I want to run around & set the world on UV

[colinbm]

Thanks Andy,

The peak was at 368nm & 0.2767 Spectral radiance (mW/cm2).

[Andy Perrin]

The peak agrees with Ulf's 367.63nm. David found 366nm and almost 368nm for his two. (Spectral radiance probably doesn't mean too much here because it is only for the specific distance you measured at.)

 

So far we have the reported values:

 

367.63nm (Ulf)

366nm (David1)

368nm (David2)

368nm (colin)

 

It seems that this torch type is pretty close to the nominal value of 365nm and fairly consistent. One question would be: how much does the peak shift after it's been on for a while and got hot?

[colinbm]

Yes that was measured at 600mm from the entrance pupil.

Thanks Andy for your help & patience, it is muchly appreciated.

[colinbm]

A close-up of the ? LED......

 

[ulf]

The black aluminium base looks different, (modified?). It is lacking the tube normally surrounding the reflector and seams to be painted.

This affect the cooling.

If the paint is behind the metal PCB it is an added thermal barrier and the removed tube gives less cooling area.

Without anything pushing the PCB against the aluminium, the thermal paste layer will be thicker than normal.

 

There is a risk that the LED is not as efficiently cooled as the complete torch, but I cannot estimate how big this risk is.

 

I bet that the light pattern is nice

[colinbm]

Ulf, all components of this torch, except the LED & its driver seems to be made of Aluminium/Magnesiun die cast, painted black.

The black that you are referring to is an Aluminium/Magnesiun die cast part, painted black.

[ulf]

That is what I thought.

The Nemo torch's main metal aluminium parts are not made by die casting but NC-tooling in a lathe.

It is very unlikely that magnesium is used as that only would increase the cost and make cooling less efficient.

 

The original black colour is an anodisation that is very thin.

The Nemo has a standard type of black-anodisation while the Convoy has a tougher more expensive matted hard-anodisation.

They are different galvanic processes.

 

On your picture the structure is mottled, making me guess it has been painted.

Paint is much thicker than anodisation and act as a thermal barrier.

[colinbm]

Ulf, the heat generated by this LED seems to be under-control. If it was getting too hot the white plastic surround that centres the reflector would melt away.

[ulf]

No the plastic is not a good indicator that things are OK thermally.

 

Most plastic melts at around 200°C.

The internal semiconductor barrier in a LED, generating the light will start decaying at around 130-140°C. Then normally the case of the LED can be around 100°C - 110°C

If the LED still conducted any current when the plastic melted it would long before have stopped emitting meaningful amounts of UV.

 

Do not confuse the short time of high temperature during soldering with any operational temperatures. Then there is no current passing through the device.

[dabateman]

Seeing that 20mm disc reminds me of your 390/380/370/360nm light. I wonder if the driver is the same.

If you could just flip out this led disc for yours.

 

I did that with a my bad convoy. I flipped out the bad 16mm led board for a LG 3535 385nm led unit. It works amazingly well. But I researched the driver details and they were the same for the convoy and the cheap 5 pack of leds I ended up buying.

I don't know any of the details for the Nemo.

[colinbm]

Thanks Dave, my head is swimming with ideas.....;-)

[colinbm]

I have taken a multi-meter to the torch & with a full battery at 3.7 Volts the torch is drawing 1.1 Amps, that equates to 4.07 Watts.

[Andy Perrin]

4 Watts is not 15 Watts...hmmm.

[Stefano]

I have taken a multi-meter to the torch & with a full battery at 3.7 Volts the torch is drawing 1.1 Amps, that equates to 4.07 Watts.

If with "full battery" you meant "fully charged battery", the voltage is a bit higher, 4.2 V. 3.7 V is the nominal voltage, the "average" voltage the battery has. When measuring current, a multimeter does have some resistance, and you have to be careful about your contacts. Still, this is not 15 W.

[dabateman]

This article should help:

https://www.google.com/amp/s/www.naturesrainbows.com/single-post/2015/05/01/Convoy-S2-365nm-False-Power-Claims---It-Aint-5-Watts-4-35-or-even-3%3f_amp_

 

The Nemo is 4x the convoy as far as I can make out the marketing. Reality is clearly different. So maximum convoy output is 4W, thus Nemo at 4x would be 16W right? Reality maybe quite different though.

Most likely not hitting that.

 

[ulf]

That is a good article that might clarify what I tried to explain here several times before.

[ulf]

So maximum convoy output is 4W, thus Nemo at 4x would be 16W right? Reality maybe quite different though.

No that is not true, if you are talking about optical power.

 

All specifications used by marketing is using the high numbered values, the electrical INPUT power.

The interesting parameter for us is the optical output power.

 

A half dead LED might still consume 4W or 10W or 15W... producing a lot of heat, without generating much light.

It is the combination of a good conversion efficiency with a high power input that is interesting.

 

However it is a bit difficult to measure properly due to very differing beam patterns.

[Andy Perrin]

Ulf and David: Colin measured the input power, which is what the marketing uses. Even the INPUT is not 15W or anything close. For certain the output is not anything like that, it must be far less. I find it hard to see this as anything but false advertising.

[ulf]

Quite true.

 

Looking at the input power and comparing to marketing text is a good start.

 

I was just speaking about general measurements and comparisons, where the essential thing for us is the optical output power

If the torch for some reason consumes that little, it is not right saying it is a 15W torch.

 

However the LED itself might still be a 10W or 15W device that for some reason isn't driven to it's full potential.

That is what the LED looks like. LEDs with that appearance ar often used in professional UV-curing arrays.

[Stefano]

However the LED itself might still be a 10W or 15W device that for some reason isn't driven to it's full potential.

That is what the LED looks like. LEDs with that appearance ar often used in professional UV-curing arrays.

Yes, it can be a 10 W LED. Mine is (the super powerful paper-burning LED) and it does actually draw 10 W of power. I don't know if it can be a 15 W one, that's quite a lot. I certainly think it can be driven to at least 8-10 W. Can it be that the battery can't supply that power? Lithium-ion batteries are capable of much more sometimes, but even 20 W shouldn't be a problem for a 26650 cell.

[dabateman]

 

No that is not true, if you are talking about optical power.

 

All specifications used by marketing is using the high numbered values, the electrical INPUT power.

The interesting parameter for us is the optical output power.

 

A half dead LED might still consume 4W or 10W or 15W... producing a lot of heat, without generating much light.

It is the combination of a good conversion efficiency with a high power input that is interesting.

 

However it is a bit difficult to measure properly due to very differing beam patterns.

 

Ulf I was joking. That is why I said marketing. If marketing says 15W expect no more than 1W in my mind.

[ulf]

But you have a valid point.

Sometimes I tend to go too deeply into the measurement stuff and get derailed about what was discussed.

 

Eventually I will map the current/voltage behaviour of the torch's LED head without batteries.

[Andy Perrin]

Ulf, yes, that would be excellent, especially if you could simultaneously measure the optical power output. Together that would let us estimate the efficiency.

[Stefano]

Eventually I will map the current/voltage behaviour of the torch's LED head without batteries.

Do you have a bench power supply?

[ulf]

No, I have several, each with different benefits.

Here several = 4.