Newer Convoy S2+ with newer 365nm UV Nichia LED

[SteveE]

Perhaps there is some regulatory limit on output in some jurisdiction? They might have had to switch parts, so they adjusted the circuitry to provide the same output?

[ulf]

The two 7135 chips are linear low dropout constant current regulators. I commented on them in some other thread.

If the old and new torches have the same type of board, the current to the LEDs will be rather similar.

 

I guess this is the case:

Power-LEDs can often be bought in different selection batches, both for efficiency and wavelength.

The less efficient ones are often cheaper.

If the older type of Convoy used a higher selection of efficiency for the LEDs, the newer type could be produced at a lower cost, giving the same performance as the old one.

[Stefano]

20 minutes, the newer one is running slightly warmer.

New Old

107 F 98 F

42 C 37 C

That's exactly the opposite of what I guesses would have happened in post #5. If the technical data about the LEDs is true, we shouldn't have this situation in real life. There is something wrong, I don't know where, but definitely these real life tests don't match the numbers you provided (provided by Nichia). This needs more investigation.

[Cadmium]

Done a quick search, it seems to be a multimeter, just like the one I have.

 

Set it on high current (it should go up to 10 A if I am not mistaken). Connect the LED assebly to your battery, maybe use jumper wires for that. Connect the multimeter in series, where you want, the position doesn't matter. This will give you the amount of current you are drawing from your battery, basically how much power (I*V) you are drawing. It should be the same for both LEDs. Tell me if this was clear enough.

 

I move the red probe connection to the left hand side, 10A MAX FUSED, then I set the dial to mA / A, correct for far?

So, I can put the meter in series between the battery + and the spring contact (assuming I have the body of the brass housing connected to the battery -, is that all correct?

Or do I need to put the meter in series with one of the LED wires?

Let me know, got it all ready to go, but I don't want to blow myself up, I await your direction.

[Stefano]

Don't worry, I already shorted a car battery and the mains in my old house by forgetting to change my multimeter from current to voltage, so I have experience messing thing up (with the car battery I literally melted my wires together while holding the multimeter, but I instinctively threw it and I did not get hurt).

 

Since the current will surely be higher than 200 mA or the maximun current your multimeter can measure in "normal" mode, you have to keep it set in the 10 A mode (remember to change plug for this mode). Basically you want to build a loop containing your battery, the LED assembly (connect the driver, not the bare LED), and your multimeter. Current must flow through everything, like water in a tube.

 

If you want to think it in another way, connect the battery and the LED driver. This will turn on the LED, like it normally does inside the torch. But now introduce your multimeter, in 10 A mode into the loop, and make the current flow through it too. You should have only one loop.

[Cadmium]

They vary, mostly depending on the battery I use, all of which are freshly charged, but not all the same batteries.

Using the same battery in one old, and three new, Convoys, I get about 600 mA up to about 800 mA depending on the particular battery.

The difference between between the one old, and the three new Convoys (using the same battery in all 4) differs slightly, but not much.

Mostly the difference seems to be pertaining to the battery used.

[Stefano]

At 4.2 V (voltage of a fully charged Li-ion cell) you get 2.52 W at 600 mA and 3.36 W at 800 mA (the voltage may sag a bit under load, but this isn't a big load. Also, your multimeter and the wires you used (especially if thin) have some resistance, but hopefully it didn't affect your measurements too much). So your Convoys are using about 3 W of power, and this is also about the amount of power getting into the LEDs (the output power of the driver is the input power (what you basically measured) times the driver efficiency, which (I hope) is at least 80%, so 0.8). What matters is that you didn't see much of a difference between the LEDs, but more between the batteries. The two LED generations have the same forward voltage and the same forward current, so they use the same amount of power, and it makes sense that both assemblies draw the same amount of power from the battery.

[Cadmium]

Stefano, Wow, I am impressed. Thank you!

 

Ulf, I opened one of the new board housings (brass unit), pulled the board out (I was able to do that without unsoldering the LED like I did with the older one),

and it uses the exact same circuit board, but it has 3 of the 7135 35S chips. The older one has 2 of the 7135 chips.

By the way, the older one has 7135 35N chips and the new one has 7135 35S chips, also both have a smaller 'chip' (maybe a diode or something else?) the new one is labeled as S4.

Here is a photo of the board in the newer Convoy S2+ with three 7135 chips.

 

[Cadmium]

By the way, here is the old and new Nichia LED pdf.

They show the new pdf on one of the previous links I posted, but I found the older pdf bookmarked, so here are both.

There may be more info in the pdf that might be interesting.

 

OLD:

https://www.nichia.co.jp/specification/products/led/NCSU276A-E.pdf

 

NEW

http://www.nichia.co.jp/specification/products/led/NCSU276C-E.pdf

[dabateman]

All this has made me almost want to take my convoys apart to see what I got.

I purchased both of mine from Gearbeast in May 2018. One is very bright. The other one is quite dim and has significant IR output. Ihave wondered if different generations or just a bad unlucky flashlight.

 

 

[ulf]

Stefano -

 

I think you are confusing this circuit with the switching driver circuits used for torches with more than one battery. They operate by converting one set of input voltage and current to another, just like a transformer do for AC. During that conversion there are some losses in energy.

 

These driver chips are linear constant current ICs. They just control the current by variating their forward voltage if the battery voltage is high enough. V_LED + V_IC = V_Bat.

 

Essentially the efficiency is 0% as all the losses it must generate doing it's job will be converted to heat. An efficiency value of this type of IC is meaningless as it is operating as a self-controlled serial resistor.

 

https://www.sunrom.c...c-led-driver-cc

https://www.sunrom.com/get/303900

 

For a fixed current and temperature, the voltage over the LED is constant.

The ICs function is to keep the current through the LED constant, avoiding harmful over-currents destroying the LED when the battery's voltage is high.

 

The small extra components on both boards are diodes for protecting against reversed batteries.

The new type is mounted in a technically working way, even if the board's layout wasn't intended for this component's shape.

 

The markings 35N and 35S could be some production date code.

[Stefano]

So basically the driver is just a variable resistor?

 

You can still calculate an efficiency by doing output power/input power (that's a division).

[ulf]

So basically the driver is just a variable resistor?

 

You can still calculate an efficiency by doing output power/input power (that's a division).

No offence.

 

Yes you can do that calculation, but as the result is not meaningful, opposed to energy conversion calculations, like for for dcdc-converters.

 

The board is a part of a voltage divider-chain with the LED. it is not converting electrical energy from one state to another with losses in that control process.

You do not do efficiency calculations for a resistor, only power loss calculations from current and voltage.

[Stefano]

Ok, got it. Don't worry, you weren't offensive.

[Stefano]

Basically a resistor does a sort of conversion, but passively, while a SMPS or a DC-DC buck/boost converter actively converts a set of voltage and current into another, by using an oscillator, an inductor, a diode and a capacitor (to simplify things a lot).

[Cadmium]

All this has made me almost want to take my convoys apart to see what I got.

I purchased both of mine from Gearbeast in May 2018. One is very bright. The other one is quite dim and has significant IR output. Ihave wondered if different generations or just a bad unlucky flashlight.

 

David, I think your Convoys are both the same in all ways as the one Stefano and I have, those have been the same for quite a few years.

It sounds like you just have a bad LED. You have looked at ti and compared it to the Convoy version 1 Nichia LED that I show in the pic?

It should work fine if you can find an LED, and they also sell boards to match up with the newer LED...

So you could convert that Convoy, update it so to speak.

[dabateman]

I briefly looked at the bad one and the black leed looks to be slightly stripped inside due to rotation of the head board containing the led. So it might be worth unsoldering and taking apart to fix.

Yes I think its the same as your first generation as it looks like I need to unsolder the red and black wires to pop the led top off.

 

Although I didn't clamp down the threads to see if I can unscrew the dual hole ring yet. Its on crazy tight on the head.

[Cadmium]

You don't need to do anything on the LED side, leave it be if the wires there are well attached.

Unscrew the rear brass retaining ring. The circuit board will pop out, the wires are twisted, so you can rotate the board and untwist those wires, that should give you enough slack to assess the situation with the board.

Then if you need to you can unsolder or resolder the wires at the board.

The retaining ring is tight, you will need a firm grip, like pliers on the outside of the brass container, and an actual spanner type for the retainer ring holes.

Good luck.

[dabateman]

This is what the board looks like on my bad convoy.

Very different than yours. The chips say D7135 35M

[dabateman]

This is the board on my good convoy.

[Hornblende]

Cadmium,

Did you take a look at this? http://budgetlightforum.com/node/57798

It appears that the SST-10 UV led made by Luminus is also a promising UV LED.

[dabateman]

The Led board seems to be 16mm in diameter.

Would it be possible to solder one of these LG3535 385nm leds to my driver in the convoy?

 

Ebay item:

231933293669

[Cadmium]

David, Yes your board is different, but the same two ships + diode, so it looks the same functionally.

I think your LED is bad. Replace the LED.

[Cadmium]

Cadmium,

Did you take a look at this? http://budgetlightforum.com/node/57798

It appears that the SST-10 UV led made by Luminus is also a promising UV LED.

 

Hornblende, Thanks, looks interesting.

[Hornblende]

Cadmium,

I purchased this UV flashlight with the SST-10 emitter: https://malkoffdevic...2-uv-flashlight

I know this model is very expensive but this is one of the highest quality flashlight brand made in the US and I need a very tough light to bring with me on the field.

 

Anyway, the light produces 1800mw of UV with fresh batteries, I think this is quite powerful. It also appears that the SST-10 emitter does not produce as much visible light compared to other UV emitters. I contacted the owner of the website, he told me the SST-10 is a little overdriven. Given the size the the flashlight there is enough thermal mass to prevent the emitter to overheat.

 

When I receive it I can report to you how it compares to the Convoy S2+ with the old Nichia model.

 

From what I heard the Convoy S2+ with the LG UV emitter is better than the Nichia one.

 

Also, here are more emitter tests:

http://budgetlightforum.com/node/65084

http://budgetlightforum.com/node/59304