it may work, you would have to test how current spreads between xmls and mtg. you wont know for sure until you try
I’m interested because it hasn’t been done before. Isn’t that enough.
Not for me. Lots of things haven’t been done before - that by itself does not give them value.
In parallel the voltage is the same across each of the chains. Current gets split up between chains. With the MT-G
having lower Vf than the 2 serial chains (assuming near typical Vf) it will get more than it’s share of current. You might look at putting a small resistor in series with the MT-G to get a little bit better balance between the chains.
Im fairly sure I saw a 5 emitter light with 1 MT-G2 in the middle and 4 XM-L2s around it here on BLF...
Thanks. I get that. Voltage stays the same and current is shared.
So current is determined by Vf. Lookup Vf in the appropriate chart to see what current you’ll be getting. Account for sag at that current with the batteries you are looking at. Lookup in the chart again. Go back and forth until things make sense. If what you want is just to light up those 5 emitters then it can be done with 2s1p batteries instead of 3s1p batteries. The draw of one MT-G2 seems to track the draw of 2s2p XM-L2 more closely than you might imagine. Each emitter is at a more or less appropriate-to-one-another current when you put a given voltage across the whole shebang.
Use these two posts:
https://budgetlightforum.com/t/-/19600#comment-417559
https://budgetlightforum.com/t/-/19546?page=1#comment-416063
and compare with the numbers from HKJ’s KK ICR 4200mAh review (these were the best results of the KK cells he tested):
http://lygte-info.dk/review/batteries2012/King%20Kong%20ICR26650%204200mAh%20(Gold)%20UK.html
There are a lot of variables here, but really it's just an exaggeration of the normal DD issues. I assume that this hypothetical situation involves normal cells, such as the KK's you used on the 9x Yupard. Let's scale up the normal single-emitter/single-cell DD problems to a situation with a battery which only handles around 10A but emitters that can draw a combined 30A+ before you account for voltage sag at the battery (real numbers, not a miscalculation)... Of course the voltage nose-dives and then you get to redo the calculation at some sort of unhappy medium. If you allow for an initial voltage sag of 3.6v per cell, you are still looking at a combined 16A draw at the battery. Still more than it can handle, but I feel like that's only slightly higher than where it would settle, at least for the first 10-30sec or so. Probably/maybe it would settle at around a 13A draw, like your Yupard. I wouldn't make a habit out of it. Note the tracking issues that start to show up later in HKJ's 10A test, this is due to the chemistry not supporting 10A very well. Compare that with what happens here and here in HKJ's tests of less impressive KK cells.
When you get a lot of voltage sag, that is the battery's way of saying "don't do that". When the battery gets hot, that's the battery's way of saying "no means no".
If you want regulation you could do a master/slave mod. Zener mod the master to control the current to the mtg and one XML and control 3 other XML’s with the slave. All running from 3s(x)p. Not sure how the Zener does with 12V but it is doable with a voltage regulator for the mcu instead.