Review: Ultrafire WF-501B with 5-mode MC-E drop-in

Ultrafire WF-501B with 5-mode MC-E drop-in

Reviewer's Overall Rating: ★☆☆

Summary:

Battery: 1x18650
Switch: Reverse clicky
Modes: 5 (H-M-L-S-SOS)
LED Type: Cree MC-E M bin
Lens: Glass
Tailstands: No
Price Payed: $20.00
From: DealExtreme

Pros:

  • Bright
  • Cheap
  • Mode memory works well
  • Nice Low

Cons:

  • Under-driven
  • Dark donut hole in beam
  • Current not well-regulated
  • Whines on medium
  • Host can't handle heat

Features / Value: ★☆☆

***Update*** I realized after doing these tests that my multimeter is no good at measuring the currents this light uses, so I have removed those measurements and can only say I don't really know what current the light draws until I get some new leads for my multimeter.

DealExtreme introduced these MC-E “Bin M” lights at an irresistible price of $20. A drop-in with a P7 is over $17, and the M bin promises to be just as bright as the C bin P7. Plus I have XR-E, XP-E, XP-G, and P7 LED lights already, so having an MC-E would add another type of LED to the collection. Lastly, I had the Ultrafire 504B and 502B, so I figured it wouldn’t be bad to have a 501B to go along with those.

Don’s review said he was very impressed with the light and that it was drawing 2.5A at the tail and didn’t have much a donut hole. Knowing that DX quality varies greatly, I went ahead and bought one. It took a little over a week for the item to get in stock which worries me that they had run out of the initial run of lights and were hard at work scrounging cheaper parts to make more.

I wasn’t crazy about having 5 modes, but with such a good current draw, I figured I could live with the SOS and strobe. The brightness levels seemed spaced out pretty well too. The driver also has well-functioning memory, where the last mode is memorized if the light stays off for more than about 2 seconds. Anything less than that and it skips to the next mode. One second would be better, but I can definitely live with 2-second memory. And it beats the memory that is based on how light the light is on.

So on paper this light is a great value. However, the light I got doesn’t have anywhere near the 2.5A current draw (1.4-1.6A) and therefore is much less bright. I have no idea if it is really a bin M brightness. Also the beam has an ugly dark donut hole in the middle that is pretty obvious (on white walls at least) at any distance over 1 meter. Those two things really knock the value down for me.

Build Quality: ★★☆

I was wary of the Ultrafire WF-501B host. I knew it was lighter weight with less heat sinking than the 502B or 504B. However, I was impressed with the overall quality of the host. Yes, it is lightweight, but the machining seems pretty good. There is a small ding in the knurling to the right of the name label, but otherwise it is pretty much perfect. Like the 502B, the light is in 4 pieces: head, throat, body, and tail. There are o-rings at each location and they were lubed. The threads were pretty rough and it can be a little difficult getting started when screwing two pieces together. There aren’t many threads in each location, particularly the tail which only has 3 threads. The throat is a little big for the drop-in so, as with my other hosts, I wound up putting in a few layers of aluminum cut from a soda can. In this case, 4 layers seems to do it which is the same as the 504B and less than the 502B. This seems to help the heat sinking, but honestly it would be crazy to drive this thing at 2.5A or more for very long. My P7 light weighs a lot more and has amazing heat sinking and that whole light gets very hot in about 10 minutes.



The LED itself seems soldered in okay. It is pretty well centered on the pill and has thermal paste underneath. The pill itself gets hot quickly, so I think the LED is mounted properly. The soldering on the back of the board seems fine too and makes a good smooth connection with the pill.

The tail switch is very firm, but works fine and is easy to do half-presses to change modes. The only thing I don’t like there is they use a smooth black rubber switch cover and I would rather have one with texture on it like the 504B has. The 502B was the same way so I already have some switch covers to use as replacements. Although the light seems like it is very similar to the 502B, there are a lot of cosmetic differences which you can see in the picture.

Battery Life: ★★☆

Even though this light draws less current than it should, I don’t know if it is a good idea to try a duration test on it. With a draw of 1.4A, it should last at least an hour, but if it really drew 2.5A it would be more like 30 minutes and I don’t think you could get rid of the heat fast enough to avoid damage to the LED.

Instead I charged some batteries down to different voltage levels and took current draws at each. I am using gray 2400mAh protected 18650’s: two are Ultrafire and two are Trustfire. These are the same batteries I have used in other tests. I am not testing output current or voltage drop across the LED, so I can’t measure driver efficiency (maybe later when I replace the driver). These are just tailcap draws. Out of curiosity I also did a measurement with just a battery and the drop-in and got maybe 100mAh higher draw, so I don’t think there are any bad contacts in the host body itself.

Voltage High Medium Low
4.16V ????mA 510mA 60mA
4.06V ????mA 480mA 60mA
3.78V ????mA 340mA 40mA
3.63V ????mA 300mA 40mA

There doesn't seem to be any real regulation going on as the current drops with the battery voltage. A light like this that requires a high current to be at its best depends heavily on its driver and the driver that came with my light just isn't up to the task.

After writing the review, I replaced the driver with a 2.8A 3-mode driver from Shiningbeam, so I took the original driver out. I measured 1.25A going to the LED with a 1.4A draw at the tail, so the driver is pretty efficient. The driver itself had silicone thermal adhesive all over it. I'm not sure if this is for heat transfer (doubtful), ruggedness (keeps the pieces from coming loose), or an attempt to stop PWM vibration and whine. I scraped off as much as I could and the driver seems pretty simple:

It seems to be basically the same driver as the one in my Ultrafire WF-504B with XR-E R2, but that one was rated at 1000mA:

Light Output: ★★☆

Compared with my 502B which has a XP-G R5 driven at 1.05A, the MC-E puts out more light. But it should be putting out almost twice as much light and that definitely isn’t happening. Also the intensity is less and of course there is the donut hole. I knew that most multi-emitter LED’s are going to have a donut hole or a plus in the beam, so this was expected, but it is more prominent than I thought. Both of these lights have a cool white tint.

Here is a series of pictures comparing the 502B XP-G R5 on the left with the 501B MC-E on the right. The pictures are taken at an ISO of 100 f/2.8 and I step down the shutter speed as noted. The lights are 0.5m from the wall and both have a fresh battery. Here are both lights on High with the exposure at 1/25th second:

Now at 1/200th second:

and at 1/1600th second (now the donut shows up, but really it is always visible):

I will save a comparison of Medium and Low for outdoor shots later on. Now here is my Ultrafire MCU WF-1200L with a P7 LED driven by 2 18650 batteries vs. the 501B. These should be about the same brightness overall, but the bigger reflector of the 1200L will give it a tighter hotspot. The 1200L has a plus-shaped hole in the beam, but it is nowhere near as prominent as the one in the 501B. Here are both lights on High at 1/25th second:

And at 1/200th second. The P7 has a greenish tint:

And at 1/1600th second.

You can see the donut hole at 0.5m, but it becomes more prominent as you get further from the wall. Here it is at 1.5 meter from the wall and a 1/400th second exposure. Later I swapped out a reflector I had in a XR-E drop-in and it made the hole a little smaller. The first picture is the original hotspot and at right is the hotspot with the replacement reflector.

The light uses PWM on Medium and Low, but it isn’t noticeable and a time exposure photo yields a smooth line when moving the light quickly. The Low is pretty decent. The light tends to have an audible whine on Medium, but not Low.

One tree in my backyard is really overgrown right now, so I had to shift my usual range. The fence post is now 80 feet away (in other reviews it is 120 feet away). These pictures are taken with a 4 second exposure to get something like what it looks like in person. For comparison, here is the 502B first with a XP-G R5:

Now here is the 501B with the MC-E on High. The fence post is actually in the donut hole so it doesn't look quite as bright, but otherwise the 501B puts out more light than the 502B. It is hard to see the donut hole in this picture, but when you are moving the light around it is almost like a pointer or like a bug is on the lens:

Now here is my Ultrafire WF-504B with a Cree XP-G R4 neutral white LED. The camera compensates for the tint, so you can't much of a difference in tint. This LED is driven at 1400mA whereas the 502B is driven at 1050mA.

Now here is the Ultrafire WF-1200L with a P7. It has great throw:

Now here is the fence at the middle of the picture zoomed in (502B, 501B, you can't open these for larger versions):

Now the 504B which doesn't throw that well, and the 1200L which lights up trees well past the neighbor's tomato plants:

Here are the same 4 lights on Medium. First the 502B:

Now the 501B (again, the fence post is in the donut hole):

Now the 504B:

And the P7:

Now three lights on Low (the 1200L doesn't really have a Low). First the 502B:

Now the 501B. The Low is significantly brighter than the 502B, but is still pretty good:

Now the 504B

Summary: ★☆☆

This light would be a great value if it could correct the big donut hole and drive the LED per the specifications. It shouldn't be a problem getting a better driver, but it will cost about $9, a big chunk of the price of the light. I may be able to tweak the donut hole by unscrewing the reflector a little, but I don't think the 501B host has room for a longer drop-in.

nice reivew,thanks !

Awesome review and great pics......

I got the 3-mode driver from Shiningbeam yesterday. Before taking out the old driver, I measured the current to the LED and it was about 1.25A vs. 1.45A at the tail. I didn't do more detailed measurements for voltage and power. The light is actually very bright and I was able to get the hole to go away some by swapping the reflector (I added a picture to the review above).

Anyway, the Shiningbeam driver doesn't draw that much more current than the original one. I measure about 1.8A instead of 2.8A. I would have thought the batteries would be able to crank out more amperage than that, but they are gray Ultrafires and Trustfires, plus some flame colored Trustfires which I don't think are quite as good at high drain as the gray ones even though they last longer. Maybe I should get some LiMn batteries (IMR)? Any budget sources for those? I know I can get them for $11 each from AW, plus shipping. Two batteries won't work because both drivers have a max of 4.2V. I could try hooking up 3 AA NiMH's temporarily and see what happens . . . I know eneloops are good for a couple of amps.

The other possibilities for the lower current that I can think of is my meter isn't measuring correctly, though I am using the unfused 10A jack so it seems like it should be okay. Other than that, I guess it could be the LED or circuitry. Maybe I should use heavier duty lead wires? You look at how small some of the components are on the driver board and it seems unlikely that a small lead wire would be smaller than those parts, but I know some people swear by larger gage leads. I'm using the leads that Shininbeam supplied already soldered to the board. It can't be the spring because the spring is already part of the board since it has chips on the bottom (I'll include pictures and do a more thorough review of the SB driver later).

Like I say, I'm a lot happier with the light than I was originally. Now that I've taken it out on some walks with the dogs, it really puts out a ton of light. Maybe a little too much: most of the sidewalks are tree-lined so I get a lot of reflection off of the trees close up, but that prevents me from seeing all that far down the sidewalk where it isn't going to light up quite as much. It is impressive though. I'd just like to think I'm getting all I can out of it.

Here's the 3-mode Shiningbeam driver. I believe the name might be NANJG 106. It has stars on the bottom which makes me think there are different sets of modes available like with some other NANJG drivers. The top (inside the pill):

And the bottom, with the built-in spring:

Could it be something else limiting the current? The switch maybe? Or may be your MC-E is a bit shit with a very high Vf? Perhaps try 2xCR123 if you dare?

It can't be the switch because I'm measuring the current at the tail without the switch being in place. It could be the LED, but I don't think more voltage would help the SB driver because it is a linear regulator and is going to provide some set amount of voltage regardless of the input voltage (I think).

Let's not talk about the LEDs that have died in contact with my 7.2V 85A capable battery pack. It is claimed to hold its voltage at 80A though I've never attempted to discharge it at that rate. It is rather too far past the Vf of most LEDs to be a sensible test pack. It will run a 500W device briefly. 500W through any LED is definitely cruelty.

It’s my understanding a linear regulator like the 7135 works a bit like a resistor but provides a constant current (which also means it’s inefficient with high voltages since it just burns off the extra voltage as heat). This also means it’s also a bit like a buck regulator, although doesn’t have much of a voltage drop but perhaps most importantly, it doesn’t perform like a boost driver and can’t provide a voltage higher then what you start with (basically it becomes direct drive if the voltage is too low). So if the Vf in your LED is high enough, the driver can’t do anything. http://www.candlepowerforums.com/vb/showthread.php?t=247543 Aqualab: LED Driver List - LED Drivers and Regulator Boards (database driven) provide some details. That’s the why in most lights the brightness drops off as the battery voltage runs down (preferably your light will start off at maximum brightness and only reduce near the end but depending on the Vf it can easily never reach there).

You could try bridging the reverse polarity protection diode, as the 7135s get their reference voltage via the MCU. I tried this on a NANJG 101-AK, modded with a fourth 7135 to drive an XP-G R5, because it would only draw about 1.2A. Worked great, now it draws the expected 1.4A.
Tido

Did this help with fresh 18650 batteries? Or some other kind of batteries or unfresh ones? Reason being from what I understand it’s unlikely to help with fresh 18650 batteries and in fact will generally only help if the Vf of the battery is really low (between 2.7-3.3V) and the voltage is within that range. See the posts by Torchboy http://www.candlepowerforums.com/vb/showthread.php?t=192677&page=2 (several of them) which offer some explaination.
Although this is a bit late, it occurs to me that it’s probably best to test your LED under direct drive. With a P7 or MC-E this is usually safe I think, just be a bit careful. If the current is above 3A then using a linear regulator may work well (and you should also not test for too long). If the current is below 2.8A then you are already driving the LED at the maximum recommend current under DD so adding a linear regulator or a buck driver won’t help unless you want it at that constant low level or for different cells. Of course your best bet is probably to test the Vf under different currents with a reliable constant current source if you have that.
BTW some more thoughts to try to improve output. I thought SB would provide a driver with good soldering, wires etc for the output current but may be not. See http://www.candlepowerforums.com/vb/showthread.php?p=3279436#post3279436 http://www.candlepowerforums.com/vb/showthread.php?p=3299582#post3299582 So yeah replacing the lead wires as you mentioned in your earlier post is likely to be a very good idea.

Hmmm . . . . I tried direct driving the LED from an 18650 at 4.15V and the current draw was only 1.56A. I was using some heavier leads from some solid wire phone cord I had. So it isn't the driver. Now it's down to either the meter or the batteries just can't produce that kind of amperage.

Disconnect the meter and see if it gets any brighter with a wall bounce or something? You could use your camera to help presuming you can set everything manually.

BTW is that voltage when being driven?

Awesome review brted! Very professional and great pics and usual. Thanks! Sticky'd and Frontpage'd. Sorry I missed it until now, feel free to PM me if a nice review slips by me.

Sorry for the late answer, I've been a bit busy the last few days. The batteries where fresh of the charger at 4,2V. After looking at the driver schematics and dusting off the knowledge from the basic EE course I had to take in university, I can see no obvious reason why removing the diode should increase the current through the LED.

You are right, as long as VBatt - VDiode - VMCU > 2,7V, the AMC should work as expected. Without a component level schematic of the AMC it's really impossible to tell what's going on here. All I can offer is an educated guess.

According to the data sheet, the AMC's OUT-pin needs to be buffered with a capacitor if the leads to and from the LED are longer than 10cm and 3cm, resp. The 101-AK is missing these caps and due to a seperate pill and reflector assembly in my torch, the leads are more than 10cm long. Maybe this caused some instability in regulation that vanished when Vdd was slightly raised? I should try soldering some 1µF SMD caps between the chip's OUT and GND pins...

I borrowed a friend's DMM and it has big 18 gauge leads (rated for 20A). Using his DMM I got about 2.8A and then using his leads on my meter I was getting about 3A direct driving the LED with a 18650 battery at 4.06V (at rest). So it's all in the leads. I won't have a chance to solder the drop-in back together and do tests on it for a little while, so for now just ignore my current readings because they are not right. I didn't realize it would make that big a difference. Also I kind of wonder if my DMM is calibrated somehow for its tiny leads and if I start using big leads, will I be getting numbers that are wrong in the other direction?

Best bet is to check it against something known, or at least something you can trust more. Low resistance in the meter and leads is not important for voltage measurements where you want the resistance to be as high as possible. For current measurements you want it to be as low as possible. When a meter is measuring current it is in fact measuring voltage over a shunt which is a (hopefully) high precision resistor with a low value, usually around 0.05-0.5 ohms. An external shunt is probably the best way to go and some nice heavy wire to connect the shunt into the circuit.

I think I will have to invest in new leads for my multi meter as well got my MC-E today shows 1.4a draw.

How bright is your light? Do you have other lights you can compare it to? Does it have that donut hole in the beam? Sorry for all the questions

These leads at DX seem like they are pretty good. I have some on the way now. The writing on the wires in the user picture shows 18AWG which would be 18 gauge wire. That should be plenty.

http://www.dealextreme.com/details.dx/sku.33451

I also bought some leads on eBay that look just like those, but even though they are listed as 10 amp leads, they have thick looking leads that are all insulation around very thin wires and can't measure the current either (I stripped the insulation and figure they are about 26 gauge wire). I told the seller about this and they agreed to send me 20 amp leads which they confirmed had much thicker copper inside. I got the 10A leads for $1.28 at auction whereas they sell the 20A leads for $6.99 so that was pretty nice of them (but they got me to change my negative review of them). Once I get some of the new leads I will do the tests again and post the results.

It is BRIGHT. When it stops raining I will do a comparison shots with it up against my R5's. Yes has a hole, but I suspect that can be fixed by removing alot of heat sink goop.(its everywhere emitter too high)

no worries

Ordering a better set leads from dx .