Some questions about the Shark driver

[LED Zeppelin]

Hi Wayne,

In your Shark Buck thread you offered up the Blue Shark in a tri-board configuration, 1 Blue Shark w/Remora as master + Blue Shark slave + Blue Shark slave, as a possible solution to my seven P7 in series drive problem.

Do you have any schematic example of how to connect these three boards for that build?

Thanks,
Mike

Mike, here's some info on wiring multiple Sharks. The Sharks pictured are early versions without the copper C, but the wiring should be the same.

This is a pic of a Remora controlled tri-Shark mod (9 X Cree in a 3" Delghi head on a Leef neck adapter). The brass shapes are topside sinks I made to help cool the IC and diode.




This is the wiring of the Sharks for use with a pot:



How to wire multiple Sharks to one dimming control:

- Order the Sharks with the on-board pot removed, or remove the pot yourself.
- Locate the three pads under the pot. Viewing the board from the component side with the pads at lower right, you will see that the larger central wiper pad connects to the top soldering hole just to the right. There are two smaller pads under the large pad. The lower right pad connects to the lower right soldering hole, and the lower left pad does not connect to any hole.
- If using a pot, order a 20K log pot. I use these from Digikey.
- Connect the pads of one Shark to the pot. Viewing the pot with the shaft at top and the leads facing you, connect the large wiper pad (or top right hole) to the center pot lead, the lower right pad (or lower right hole) to the left lead, and the lower left pad to the right lead. I find it just as easy to solder to the pad as to the hole. The wires do not pass any current, so I use very thin flexible wire, 30 ga or so.
- The other Sharks are wired to the mother by connecting the large wiper pads together. Any number of Sharks can be wired together. Alternatively, all three pads of each Shark can be connected in parallel, or all connected to the pot.
- If a Remora is used, connect the Remora to one Shark, and again wire the wiper pads together. All Sharks must have the on-board pot removed.

[Nanomiser]

Hi LZ,

First let me just say you do some very nice work! The type I aspire to do.
Where did you get the aluminum half round driver sinks? I could use some of those.

The picture of the Remora controlled tri-Shark config is probably closer to what I am trying to do. It looks like you wired the three banks of 3 Crees in parallel controlling each bank with one Shark, is that correct? I need to combine the total current output of the Sharks in order to get the 2.8amps necessary to drive the P7s, but some how divide the total voltage supply among the 3 boards while only contolling two banks of 3 & 4 P7s. I plan to use seven AW18650 cells.

I think I have a basic understanding of how the Shark works, but I just can't seem to get my head around this master and 2 slave set up. Any further clarification would realy be appreciated.

Thank you for your help so far,

Mike

P.S. Does the Remora need the use of a momentary switch or does a stock mag switch work fine? Also, what is the max amp rating for a Mag-D switch?

[LED Zeppelin]

Nanomiser, the Sharksinks are available at the Shoppe here.

Driving (7) P7s can be done by using (3) Sharks all wired in parallel off the cells, and parallel to the string of 7 LEDs. The combined output will be about 2.8A. Since the Shark is a boost only, the cell voltage should be less than the Vf of the string for regulation. (7) 18650s will give about 29.4V hot off the charger, and the Vf of (7) P7s about 25V. So the light might be in DD for a short while, but maybe not given the voltage sag of the cells.

What this does mean is that you will not have much dimming, perhaps none since the cell voltage will be greater than the Vf at lower currents. All levels will be hi. A battery configuration of 8 X 18650, 4s2p, 14.8V would be a better cell solution with full dimming.

Using the Shark Buck, you'd need (2) boards, one driving (3) and one driving (4) P7s. Then you could use your 7 X 18650s with full dimming.

The stock Mag switch does not need any modification for use with the Remora, and I've had 5 amps through them without issue.

You want to make sure your cells can deliver the required current, which should be fine with what's been discussed.

Wiring multi-Sharks isn't as bad as it looks. Just wire one normally to the Remora. The other Sharks connect to the first with one wire jumpering the wiper pads of the on-board pot (pot removed). The wiper pad is the larger central one. Make sure you bench test the entire system before doing anything permanent. There are that many more things that can go wrong using so many LEDs and drivers.

[Nanomiser]

In order to stay with the Blue Shark I have the option to use 6x18650 dropping my supply voltage to 25V, fresh off the charger, hopefully avoiding a DD condition and potential damage to the converters. I need to stay with the 18650 cells because of the Mammoth body size and my FM holders.

Once the cells settle my supply should be @22V with a 25 Vf; would that allow for proper dimming? I would also like to avoid drawing high current through my Sharks when the 6 cells hit their cut off at ~16V; I'm not sure if that applies with this configuration.

I keep thinking series circuit voltage divider / current constant, parallel circuit current divider / voltage constant. So wiring the Sharks parallel to the batteries means that each board Vin will be 25V to 22V, but the 18650s would see ~3.1A draw @3.7V and ~4.3A @2.7V correct?

But how does the parallel tri-Shark set up get me 2.8A combine to each P7? Please forgive my beginner-like understanding as I am still on a big learning curve compared to a lot of CPF members. Thank you for the info on the Mag-D switch & Remora in addition to the continued clarification.

Mike

[LED Zeppelin]

I'm not sure you'll get a proper low even going to 6 cells. You might just get the med and hi on full cells. As the cells deplete, you'll get a lower lo.

I think the cells will hit the 2C limit before they reach the 2.7V cutoff.

When the LED leads are connected in parallel with each Shark output they will see the sum. So if each Shark delivers 930 mA, the LEDs will be driven at 2.8A.

[Nanomiser]

I'm concerned that if I go with any less supply voltage, i.e. 5x18650, I may draw too much current through the Sharks as cells deplete. What is the 2C limit? For now I will have to forego completely understanding the combine 2.8A from 3 Sharks in parallel and focus on what physical connections I will need to make in order to wire up these 3 drivers.

So once I wire the 3 Blue Sharks in parallel to the cells I would then take the Blue Shark w/Remora (Master) and parallel wire the Remora GRN & PWM to the 2 slaves GRN & PWM for dimming control, correct? The LED+ & LED- of the master are then some how wired to the 3 P7 & 4 P7 strings or to the slaves, I'm not clear. However, the LED+ & LED- of the slaves are wired to the strings of that I'm fairly sure.

You gave a good suggestion to bench test every thing before permanently mounting all it; what's the best way to heat sink drivers without epoxy?

Can I epoxy the Blue Shark copper C in direct contact with the heat sink surface (metal to metal) or must it be electrically isolated?

Do you advise the use of topside sinks to help cool the IC & diode or does the copper C take care of that?

Thanks,
Mike

[LED Zeppelin]

Nanomiser, the 2C limit applies to protected Li-ion cells - you don't want to discharge them at more than twice the capacity. Since AW 18650s are rated at 2.2AH, you'd want to keep the current below 4.4A.

Once you wire up one Remora Shark normally, then simply wire the wiper pads of all Sharks together. The wiper pad is the largest soldering pad right under the removed on-board pot. You can solder a lead to the surface of the pad, or to the connected soldering hole just to the right of it. There are two holes and the wiper pad is connected to the upper one. Wire all the Sharks to the cells, and all three sets of LED outputs to the LED leads. So you would take all three LED (+) leads from the Sharks and solder them to the LED (+) of your string.

For bench testing I don't heatsink the Sharks, it's only brief testing to verify operation. You can touch underside the boards to see how hot they are getting.

The copper C can be directly epoxied to a conductive surface, just be careful no leads are touching the surface.

You won't need topside sinks. That was something I did before the Sharks had the copper C for high-boost applications. And yours should run relatively cool sice they are not boosting hard (cell votage is relatively hi vs. Vf of LEDs).

[Nanomiser]

Hey LZ,

Something like this?



However, I'm not totaly sure about the Remora GNDs. Should those be connected as well or just leave them the way I have them here?

Thanks,
Mike

[LED Zeppelin]

Mike, you need to remove the on-board pots from the slave Sharks and jumper the pads like the master. Otherwise you're good to go.

[Nanomiser]

Hi LZ,

Remove the slave pots, very good point! I tried to find a shark image w/o the on-board pot, but couldn't. I will go back and add a notation to the diagram to reflect this. Thanks again for baring with me on my learning curve. I hope for the opportunity to return the favor one day.

I am still waiting for a number of parts to complete this build, but when I do I'll start a thread and give you a heads up.

Mike

[Nanomiser]

Revision done!

[LED Zeppelin]

Doh!

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.

[Nanomiser]

Hi LZ,

Tony & I just noticed that if I try to use a Remora / Blue Shark master wouldn't 6x18650 exceed the 20V max Vin for the Remora? What exactly did you use to power your 9-Cree 3-Shark Mody Dick 3" Delghi build?

Mike

[LED Zeppelin]

Hi LZ,

Tony & I just noticed that if I try to use a Remora / Blue Shark master wouldn't 6x18650 exceed the 20V max Vin for the Remora? What exactly did you use to power your 9-Cree 3-Shark Mody Dick 3" Delghi build?

Mike

Mike, the 9-Cree Delghi was powered by (2) A123 cells. Each Shark was wired to (3) Crees in series. I haven't used a Remora with high voltage yet, so I don't know if you'll encounter a problem. I'd bench test the setup, maybe it will be fine.

[Nanomiser]

WOW LZ,

That's impressive! No wonder you were pulling 5.4Amps, but I thought that would be entering the danger zone for cell discharge? What kind of run times did you get? With 25Vin+ my assumtion is that the Remora would not handle it and probably release it's magic smoke . I think implementing an external pot for this build makes more sense. Would that 20K Digikey pot work for this application?

Mike, the 9-Cree Delghi was powered by (2) A123 cells. Each Shark was wired to (3) Crees in series. I haven't used a Remora with high voltage yet, so I don't know if you'll encounter a problem. I'd bench test the setup, maybe it will be fine.

[LED Zeppelin]

Mike, the A123s are capable of lots of current. I actually don't own any, the 9-Cree Delghi light is Donn's. I think he uses both 2 and 3 A123s to power it, loading them into the modular Leef tubes.

This is an excerpt from another thread post I made some time ago that might help with a pot/Shark mod:

How to wire multiple Sharks to one dimming control:

- Order the Sharks with the on-board pot removed, or remove the pot yourself.
- Locate the three pads under the pot. Viewing the board from the component side with the pads at lower right, you will see that the larger central wiper pad connects to the top soldering hole just to the right. There are two smaller pads under the large pad. The lower right pad connects to the lower right soldering hole, and the lower left pad does not connect to any hole.
- If using a pot, order a 20K log pot. I use these from Digikey. Be sure to measure the resistance of the pot (across the outer leads). The production tolerance is wide (20%!), and you'll want to use one at least 19 kohms for max output. I cut off the panel mounting tabs with small sidecuts, and tap a 9mm X 0.75 hole in the tube of the light. In a Mag tube, this hole is 1/4" in front of the oval switch recess (not the switch hole itself). The pot threads are coated with epoxy, then threaded into the tapped hole from the inside until the housing just protrudes from the tube. The tap is very rare and hard to find; it is acceptable to drill a tight clearance hole and epoxy without threading as well. I don't use the jam nut and washer supplied with the pot, but if the clearance hole method is used, they might come in handy as additional surface for the epoxy to adhere to.
- Connect the pads of one Shark to the pot. Viewing the pot with the shaft at top and the leads facing you, connect the large wiper pad (or top right hole) to the center pot lead, the lower right pad (or lower right hole) to the left lead, and the lower left pad to the right lead. I find it just as easy to solder to the pad as to the hole. The wires do not pass any current, so I use very thin flexible wire, 30 ga or so.
- The other Sharks are wired to the mother by connecting the large wiper pads together. Any number of Sharks can be wired together. Alternatively, all three pads of each Shark can be connected in parallel, or all connected to the pot. From my measurements, it seems that slightly more max current is delivered when connecting all the pads, 960 - 980 mA vs. connecting just the wiper pads, 900 - 920 mA. I'm not certain of the reason, but the difference is arguably negligible.

[Nanomiser]

LZ,

I guess I have allot more to learn about Li-ion cells. As usual, you’ve provided me with some more great information. The implementation of an external pot with multiple-Sharks will be an invaluable reference when it comes time for me to wire up this flame thrower. I will do my best to reflect your instructions in the next revision of my wiring diagram. In the case of the Mammoth, how many watts would the pot see at full throttle?

Thanks,

Mike

[LED Zeppelin]

LZ,

In the case of the Mammoth, how many watts would the pot see at full throttle?

Thanks,

Mike

The pot sees negligible current, you can use teenie wire.

The MD Trinity is a high-power tri-Shark/pot mod wired similarly. On full blast the cells are delivering about 80 watts. I built that before the Blue Sharks' availability and without the copper "C", hence the need for the topside sinks.

[Nanomiser]

That MD Trinity is quite the light! Very nice beam shot comparison. What kind of temperatures did it hit at high output?

I found this Bourns 20K cermet pot on eBay last night and bought them.



-Mike

[Jarl]

What's the absolute max of the blue shark? I'm looking at boosting to 34V/2A (per shark). Would this be too much? The alternative is to use 8 sharks instead of 4 (erg, $$$) or to drop the max power down, though if I do that, I probably won't even hit 9K lumens.

[Greg G]

While we're on the subject of driving multiple CSWOI P7's I need some help diagnosing a light I finished last night.

It's a 3D Mag with 5 P7's driven by 3 Blue Sharks in parrallel. The battery pack, 4 AW "C" cells, was not fully charged and had 15.4 volts resting. In testing, my power supply showed 16.2 volts was needed to run the string at 2.8 amps, so I was in a boost situation. I wired the Sharks up, AA'd them to the mag tube using the Led Zep sinks, and soldered all the LED+ inputs to the positive lead on the first P7.

I turned it on and within seconds I smelled smoke. All the Shark IC's burned up. I can't find anything that is shorted. I checked that carefully while building, and reinspected again afterwards.

The light does come on, but it only has a little over 15 volts at the LED's, meaning I'm in DD?

What did I do wrong? I'm at a loss.

Thanks.

[Nanomiser]

Hi Greg,

Given my current project involves a very similar setup I'm very interested in what you discover. Your hardware configuration sounds reasonable so it is difficult to identify the root cause. Maybe some pictures could help in sorting this out. Unfortunately, I'm still fairly new at this so I don't have any really good suggestions other then to retrace your steps.

Mike

[Greg G]

It's wired like your diagram, except I have 4 C LiIon's (15.4 v), 5 P7's (16.2 vf), and no Remora or pot. Before I glued the Sharks to the sinks I made darn sure no wires protruded from the bottom of the board.

I'm going to salvage the emitters from the light for other projects and scrap this idea. The beam was not very useful.

[Greg G]

Well, I put some magnifying glasses on and one Shark did not have soot around the IC and is still working. I popped them loose from their sinks and desoldered their wires, removing them from the light.

[Nanomiser]

Even though my diagram is flawed for the use of six Li-ions it will certainly work with 4; I still need to revise it. It sounds like you have taken all the necessary precautions during your build; however the magic smoke still escaped. Have you contacted Wayne yet for some Blue Shark Q&A?

[Greg G]

Being a holiday weekend I haven't contacted Wayne.

I have fried several Sharks in the past but I figured out why, the LED+ grounded to the flashlight body (to ground). That did not happen this time. I know it was something I did, I just haven't figured it out yet. I'd like to figure it out so it doesn't happen again. What has me baffled is that one of the three Blue Sharks lived and works fine. They are all wired the same.

[Nanomiser]

Sounds like the sole surviving Blue Shark is the only clue being offered. Follow that lead and see where it takes you. Sorry I don't have any better suggestions.

Mike

[mofiki]

I have bought four of these shark boards in the last year and a half and all of them fried shortly after the light got heated up. This last build utilized the perfect sink 19 with four Cree Q4's, epoxied shark to the sink with a piece of heatsink aquired from the back of a processor chip. LED's wired in series. Batteries consisted of 6 series Duracell AA rechargeables (2000mAh)for a voltage of slightly more than eight volts. On/Off was through the maglite's original switch. Why. Why. Why. If others are having success with this why are mine blowing up?They look every bit as good as some of the best I've seen posted in CPF. I'm no amature when it comes to electronics and assembly. I'm currious about the J1 being jumpered. I know it's to control current in a multiple led string so I've been jumpering J1. This last build I even set the pot lower (didn't run at full power) maybe 750mA supplied to the leds. I'm ready to give up.

[dat2zip]

It hard to tell what is happening. Each case must be looked at carefully.

If the output is shorted that would cause damage to the Shark. You can use an ohmmeter and make sure the LED+ is not shorted to GND.

Other checks are to verify the LED slugs are not contacting the heatsink. Isolation of the LED slugs especially on Seoul LEDs are necessary.

It's best the check the setup with a power supply and set the current limit to 100mA or so. If the LEDs don't light up a little bit there is a problem and troubleshoot this before going to full power.

Another good check would be to power up the LEDs separately before connecting the Shark(s).

Greg,

Can you provide some pictures of your Sharks either here or privately. Any additional information regarding your configuration would be helpful.

Wayne

[Greg G]

Thank you Wayne for helping me figure this out.

The LED+ was not shorting to the bottom of the LED Zep heatsink. When I solder the wires in they all end up flush with the bottom of the converter board. The copper C then spaces the wires further from the sink. I ended up having AA fill up those voids when glueing them down.

As you suggested, I test my LED/heatsink combo with a power supply before hooking it up to the converter board(s). I did have a problem at first with one or more emitters grounding to the sink. The sink is one I turned on my lathe, and is not anodized. Only the first and last emitter lit up, and there was voltage in the sink when checking with a MM. I popped them all off, removed the AA that was still stuck to them, spread a very thin skim coat of AA on the bottom of the emitters and let it cure, then glued them all back down. Everything checked good then, and checks good now.

Again, I believe this is something I did. I have fried Sharks before, but diagnosed why (it was always was the LED+ shorted to ground). this one I can't figure out.

I don't have my camera handy but I'll take some pics next week for you Wayne.

Thanks for the support.