The Quark lights thread! (Part 5)

[Owen]

..it should be about 3.6 Vf. That is also the nominal voltage of a single RCR123. Put the cell under load, and the voltage will depress even lower.

"Nominal" is usually considered synonymous with under load. It will be in regulation, but how long that remains true depends on a variety of factors. Keep in mind that if it were DD, the emitter would be getting closer to an amp, possibly higher(?), of current-more than it does in regulation, and current draw from the battery would be somewhere in the neighborhood of 900mA.
Basically, as long as the current and output seen at the emitter would be greater in direct drive than the level designated by the buck circuit, the circuit should be able to operate in regulation. That is not 100% correct, because it doesn't take things like temperature and driver efficiency into account, but might serve as a simplistic way of illustrating how a single Li-ion can work with a buck circuit.

 

[clipboard]

hey guys I have a question. First of all I own 2 quarks both are tactical one is a 123 and the other is a 123 squared. What is the best rechargeable li-on to run in the 123 squared ? 17670 at 4.2 volts or 2 rcr123's at 8.35 volts. I would think the to rcr123's would be better since it is a 9 volt head. But selfbuilt's review would indicate almost the same brightness and better runtimes with the 17670 aw's.....

 

[Marduke]

But selfbuilt's review would indicate almost the same brightness and better runtimes with the 17670 aw's.....

There is your answer.

 

[wapkil]

I can't see the voltage drop being that severe, especially with a good cell capable of decent discharge. If a regular AW RCR123 in my L2m can keep a 3.4-16V deree module happy for over 15min (and it's an XR-E), I don't see why one couldn't keep the Quark bright.

I think that for a typical RCR the voltage will be around 3.4V-3.5V. You can see how some RCRs performed in SilverFox tests. It's not, as you say, something severe - I wrote that the output will probably still be around 80%-90% of the maximum but the light will run out of regulation.

 

[wapkil]

The Quark normal head is 0.9 to 4.2 Volts, Quark 123² head is 3.0 to 9.0Volts.

Do you have additional information or for some reason decided to simply quote the specification?

If the circuit in the 123-2 head is buck only (as David wrote) the 3.0V level for an XP-E LED is absurd. The circuit will not raise the voltage (because it's buck only) and the LED needs at least 3.3V-3.4V to work correctly in the Turbo mode. It will be direct driven and even will have some output but it won't be regulated and won't be anywhere near the specification. As was BTW demonstrated in HKJ's test's, higher than I expected, estimated 100lm output.

 

[Marduke]

However lower modes will still be regulated at only 3v.

 

[HKJ]

If the circuit in the 123-2 head is buck only (as David wrote) the 3.0V level for an XP-E LED is absurd.

No, it is not. It is the same as the 0.9 to 4.2 volt specification. The head will send out light within these voltages, but it might not be in full regulation.

 

[kwkarth]

No, it is not. It is the same as the 0.9 to 4.2 volt specification. The head will send out light within these voltages, but it might not be in full regulation.

But what about what David said? Below 3.0v input the head does not boost, and above 3.0v it regulates up to the point that is needs to buck to continue to deliver specified output to the led. Is he not correct? He said that a single 17670 works well, and that makes sense. It does not make sense to use a single 123 with the 123x2 head. Sure, it might light up for a while, but it will never be anywhere near optimum, and never in regulation.

 

[Marduke]

But what about what David said? Below 3.0v input the head does not boost, and above 3.0v it regulates up to the point that is needs to buck to continue to deliver specified output to the led. Is he not correct?

The voltage threshold of buck/boost is a moving target, dependant on the Vf at the drive current of the level in question.

 

[wapkil]

However lower modes will still be regulated at only 3v.

No, it is not. It is the same as the 0.9 to 4.2 volt specification. The head will send out light within these voltages, but it might not be in full regulation.

This is not a device that is specified to produce some undefined output in some undefined way. The turbo mode is specified to produce 190lm, in the regulated fashion at the input range of 3.0V-9.0V. If the device is incapable of producing 190lm at 3.0V then the specification is incorrect. I believe that such a description is misleading for users less familiar with flashlights and absurd for those who know that it is impossible for this light to work this way. The specifications should not be written to force users to guess which parts of them are applicable and when.

I don't say that it is wrong for the light to work this way. It is not, the device is dedicated for two batteries so if it wasn't written in the specification I would never expect it to work correctly at 3.0V but the specification is simply incorrect.

 

[HKJ]

But what about what David said? Below 3.0v input the head does not boost, and above 3.0v it regulates up to the point that is needs to buck to continue to deliver specified output to the led. Is he not correct? He said that a single 17670 works well, and that makes sense. It does not make sense to use a single 123 with the 123x2 head. Sure, it might light up for a while, but it will never be anywhere near optimum, and never in regulation.

I do not believe it has anything but buck regulation, but at lower settings the led has a lower Vf and it can get into regulation at a lower battery voltage.
Using a single LiIon battery with this light is equivalent to using a single LiIon in many other lights with buck regulator, they all stay at the edge of regulation. This will also get a long runtime, because the current is a bit below specification.

This is not a device that is specified to produce some undefined output in some undefined way. The turbo mode is specified to produce 190lm, in the regulated fashion at the input range of 3.0V-9.0V. If the device is incapable of producing 190lm at 3.0V then the specification is incorrect.

I do not believe that 4sevens has specified either head to produce full output at the the full voltage range, he has specified. Just because both specifications are on the same page, it does not follow that they are linked together.
When I read the specification on the webpage, I do not see anything about about support for LiIon either, only support for 2xCR123 batteries.

 

[Biginboca]

I agree wapkil. I originally purchased the 123 body thinking it could operate my 123x2 head with a 123 primary.

I think based on the graph we have seen the spec should be 3.6 volts to 9 volts since that is what is actually needed for the 190 lumens.

 

[wapkil]

I do not believe that 4sevens has specified either head to produce full output at the the full voltage range, he has specified. Just because both specifications are on the same page, it does not follow that they are linked together.

Oh come on, it's not only the same page, it's also the same section called "Specifications". And even if it wasn't the same section, it would still be a document describing the features of a single product. It's like with everything else. If I buy a device with the "Input: 100V-240V", I expect it to work the same way with the US and European voltages. If it only partially worked in the US, it would be an incorrect description. Even if the functionality is described on a different page of the manual...

When I read the specification on the webpage, I do not see anything about about support for LiIon either, only support for 2xCR123 batteries.

I haven't noticed it but I think it's another problem with this specification. It doesn't have to be an error, it could be omitted on purpose but I think that in the Market Place design threads the Quarks were advertised as working with Li-Ions. Probably it's just a mistake. BTW, it looks like recently the Quark design threads also mysteriously disappeared...

To be clear, I believe that the Quarks specifications are better than many documets published by the competition. I think the output numbers probably correctly represent the actual typical outputs. The runtimes are also usually close to what was tested (although a few random published runtimes are much lower than they should be). Unfortunately it doesn't mean that this specification is good but only that many others are much worse. I believe that there is nothing wrong in pointing out the errors. Especially since they are evidently not intended as a false advertisement. Just normal errors and omissions but I don't see the point in pretending that they are not there.

 

[bcwang]

Even in seflbuilt's review, the 123x2 head with a 17670 doesn't regulate well at all, so a rcr123 is only going to be worse. With 2 you are much better off.

Talking about specs, my AAx2 box mentions 1.8 hour runtime at max. The website 1.3 hours. One of those has to be wrong, or both are.

 

[wapkil]

Talking about specs, my AAx2 box mentions 1.8 hour runtime at max. The website 1.3 hours. One of those has to be wrong, or both are.

I dunno. In the selfbuilt test it ran only 1h 5min. I believe that the 123 box (if they weren't changed) says 1h at max while the web page has 0.8h. I think 4sevens probably expected better runtimes when the boxes were printed. Mistakes in the last minute changes could also somehow explain some strange runtimes in the specification - around twice as low as they should be.

 

[uplite]

"Nominal" is usually considered synonymous with under load.

You're right, I should have said "under this load".

Nominal voltage is typically quoted for a fairly low load...0.05C for some cells...more likely 0.2C for a li-ion. In this case though, the discharge rate is more than 1C.

For some good graphs of RCR123 discharge and voltage sag, see this thread. At 1000mA, most RCR123 cells fall below 3.6V within about 5 minutes.

-Jeff

 

[bcwang]

I'm going to try to see if my efficiency data is easier to understand in this form, though the detail is almost unreadable. To get the original numbers and my interpretation, look back like 25 posts.

nimh=2000mah
nizn=1500mah
14500=750mah
17670=1600mah

 

[kwkarth]

I do not believe it has anything but buck regulation, but at lower settings the led has a lower Vf and it can get into regulation at a lower battery voltage.
Using a single LiIon battery with this light is equivalent to using a single LiIon in many other lights with buck regulator, they all stay at the edge of regulation. This will also get a long runtime, because the current is a bit below specification.

I do not believe that 4sevens has specified either head to produce full output at the the full voltage range, he has specified. Just because both specifications are on the same page, it does not follow that they are linked together.
When I read the specification on the webpage, I do not see anything about about support for LiIon either, only support for 2xCR123 batteries.

Both of your points make good sense to me HKJ, thank you.

 

[Xak]

Even in seflbuilt's review, the 123x2 head with a 17670 doesn't regulate well at all, so a rcr123 is only going to be worse. With 2 you are much better off.

Talking about specs, my AAx2 box mentions 1.8 hour runtime at max. The website 1.3 hours. One of those has to be wrong, or both are.

IIRC in Selfbuilt's review the Q123-2 17670 ran something crazy like 1 1/2hrs or so to 50% and was the best light/battery combo in the Quark series, the 2xRCR123 in the Q123-2 only lasted like 15min.

 

[heathen]

kwkarth - I did say that I have a voltmeter if you will go back and read my post. I don't however have any alkaline batteries to test. I have tried the tactical head on the AA body and it works fine. I have legoed these things everyway possible and the regular head is the only one I am having trouble with. My batteries read 1.46 volts fresh off of the charger. What I am getting at is that this head should work with these batteries and it is not. I think I am going to have to send it back in. Thanks for your help.