Yezl Y3 - a picture breakdown

The leads have resistance and resistance cause a voltage drop if you measure current. This dropped voltage is “stolen” and less voltage gives different current.
How the lower voltage behave on a driver depends on the driver.
The stock y3 behaves like that:
4.2V gives 2.3A
4V gives 2.4A
3.8 gives 2.6A
3.5V gives 2A

It’s regulated to some point(around3.7V) and after that power drops.

I just plugged it to a power supply and noted the display current and voltage which is not absolute accurate but enough to enlighten all the differences in tail amps a bit.

I bought 14 AWG silicone multi-strand high temp wire and directly soldered them into bare DMM connectors. On the other ends, I strip off bout 1/2 inch then solder the exposed ends up so it's stiff and chunky. The 14 AWG silicone is 12" long, nice and flexible, easy to work with, but not too long - again, longer it is the more the resistance. I bought the wire here: http://www.buddyrc.com/black-14awg-silicone-wire.html and red, cheap, good quality -- probably best deal around actually on wire at BuddyRC.

For the connectors, dunno - I got mine @work, but you could buy a spare set of leads cheap ($3 - $7 eBay, FastTech, etc.) and cut off the wires.

This is a sweet cheap DMM I have: http://www.fasttech.com/products/0/10002748/1483104-uni-t-ut33d-1-9-lcd-palm-size-digital-multimeter, but I leave it on and kill the battery all the time. So I bought this one: http://www.fasttech.com/products/1004/10002748/1204400 with a battery saver, and 20A capable Wink.

Well I made a few mistakes with the Y3. Wanted to put in a MTG2. First mistake was buying protected batteries for it, well they don’t fit at all but enough to get power through. Can only screw it up 50%. Once I got everything together the beam was just too tight for my liking, I should’ve known this though based on everyone else’s comments. I did double up a R100 like the rest of you. Output was good but it produces too much heat too quickly to be something I want to carry around. I still prefer my older resistor-modded TR3T6 which floods more and is brighter to my eyes.

Anyway took the G2 right out and put the XML back in. Back on the shelf for that light probably!

Oh and instead of shimming the G2 up, I ground down the head of the light at the base. This allowed me to screw the pill further into the body and press up against the reflector, eliminating rattle and pressing the MCPCB against the aluminum.

grantman321 wrote:

. . . Bottom line… I need to figure out how to get this DMM set up right and then go from there. Are there any writeups on BLF about modifying budget DMMs for better accuracy?

I think we cross posted. Did you see my post 141. Your measurements are in line with Werner's measurements using a power supply (Post 143).

But in my post #132 I said I measured 3.55A to 4.0A with the R100 resistor added (just like grantman321's Y3) - Werner's is with no resistor mod. There is still a difference in amps measured between his R100 modded Y3 and my R100 modded Y3.

@ImA4Wheelr, Werner & Tom E — Thank you for the replies! So basically it seems to be reading somewhat correctly, and the R100 doesn’t have much effect when running the light on just one cell? I’ll desolder an LED lead on each one and see what the current is at the emitter with two cells in series. My end goal with these is to have one with an MT-G2 and one with an XM-L2, both on 26mm maxtoch mcpcbs putting out in the neighborhood of 4-5A on high. I’d prefer to run the XM-L2 on one cell, but if two is what it takes, so be it. I went ahead and grabbed two extension tubes when I got the lights during the GB anyway. What got me to the point I’m at now is I’m trying to get the driver as close to where I want as I can before going whole-hog and reflowing emitters and all that. Haven’t even braided the springs yet.

Still definitely going to solder in beefier leads in the DMM, it’s something I figured I’d need to do eventually and now seems like as good of a time as any. Y’all reckon a sizeable local RC/hobby shop would stock wires like that (high temp, low enough gauge)? My understanding of electronics is just marginally decent enough to mod lights with a small degree of competency — I hadn’t even considered the DMM leads would cause a voltage drop which would lead to different readings… I figured they were primarily limiting current.

@aoeu — the MT-G2 wasn’t floody enough?? I find it hard to believe it’d be less than twice as floody as a Y3 with an XM-L2. I’m kinda hoping for a decent balance between throw/flood with my eventual MT-G2 Y3. The Convoy C8 I’ve built with an MT-G2 is mostly all flood… doesn’t throw hardly at all compared to a dedomed XM-L2 in a C8. Hoping the much-larger Y3 reflector will be big enough to give it a decent balance with the bigger emitter.

Were you measuring with two cells or one?

aoeu - I believe the tube extensions are not sized correctly. With the stock light, a 26700 cell fits fine. With 1 extension, things get tight (maybe too tight) for 2 protected cells. With 2 extensions, 3 unprotected get very tight. Ugh.... You can see as you add extensions and cells, the last cell sticks further and further out of the end... This is sort of a disaster really for use of the extensions. Suppose you can mod something somhow to get some space back...

Edit: I'll have to look at that head base - bout grinding it down. Nice tip!

gman - I used one cell, clearly stated in post #132 ("one" and "a"), so we are still getting different results. Ohhh - my EVVA unprotected is a 4000 mAh (good re-labeled KK).

For the MaxToch 26 - I didn't finish it yet, but sized it up for the Y3 I'm working on, I'll have to re-drill one LED wire hole (poor positioning in the stock light) and possibly add a copper spacer to make up vertical space. I'll be modding the plastic alignment piece by cutting it down, then sanding the bottom to make it much thinner - it's way too thick now, keeping the LED further out of the reflector than it should be. It looks like the wires will clear the reflector bottom, so if the alignment piece is filed down, you can really get a nice positioned LED out of it -- meaning more inside the reflector.

if we add a resistor marked “0” will we gain anything ?

ImA4Wheelr wrote:

grantman321 wrote:

. . . Bottom line… I need to figure out how to get this DMM set up right and then go from there. Are there any writeups on BLF about modifying budget DMMs for better accuracy?

I think we cross posted. Did you see my post 141. Your measurements are in line with Werner's measurements using a power supply (Post 143).

Tom E wrote:

But in my post #132 I said I measured 3.55A to 4.0A with the R100 resistor added (just like grantman321's Y3) - Werner's is with no resistor mod. There is still a difference in amps measured between his R100 modded Y3 and my R100 modded Y3.

That's why I asked in my previous post if anyone has measured with one cell. I don't know what the difference is about. It seems like what you said about the stock DMM's leads could be a big part of it. Stock DMM leads become a bigger bottleneck as current increases.

"0" resistors are jumpers - same effect as jumpering a wire across the pads. So if you do that over one of the driver resistors, you are nulling out the effect of those resistors. I didn't try it - might work at high amps, might fry something because of super high amps, might simply fry something because the driver can't take it - a risk, unless you don't mind losing the driver and/or LED, etc. Maybe someone tried it already? Not sure...

thanks for the info TomE , i prefer not to take the risk :slight_smile: ,especialy for the driver will be very dificult to find another .

If someone mods this, please make some pictures so that I can just do the same without hassling around to much to find a good focus.

Werner explained this above, but it’s quite possible to measure high current with thin leads… as long as there is plenty of voltage overhead available. Between voltage sag and the relatively high sense voltage which these buck converters consume and the voltage dropped on the thin leads grantman321 has run into his problem. When he measures 3A or 4A on a linear driver he’s probably using a very strong cell (doesn’t drop much under load) and a relatively low Vf emitter and just barely getting away with it. As you guys have all already mentioned, the problem is certainly going to be the leads. I just thought I’d expound on why grantman321 may have gotten away with 3A measurements.

An exaggerated example:
Stable 5v input for 3v output means the linear driver “burns” 2v. If small multimeter leads “burn” 1.5v, there is no perceptible difference at the 3v device. Current remains the same, output remains the same, and heat generation at the linear driver is reduced. Using the same setup with a 4v supply you’d notice a drop in output when you inserted the multimeter.

As I commented in the GB thread, there is never a good reason to short the sense resistors. There is nothing a short can achieve that a standard resistor mod could not. Figure out the drive current you would like to achieve and then do the math I posted in MRsDNF’s GB thread (post #417). Don’t forget that after you determine the target resistance you must use a parallel resistor calculator to figure out what you should stack!

missed that ,thanks.

Figured it out - as explained by wight, voltage drop was the issue. Tested both lights with two cells. Stock driver pulled 2.5A on high and the driver with an R100 added pulled 4.6A on high. Looks like two cells (or three) are needed to get the driver to perform with the resistor mod.

Not true for my R100 mod. I get 3.5 to 4.0A with one (only one) cell. You really need the heavy gauge DMM wire leads. I got it torn apart right now - in the middle of the MaxToch MCPCB upgrade - will use a XM-L2 U2 1A keeping the dome on.

As Tom E said, I don’t think you really grasped my explanation. The key is that we don’t have a lot of “extra” voltage (headroom) to work with when driving 1 emitter on 1 cell, especially with a buck driver. That doesn’t mean that the resistor mod won’t work. It does mean that you can’t accurately measure current the way you are attempting to.

That said, I can’t guarantee that every light is identical. Maybe you did get a bum light: one which behaves unlike the ones tested by Tom E and Werner. We know from collective experience though that standard DMM leads are always a problem, so I think it’s safe to assume that we’re right about it this time. Not that it means a ton, but look at the post counts of these people telling you that the DMM leads are the problem ;-).

Nah — I got what you were saying… Just looking towards the next step - if the DMM leads are causing a voltage drop such that the buck driver doesn’t have a high enough voltage to put higher amps to the LED at the LED’s forward voltage - the performance of the light (absent any interference from the DMM) would be good on a fully charged battery, but should degrade quickly as the battery drains given the voltage sag of the battery under load. So I was thinking that using two cells would ensure decent driver performance since Vin will always be sufficiently higher than Vout. Of course, as I said before, I’ve got barely enough grasp of all this to be dangerous, so I could be completely wrong.

On the other hand, a good cell with less voltage sag might be fine. But two cells would keep that Vin up high for sure.