Tri-Shark & Remora Wiring Diagram

[Nanomiser]

Being a newer member myself, I thought this may be helpful to other new folks in getting started with their mods. A BIG thank you goes out to LED Zeppelin for being gracious enough to hold my hand through this process and to dat2zip for his time spent in helping me to understand the fundamentals of the Shark and offering up this particular solution in the first place. Please feel free to offer any suggestions or improvements to this configuration. From original thread P.10 here.

 

[DaFABRICATA]

Subscribed...Thanks!!

Nice easy to read lay-out!!

 

[Nanomiser]

Glad I could help!

:thumbsup:

 

[LED Zeppelin]

Mike, I just realized your P7s all need to be in series, not two strings. And you need to use Blue Sharks.

With a Shark Buck solution, you could use two strings, but each running independent Sharks that feed the strings separately.

With the Blue Sharks and two unbalanced strings as pictured, the strings will share the current unevenly. With Shark Bucks and two unbalanced strings, the Sharks will still deliver regulated current to each string, but one Shark will pull more from the cells.

 

[Gryloc]

It is great to see that you stitched that picture together. The wiring to the LEDs are kind of confusing since the lines were not as clean as what you would see in a circuit diagram (perpendicular lines running horizontally and vertically). It is still readable, though.

I did notice one glaring issue that has me troubled though. It seemed to be hidden in there, but once you straighten things out, you see it. I went ahead and re-drew the electrical paths in the following picture. I used to draw the little humps in my lines to show that they are not connected electrically, but in the following image, crossing lines do not mean that they are connected, but crossing lines with a dot does mean that they are connected.



That is my issue: Three LED emitters connected in parallel with four LED emitters. The Shark drivers are totally blind as to what is on their output. The Sharks will band together and they probably will deliver the same currents according to the signal from the Remora board.

Actually, something undesirable can happen. I am not sure on the topography of the Shark board, but I thought that I heard it is sort of a voltage regulated board. Someone please clarify on this. If the Sharks is current regulated (say each delivers 1000mA), then the voltage will adjust automatically so one string will receive 3000mA. In this case, the string with three emitters (with a lesser combined forward voltage) will receive 3A of current, and because the string with four emitters have a large combined forward voltage, will get barely a trickle of current. On the other hand, the voltage may increase to try to cater to the string of four P7s and the drivers may damage the three P7s. I am not 100% sure about what will happen since I have not owned a Shark driver.


So, I think that you intended to connect all seven of the P7s in series, judging by your large bank of li-ion cells in series (over 22.2V worth). I am not familiar with the max voltage capabilities of the Shark driver, but something seems scary about this setup. I would re-check all of your specs if I were you! The Sharks are step up drivers, but they may be unable to handle such a high input voltage. Can the Sharks deliver ~26V on the output?

Once you can confirm these voltages, then I can see this working okay. Each emitter will see 3A each. With seven P7s at this current and some decent reflectors, then you are talking about some serious light output!! I hope that you are not trying to shove all this into a Maglite. That is way too much heat (~70W) unless you are going to make it a dive light where the body is suspended in some water.

-Tony

EDIT: I guess that I was way too slow on this one. LED Zepplin beat me by a mile (and with fewer words)!

LED-Zepplin, although you are walking Nanomiser through all of this, could you please explain how you can set up the Shark to be a buck mode driver? I know that many driver can be wired in a slightly different configuration to change its topography, but I am curious on how you can do so with the Shark. Is there advantages of using the Shark as a bucking driver? Out of curiosity, is there issues with having 6 li-ion cells in series (when it comes the cells depleting at different rates), even if they are high quality cells? I know it is safe to do with NiMH and NiCad. Thanks!

 

[LED Zeppelin]

Gryloc,

The Shark Buck is a brand new converter just released by Wayne, it hit the shelf yesterday. It on the Shoppe's site here.

There's also a thread about it here.

I'm not a cell expert, but I don't believe there is any issue with 6 Li-ion cells in series. The issue is when charging - though they can all charged simultaneously with a charging jack-equipped holder, they should be balanced periodically so the states of charge are not dissimilar. If they get far out of balance there is a chance of reverse current which is the danger.

 

[Gryloc]

I looked at the Shark Buck thread and found more info. Sounds awesome! In addition, I found more about your project, nanomiser. From just this thread, it was hard to see what your intentions were. Now it is more clear. You are also in great hands at the moment, so I won't interfere too much . It looks like you will have a real monster when you are finished. I cannot wait for beamshots.

I am going to have to re-think some previous projects now that a 3A buck circuit is available :thinking:. I am so happy because usually to get any sort of dimming function (with a potentiometer), you had to use a boosting circuit if you would stick with dat2zips's or george's drivers. Thanks for the news!

As for the series connected li-ions, I just had some sour experiences with using cheap and unprotected cells. I could charge two cheap 26500 cells in parallel for guaranteed balancing, but after using them only once in series in a light that drew maybe 2A from the cells, one cell would lose its voltage faster than the other. Even with good protection (and still cheap cells), I could still see this happening. However in that case, the entire light would shut off if once the voltage of one cell drops too low. Now I have two of AW's C cells to replace those cheap cells and I will monitor this over the weeks. I probably will be fine. I guess, to me, balancing 6 cells sounds like a small chore unless you have the balancing connector built into the 6-cell li-ion battery carrier. Good luck, nanomiser!

-Tony

 

[Nanomiser]

LZ,

Thanks for the correction and my apologies for the confusion on Shark type. Yes, my intensions are to use Blue Sharks so maybe I need to clarify that on the diagram. I will go back and revise the drawing again and post it here for more scrutiny. This is exactly the kind of response I was looking for, Thanks.

Mike

Mike, I just realized your P7s all need to be in series, not two strings. And you need to use Blue Sharks.

With a Shark Buck solution, you could use two strings, but each running independent Sharks that feed the strings separately.

With the Blue Sharks and two unbalanced strings as pictured, the strings will share the current unevenly. With Shark Bucks and two unbalanced strings, the Sharks will still deliver regulated current to each string, but one Shark will pull more from the cells.

 

[Nanomiser]

Hi Tony,

Thank you for your input as well. I got a similar response from Wayne, but I think the Mammoth is just the host for this particular mod. I just need to make real sure I have all my ducks in row when I fire this one for the first time. That's were all of you expert come into play in helping newbie's to learn and hopefully do it right the first time.

I always error on the side of caution when it comes to Li-ion cells and won't buy anything else but fully protected, preferably AW. Up until your last response I wasn't completely clear about cell charge balancing, but the parallel charging comment finally made it click! I will need to reconsider my approach to how I will be charging all of my cells. How often is it necessary to charge balance a Li-ion?

I guess I am back to the virtual drawing board for another revision. I'll post new drawing when I can.

Mike

I looked at the Shark Buck thread and found more info. Sounds awesome! In addition, I found more about your project, nanomiser. From just this thread, it was hard to see what your intentions were. Now it is more clear. You are also in great hands at the moment, so I won't interfere too much . It looks like you will have a real monster when you are finished. I cannot wait for beamshots.

I am going to have to re-think some previous projects now that a 3A buck circuit is available :thinking:. I am so happy because usually to get any sort of dimming function (with a potentiometer), you had to use a boosting circuit if you would stick with dat2zips's or george's drivers. Thanks for the news!

As for the series connected li-ions, I just had some sour experiences with using cheap and unprotected cells. I could charge two cheap 26500 cells in parallel for guaranteed balancing, but after using them only once in series in a light that drew maybe 2A from the cells, one cell would lose its voltage faster than the other. Even with good protection (and still cheap cells), I could still see this happening. However in that case, the entire light would shut off if once the voltage of one cell drops too low. Now I have two of AW's C cells to replace those cheap cells and I will monitor this over the weeks. I probably will be fine. I guess, to me, balancing 6 cells sounds like a small chore unless you have the balancing connector built into the 6-cell li-ion battery carrier. Good luck, nanomiser!

-Tony

 

[Gryloc]

There are different ways to balance a pack of li-ion cells. Two that I know of is parallel charging, and by using a specialized charger.

Charging in parallel requires (sometimes, depending on design) that you remove each cell from the pack (if using a battery holder/carrier) and charge them all in parallel. The problems with that is it takes a high amperage charger to charge many cells in a timely manner. I like the method when I charge two 18650 cells or two of AW's C-cells using a laboratory power supply (do not own a high quality li-ion charger). It does not take too much time for me. I do not own any series wired li-ion packs. Some day, when I have the funds, I desire making a light on the same scale as yours.

On the other hand, some chargers charge all of your cells in series, but they use an additional connector to properly balance the cells while charging. They require an additional connector with a wire (of smaller gauge) that connects between each cell. For example, for six cells (a cell is represented below as +{]- ), you would have two main power leads, and then an additional connector that has five wires (A-E):

+{]A{]B{]C{]D{]E{]-

main leads: (+) and (-)
extra connector: [A B C D E]

The electronics will monitor each cell and increase or decrease the voltage over every cell while charging to make sure that they are about the same voltage each (I thought that I heard that they balance withing 0.02V each). The connectors are pretty small, so if you have a pack made of you six 18650 cells, and the cells are permanently connected, then you can have a small male or female connector located somewhere that you can plug into the charger. I guess the only downside is maybe the cost of one of these chargers. I am not sure how much they are, but in comparison to the overall cost of your tremendous project, it will probably be small. Of course, there are many more knowledgable people here than me about the li-ion cell technology. Just snoop around the "Flashlight Electronics" area. Good luck!

-Tony

 

[Nanomiser]

Hi LZ & Gryloc,

I hope this one has got it right! It's cleaner at least.

With this configuration wouldn't the P7 string see too much Vout from the 3 Sharks?

 

[LED Zeppelin]

Looks good to me Mike. Just remember the two top trimpot pads should be soldered together:

With this setup the P7s should see 2.7-2.9 A each.

 

[Gryloc]

Nice re-drawring. Very clean!

The three Sharks will provide what ever voltage is needed on the output as long as the current stays at what it was set at. Everything should be great as long as the total Vf does not go beyond ~32V (the max rated output voltage). If that happens (lets say you attach 2 more P7s in series), then the current output from each Shark will not be the full 1A that you intended. Come to think about it, is there enough room in the very nice configuration of seven emitter for an eighth? Electrically, it should work. Now, where to fit/shove it... :laughing:

Led_Zeppelin, what about the input voltage? Is it too high? With 6x li-ion cells hot off the charger, that is 25.2V. According to jtr1962 in his White LED lumen testing thread, with seven P7s operating at 3A, they will have a combined forward voltage of 26.29V (3.755V x 7). Is that enough overhead? This is just from one sample P7s, so there could be a variation, so the total Vf could be slightly higher or lower. I can see the voltage dropping quickly on the fresh li-ion cells (sagging?), and the cells should settle down eventually to around 22V for a while. Now, what happens as V-in gets close to V-out? Even if this is perfectly safe (which I am sure it is), then I hope you do not intend to dim your seven P7s, Nanomiser. I remember that you need a pretty big overhead to dim the LEDs down to a miniscule current. For example, I believe that yout V-in has to drop below 19.8V (2.83V x 7) before you can reduce/dim the current to only 80mA to your seven LEDs (that is less than 3.3V per li-ion cell).

Well, all that aside, I am excited for you, Nanomiser! Out of curiosity, what flux bin of P7 emitters are you using? I have been communicating with jtr1962 lately, and he will be testing some D-binned P7s that I got from PhotonFanatic to make lumen measurements. He may post his findings on his thread. He told me lately that he found that the "C-bin" P7 that he tested (the one that I linked above) was actually a D-bin P7. I wonder when he will correct the post in his thread. It is sort of a bummer because I was hoping that the performance of the D-bin P7s would be better. Even if you use his measurements, that means 5,460 lumens at 3A! Congrats! I hope the LED guys beat out the incan dudes...

-Tony

 

[Nanomiser]

Hi Tony,

Thanks for the good charging information. So far I only own loose batteries, no packs. I do however have a few of FM's custom holders that I plan to use for charging a number of my AW cells. Some months back I did make a fairly good investment on a Triton2 Electrifly charger for a very reasonable price.

I think this qualifies as a quality charger which should allow me to charge at least a couple of cells in parallel at a time or several in series. In addition I've also upgraded my cheap soldering irons with a hi-watt digital Weller that should enable me to achieve much better and smaller solder joints for all of these torch builds. Thanks again for your input here as it should prove to be helpful moving forward.

Mike

There are different ways to balance a pack of li-ion cells. Two that I know of is parallel charging, and by using a specialized charger.

Charging in parallel requires (sometimes, depending on design) that you remove each cell from the pack (if using a battery holder/carrier) and charge them all in parallel. The problems with that is it takes a high amperage charger to charge many cells in a timely manner. I like the method when I charge two 18650 cells or two of AW's C-cells using a laboratory power supply (do not own a high quality li-ion charger). It does not take too much time for me. I do not own any series wired li-ion packs. Some day, when I have the funds, I desire making a light on the same scale as yours.

On the other hand, some chargers charge all of your cells in series, but they use an additional connector to properly balance the cells while charging. They require an additional connector with a wire (of smaller gauge) that connects between each cell. For example, for six cells (a cell is represented below as +{]- ), you would have two main power leads, and then an additional connector that has five wires (A-E):

+{]A{]B{]C{]D{]E{]-

main leads: (+) and (-)
extra connector: [A B C D E]

The electronics will monitor each cell and increase or decrease the voltage over every cell while charging to make sure that they are about the same voltage each (I thought that I heard that they balance withing 0.02V each). The connectors are pretty small, so if you have a pack made of you six 18650 cells, and the cells are permanently connected, then you can have a small male or female connector located somewhere that you can plug into the charger. I guess the only downside is maybe the cost of one of these chargers. I am not sure how much they are, but in comparison to the overall cost of your tremendous project, it will probably be small. Of course, there are many more knowledgable people here than me about the li-ion cell technology. Just snoop around the "Flashlight Electronics" area. Good luck!

-Tony

 

[Nanomiser]

LZ,

Thank you for that important reminder and for all of your help.

Mike :thumbsup:

Looks good to me Mike. Just remember the two top trimpot pads should be soldered together:

With this setup the P7s should see 2.7-2.9 A each.

 

[LED Zeppelin]

Led_Zeppelin, what about the input voltage? Is it too high? With 6x li-ion cells hot off the charger, that is 25.2V. According to jtr1962 in his White LED lumen testing thread, with seven P7s operating at 3A, they will have a combined forward voltage of 26.29V (3.755V x 7). Is that enough overhead? This is just from one sample P7s, so there could be a variation, so the total Vf could be slightly higher or lower. I can see the voltage dropping quickly on the fresh li-ion cells (sagging?), and the cells should settle down eventually to around 22V for a while. Now, what happens as V-in gets close to V-out? Even if this is perfectly safe (which I am sure it is), then I hope you do not intend to dim your seven P7s, Nanomiser. I remember that you need a pretty big overhead to dim the LEDs down to a miniscule current. For example, I believe that yout V-in has to drop below 19.8V (2.83V x 7) before you can reduce/dim the current to only 80mA to your seven LEDs (that is less than 3.3V per li-ion cell).
-Tony

Tony, I figure the voltage sag will be enough to keep the Sharks in boost and not in direct drive.

And you're right about the dimming - at least on full cells I don't think that you'll get a low level with a Remora. You could use a dummy cell if that's an issue. It depends on how much the voltage sags and the actual Vf of the string. The Remora low isn't that low, so 18-19V should get you there.

 

[Nanomiser]

LZ & Tony,

I've been thinking about the number of cells I plan to use and came to the conclusion that it may not work. With Sharks in parallel to 6 cells they would all see ~25.2 which would well exceed the Blue Shark 20Vin limit in addition to possibly putting them in DD because of the ~25.2 Vf, correct?. Since I will be using FM's 4x18650 cell holders populated with both live & dummy cells, Wouldn't I be better off only using five 18650 + 3 dummies? At full charge 5(4.2) = 21Vin enables boost mode and dimming as well as getting a end of charge 13.5Vin which is still above half the string Vf. However, that condition would create an ~5.25 current draw from cells through the drivers, but divided among the 3 Sharks should be safe, maybe? Or am I better off with my original 6x18650 design?

 

[Nanomiser]

The dimentional design of the Mammoth head and the number of MCR27R reflectors I wan't to use limit me to seven P7s. Very interesting results from jtr1962 testing. I have also conducted an intial test on one of my D-bin P7s. Although not nearly as comprehensivew as jtr1962, I did observe some interesting results. My power supply would only allow the ~2.8A draw once I the P7 saw a 3.9Vin, why? If this is acurate then I should see a Vf sting well above the 25.2Vin keeping me in the window with 6x18650 cells.

Nice re-drawring. Very clean!

The three Sharks will provide what ever voltage is needed on the output as long as the current stays at what it was set at. Everything should be great as long as the total Vf does not go beyond ~32V (the max rated output voltage). If that happens (lets say you attach 2 more P7s in series), then the current output from each Shark will not be the full 1A that you intended. Come to think about it, is there enough room in the very nice configuration of seven emitter for an eighth? Electrically, it should work. Now, where to fit/shove it... :laughing:

Led_Zeppelin, what about the input voltage? Is it too high? With 6x li-ion cells hot off the charger, that is 25.2V. According to jtr1962 in his White LED lumen testing thread, with seven P7s operating at 3A, they will have a combined forward voltage of 26.29V (3.755V x 7). Is that enough overhead? This is just from one sample P7s, so there could be a variation, so the total Vf could be slightly higher or lower. I can see the voltage dropping quickly on the fresh li-ion cells (sagging?), and the cells should settle down eventually to around 22V for a while. Now, what happens as V-in gets close to V-out? Even if this is perfectly safe (which I am sure it is), then I hope you do not intend to dim your seven P7s, Nanomiser. I remember that you need a pretty big overhead to dim the LEDs down to a miniscule current. For example, I believe that yout V-in has to drop below 19.8V (2.83V x 7) before you can reduce/dim the current to only 80mA to your seven LEDs (that is less than 3.3V per li-ion cell).

Well, all that aside, I am excited for you, Nanomiser! Out of curiosity, what flux bin of P7 emitters are you using? I have been communicating with jtr1962 lately, and he will be testing some D-binned P7s that I got from PhotonFanatic to make lumen measurements. He may post his findings on his thread. He told me lately that he found that the "C-bin" P7 that he tested (the one that I linked above) was actually a D-bin P7. I wonder when he will correct the post in his thread. It is sort of a bummer because I was hoping that the performance of the D-bin P7s would be better. Even if you use his measurements, that means 5,460 lumens at 3A! Congrats! I hope the LED guys beat out the incan dudes...

-Tony

 

[Gryloc]

Nanomiser,

Boy that thing just seems so intense. Looking at the P7 in my Mag host, I don't think that I can imagine seven!

That voltage you read is not the forward voltage of the LED. It is the voltage read across the thin positive lead wire, then the LED, then the negative lead wire. There is a voltage drop in the lead wires since the wires have a bit of resistance in them. It is easy to get the true Vf.

First, do you have a separate multimeter? Second, is that power supply provide both constant voltage and constant current? If you have a spare meter, then measure the voltage directly across the emitter (at the power legs). I learned from evan9162 a nice method to measure the correct Vf of an emitter when running V vs I tests.

If that the PS can provide a constant current, then as long as you set an amount of current, then it will care less about voltage out. In that case, you can use thin, high resistance power wires if you want. If the power supply just allows you to set voltage, then you may have to tinker with the voltage a bit to get the right amount of current. Either way, you will need a second meter.

So, what evan9162 told me to do was to use two sets of probe wires soldered onto the emitter. He told me that he used that thin gauged wire found in IDE ribbon cable that was left over from a computer. He stripped it so he had individual strands. These are really thin, but can handle more than 3A without increasing in temperature. Actually, any spare hookup wire will work. Solder the power wires coming from the power supply to the (+/-) pads on the star, then solder your thinner gauged wire onto the actual legs of the emitter. I say smaller gauge wire because if the wire you use is too thick (I mean like 18AWG), then you may but mechanical stresses on the legs if you move the wires around a lot. Finally, attach the thin hookup wire to your meter. Since there is absolutely no current going through the thin wires attached to the emitter's legs, then there will be absolutely no voltage drop! At this point, do not read the voltages that the power supply gives. I noticed with my power supply, that even with thick gauged wire, there is a big voltage drop.

I hope that helps. If you do not want to do all that, then just try to clip the probe wires from the multimeter to the legs of the emitter.

BTW, it does seem silly to have a huge battery compartment filled with so many spacers. The V of the fully charged cells is 25.2V, which is close may be at the max voltage, but you could always let the cells rest a bit before putting them in the light. If you are really worried, just put a load onto the battery pack before inserting for a few seconds (like an old 24V light bulb or something). Well, after the light runs for a few minutes, then the voltage will sag down some. When the li-ions are at more than their halfway point, then the voltages will only be 22.2V (3.7V x 7). Do not forget that the seven P7s will be running at around 26.3V at 3A. The circuit wont mind as it will have to boost the voltage.

I bet Dat2zip could tell you if the voltage is too high (only .2V). The 25V rating may be just what he tells his customers, knowing that the customers will most likely push the limits. The true max voltage may be 26V. You may be able to ask Dat2zip if there is a easy fix. I remember someone taking a boost circuit and wanted to add one extra li-ion cell to it. It would mean a voltage in being slightly over the max input rating. So, they placed a cheap 3mm red LED in series with the Vcc pin on the main IC chip. It made the chip happy, while the rest of the circuit handled the slightly higher voltage just fine. I think this was done on TaskLED's Fatman boost circuit. That is one thing you can do (just a thought)...

-Tony

EDIT:
I just seen that you do have a multimeter (is that a Fluke?). Great. I just asked jtr1962 in his thread about what bin of P7 he used. I noticed that the P7 you got from KD was an I bin. Well, if jtr1962 has tested a J-bin, then your total Vf of seven P7s (if an I-bin) may be a bit lower than 26.3V after all. We'll see.

 

[Nanomiser]

Hi Gryloc,

When you say 25Vin limit are you sure your not thinking of the Shark buck? I just checked the Blue Shark specs at the Shoppe and I thought it said 20 volt max Vin. Maybe I'm missing something.

Mike