bounce)
M30 w/1*CR123 EV 4.6 (ceiling bounce)

Comparing the M60 & M30 side-by-side like this, do you notice any shift in tint at all when underdriving the M60? Maybe it wouldn't be sufficiently underdriven to notice.:shrug:

I don't have the time to do this right now, but I will at some point, perhaps next week.

Hey, that's cool, I know that would take a lot of time i.e. a large number of small segments, really chopping up your day. I was thinking like sometime next month would be cool...

 

[Justin Case]

I did a little testing. All cells used were fresh off the charger.

First, two different Solarforce LC-1 R2-M 5-mode lamps, one a 3.7v and the other a 4.2-8.4v:

Solarforce 3.7v lamp = EV of 4.9 at 29" ceiling bounce

Solarforce 4.2v-8.4v lamp = EV of 4.9 at 29" ceiling bounce

So, no difference in output from a single cell.

Then I compared them for runtime, each with a single 18650 unprotected cell. They ran for 3 hours on high output setting until I shut them off when voltage dropped to about 3v. During the 3 hour period, the output of each lamp gradually decreased. (Apparently these Solarforce lamps are not regulated? I initially assumed they were.)

Output measured in EV as a direct beam measurement (non bounce) from the lamp at about 9'. Measuring intervals were not uniform (next time they will be!), but this will give an idea of how they dropped:

=========R2 3.7v==R2 8.4v
12:25:00 PM===8.1=====8.0
12:33:00 PM=== 8.0=====8.0
12:57:00 PM=== 7.9===== 7.9
01:07:00 PM=== 7.8===== 7.8
01:20:00 PM=== 7.7===== 7.8
01:33:00 PM=== 7.6===== 7.8
01:46:00 PM=== 7.5===== 7.7
01:54:00 PM=== 7.4===== 7.6
02:01:00 PM===7.4===== 7.6
02:11:00 PM=== 7.2===== 7.5
02:25:00 PM=== 7.1===== 7.4
02:46:00 PM=== 6.8===== 7.2
03:02:00 PM=== 6.4===== 6.7
03:14:00 PM=== 5.9===== 6
03:21:00 PM=== 5.6===== 5.5
03:29:00 PM=== 5.4===== 5.1

Interestingly, the output dropped a little quicker with the 3.7v version. Which may be more of a function of the two different cells rather than a difference of the two different lamps.

Hard to say what's going on. The difference could very well be due to the different drivers.

The slight falloff in EV for the first hour to 1.5 hours could be due to LED junction heating, followed by the batteries starting to run out of gas. Or not. For the R2 8.4V, who knows if you ever were running in true full regulation. With a single Li-ion, you may not have provided enough voltage headroom for the driver to start in regulation. Some DC-DC converter ICs have a semi-regulated mode in those cases, which looks very close to full regulation. Often, you see the difference only when you actually use instruments to measure the output. To the naked eye, it looks like the light is running with constant output.

For the R2 3.7V, the driver is probably a buck-boost, but that's not a given. Some buck-boost drivers are not as efficient as buck-only drivers around the Vf transition, which could explain the faster EV drop-off after about 1.5 hrs.

It also could be that the R2 3.7V uses a buck-only driver, just like the R2 8.4V. At longer run times, the driver falls completely out of regulation and the 18650 cruises along in DD with steadily decreasing LED brightness. The R2 3.7V and R2 8.4V fall out of regulation (full or semi) at slightly different times because of manufacturing variation in LED Vf.

 

[Jay T]

So assuming the M60 is ~240 lumens in the standard configuration, it is doing maybe ~235 lumens with the single 18650, but with a fantastic runtime & a graceful (and looooong...) decline in output when depleted. Very very interesting, an awesome configuration.:thumbsup:

Using my M60 I got the following when comparing a 17670 vs 2xRCR123s during the ceiling bounce.


initial 25 29
10 min 21 28
20 min 20 27
30 min 19 27
45 min 17 died at 42 min
60 min 17

the 17670 was 3.8 V at the end so there was still light left

I'll keep running 2xRCRs

 

[Kestrel]

Using my M60 I got the following when comparing a 17670 vs 2xRCR123s during the ceiling bounce.
initial 25 29
10 min 21 28
20 min 20 27
30 min 19 27
45 min 17 died at 42 min
60 min 17
the 17670 was 3.8 V at the end so there was still light left
I'll keep running 2xRCRs

Cool, thanks for the data.

I know that the 17670 would have been running for at least a couple more hours with still-decent output, as I ran a (protected) 17500 down over a period of four hours, with it still emitting usable light of around 10 lumens or so after all that time.

 

[visigoth]

I've been reading through these results, and they seem paradoxical to me: is there *any point* in sticking with the 3.7v Solarforce drop-in, if you get the same output and runtime (as well as more versatility) with the 4.2-8.4v?

I'm trying to decide between two more recent 1-mode drop-ins: the Solarforce 3.7v LC-XML, and the Solarforce 3-6v XML. Extrapolating from these results, it would seem that I'm better off with the 3-6v: I'll be using 1x18650, probably, but the second drop-in would seem to give me alternatives. Am I missing something? There must be *some* tradeoff, or why would they bother making a 3.7v?

(The 3-6v might not have the more advanced parabolic reflector, which is a separate question -- I'm just thinking in terms of brightness and runtime.)