Fenix Outfitters        
Results 1 to 7 of 7

Thread: Canister light attempt. Zero electronics knowledge:/

  1. #1

    Default Canister light attempt. Zero electronics knowledge:/

    Hello. I'm fairly new to diving and keen on wreck diving(dove the Ex-HMAS Brisbane a few months ago, heading to Chuuk in May). I have a couple of basic dive torches. A Big Blue AL1200 main and 450 backup(model numbers refer to lumens) I have a few other lights, an Olight M20, another "tactical" light that DOESN'T need $20 worth of batteries(can't remember it's name and it's locked away), a 3D Led maglite, various head-torches...and, the reason for this thread...a busted P7 Ledlenser(pulled it apart and the plastic switch cylinder was broken allowing the switch contact to seperate from the light head contact). So I was searching for ways I can use the torch coupled with current diving needs and I came across copious amounts of dive canister builds....so I followed down the rabbit hole...and now I'm stuck in limbo between having found enough information that I want to build a powerful light for a lot less than I can buy one for, and having found enough information to realise just how much more I need to know, which leads me to asking you guys for help. I've been trawling through the forums picking up bits and pieces of info, but I cant keep it all together(and I'm easily distracted. I spent 2hrs last night looking into IR emitters...just because) I have about 40 tabs open, and can't remember what half of them are for.

    My understanding of electronics is nearly non-existent. I have a basic understanding of the relationship between volts/amps/watts from auto electrical DIY, but nowhere near enough to work it all out myself. I understand a lot of the questions are going to be rehashing old stuff for you guys, and I'm happy to be told to go and search for it(some direction would be appreciated, links would be awesome) I'll bold the questions to hopefully make it easier to find the answers(but please watch out for the questions that I don't know to ask

    So, where I'm at.(cant work out how to attach images, so more typing) I have cut the head off the torch (110mm long, worried now it's a bit short but was limited due to the shape of the torch), removed the old LED and machined a new end cap/lens holder.(the ledlenser one had breather holes behind the lens. Not ideal for a dive light) I have allowed room for extra o-rings but might end up JB welding it together anyway. There is an aluminium heat-sink(41mm OD) which threads into the torch body 15mm. Will I need some sort of thermal compound on the threads to assist in heat transfer and is this enough contact area? Does thermal paste set hard or would I still be able to unscrew the parts?

    I've been searching for 35-40mm MCPCBs, preferably with LEDs already mounted. Initially I found a 38mm, 8up XHP50.2 on Cutter Elec, so of course figured that was definitely the go(20k Lumens, 140w...good beginner project)After much soul searching, I put that idea aside for a later date. More realistically there are a couple of options in the XHP35 range that I now have my eye on. There is a 40mm 7up which is probably still a bit much, butit says they can be run in parallel, series or independently addressed. I'm assuming this means i can run them separately, which leads me to a couple more questions...If I run 2 SEPARATE series, one with 3leds and one with 4(by seperate, I mean one on, one off, not 2 parallel series)? If so, can they be run off the same CC step up driver? (different forward voltage across the circuits). If this can actually be done, where/how can the switching happen? Can I have a toggle switch after the driver? Is it possible to get optics that have different spread angles? i.e. 3 at ~10deg and 4 at ~90deg. The idea being that I use switching between circuits to change focus, giving the light more functionality, and also not trying to run 7x13w lights all at once(I get that I CAN run them at a lower power, but dont see much point having powerful lights without at least having the option to run them at full power) If this isn't a valid option, there is a 4up and 3up on 32mm(although it says they are both 36v, which is confusing. I'm assuming that's total Vf and the 36 for the 4up is a typo) which I think I can safely fall back on.

    So the other questions I can think of right now:

    Do boost drivers work more efficiently with closer Vin/Vout relationship, like buck drivers? I don't want my input to be higher than 12v and even 3 XHP35s are going to be 36v.

    Is it feasible to run 7 12v LEDs from a 12v or smaller power source? Alternatively, the XHP50.2s ARE available in 3v, that would make the 8up 24v at 6amps...batteries could be an issue though. (or was it 2 series of 4? Cant remember and too late to go looking)

    Do most LED drivers do all the low battery/high temp warning stuff? Where's a good place to find such a creature? Is this a good start?
    http://www.taskled.com/hbflex.shtml

    Is it doable to mount the driver in the canister with the battery? I would then have to allow for voltage drop across the leads when working out the driver output, which increases the difficulty, but not sure I have enough room in the head unit for too much stuff.

    Batteries...I assumed I could just get whatever batteries wanted, mounted however I wanted as long as the voltage and current was right when it got to the LEDs, but I've been reading stuff that says otherwise. How do I work out what sort of batteries I get. I was looking at LiFePo(lighter, safer?) I haven't managed to do a great deal of research into batteries.

    PWM dimming. Is it too hard for me to worry about, or is there "plug and play" options that I can use?

    A bit of daydreaming...has anybody looked into wireless controllers? I'm thinking like a wrist mounted device with selectable power level
    (touchscreen, slide the powerbar?), selectable mode (different colours, different flash sequences, type your msg and it sends it in Morse code displays battery levels, run time remaining at (x) level.

    Anyway, it's late here and I can't think of any more serious questions. Thanks in advance for anybody who can help in any way and I'll work out how to upload the pictures of my small amount of progress at a late date.

    Matt

  2. #2
    Flashaholic* DIWdiver's Avatar
    Join Date
    Jan 2010
    Location
    Connecticut, USA
    Posts
    2,464

    Default Re: Canister light attempt. Zero electronics knowledge:/

    Hi, and welcome to the forum!

    I'll try to answer some of your questions.

    To post pictures, you have to host them on another site, then click the 'insert image' button and paste a link to the picture. Forum rules state they must be limited to 800x800 pixels. If you are using dropbox, you have to edit the URL. Let me know if you need that info.

    I think most dive lights do not attempt to dissipate heat through threads. Thermal compound would be of little help. It is only 'good' at filling very tiny gaps. It's thermal conductivity is typically 5000-20000 times less than aluminum. Diamond based ones are probably a bit better than that. You pretty much want a direct metal path from the LEDs to the water. If you must thread the parts together, there should be a flat mating surface where the parts bottom out. This will conduct much better than the threads.

    There are many types of thermal interface materials. What's typically called 'compound' is a mixture of powder and oil or grease, to make a thick, greasy material, which does not harden. I have a tube of Fujik, which cures to a solid material that you can dig into with your fingernails. There are also epoxies that cure quite hard and strong. There are also sheets of all sorts of rubbery materials, some of which are adhesive.

    It is possible to switch strings as you have suggested. However, some drivers don't like the open circuit that would happen briefly during switching, and some should not be 'hot' connected to the LEDs. Again, possible, but be sure you know what you are doing. Another possibility is to run all 7 in series, then short 3 or 4 at a time. That way you never have an open driver.

    Yes, boost drivers tend to be more efficient with lower boost ratios. But a lot depends on the design and implementation of the driver too.

    According to the Cree website, the XHP50.2 has two 6-volt sections, which can be wired parallel or series. No 3V option. It's possible to run 7 12V LEDs in series from a 12V source, but there are likely better choices. That's a pretty high boost ratio.

    No, most drivers don't do warnings and protections. The ones that do are generally the high-end ones.

    Taskled makes great products, and if one of them fits your needs don't hesitate to buy them. Somewhere on the forum there's a list of a huge selection of drivers, and it is searchable by parameters. I haven't seen it in some years though.

    Most drivers are okay with the LEDs at the end of a few feet of cable, but I think I read a thread about one that was definitely not. Read the datasheet, ask, or experiment before counting on it. It isn't the voltage drop. They can handle that just fine (LEDs have resistance in them too!). I suspect it was the inductance of the cable that was a problem.

    LFR batteries are indeed lighter than most other LiIon cells, and are generally considered safer, but at a significant cost. They are both lower voltage and lower capacity than other LiIon chemistries, and should be charged in chargers designed for them.

    PWM dimming is done frequently. Any '7135 based driver with brightness control is done that way, as are many others. I'm not aware of a plug-n-play option, but you could certainly find a small driver that you like, unsolder the '7135's (or even not) and hack it into your system. Some of the Taskled drivers accept PWM inputs. One thing to keep in mind is that it can wreak havoc with digital photography and video.

  3. #3

    Default Re: Canister light attempt. Zero electronics knowledge:/

    Thanks DIWdiver, a bit of info to go through. I THINK I've worked out how to resize and add photos, was just waiting for the post to appear so that I could try...aaand fail. Using google photos. I use the share link that it provides, but the image doesnt show up here(or does it not show up until I hit post?)

    Cutters' are offering the XHP50.2 in 3v ( https://www.cutter.com.au/product/xh...000-0a00j4051/ ) Maybe the 3led 32mm in series is an option, then I can use a step down from 12v-9v? Or up from 6v-9v? Could the 6amps be an issue with finding a driver?

    Now I'm wondering why I'm using the old torch at all. Especially with the heat sink issue. Probably going to be easier/better to get aluminium rod that fits the ID of the head(~50mm) and drill/machine it out. Then the whole thing becomes a heat sink. Give me more room for bits and pieces, too.

    I had thought the driver might not like the momentary "off" during switching. Can you elaborate on the shorting suggestion? Would I still be running the lights in seperate series? Does the "short" still draw power? Sorry, this really is all foreign to me.

    One more thought. Would this thread be better off in the Light Modding forum?

  4. #4
    Flashaholic* DIWdiver's Avatar
    Join Date
    Jan 2010
    Location
    Connecticut, USA
    Posts
    2,464

    Default Re: Canister light attempt. Zero electronics knowledge:/

    Quote Originally Posted by YZed View Post
    Thanks DIWdiver, a bit of info to go through. I THINK I've worked out how to resize and add photos, was just waiting for the post to appear so that I could try...aaand fail. Using google photos. I use the share link that it provides, but the image doesnt show up here(or does it not show up until I hit post?)
    Haven't done it in a while, but IIRC it shows up if you post then go back in and read the post, or if you "Go Advanced" and click "Preview Post".

    Quote Originally Posted by YZed View Post
    Cutters' are offering the XHP50.2 in 3v ( https://www.cutter.com.au/product/xh...000-0a00j4051/ ) Maybe the 3led 32mm in series is an option, then I can use a step down from 12v-9v? Or up from 6v-9v? Could the 6amps be an issue with finding a driver?
    Huh. Cree doesn't mention them. Maybe they only made a few and Cutter got most (or all) of them. That would be a great option for some lights. The Taskled h6Flex can do 6.6A at 9V with 12V input, but if 12V comes from three LiIons in series, it's actually 9-12V and not enough to maintain full brightness to the end of discharge. You might consider that a good thing (warns you of impending shutoff).

    Quote Originally Posted by YZed View Post
    Now I'm wondering why I'm using the old torch at all. Especially with the heat sink issue. Probably going to be easier/better to get aluminium rod that fits the ID of the head(~50mm) and drill/machine it out. Then the whole thing becomes a heat sink. Give me more room for bits and pieces, too.
    A good choice if you can pull it off.

    Quote Originally Posted by YZed View Post
    I had thought the driver might not like the momentary "off" during switching. Can you elaborate on the shorting suggestion? Would I still be running the lights in seperate series? Does the "short" still draw power? Sorry, this really is all foreign to me.
    Say you run a string of 7 LEDs. If you short across the lower three, the string now looks like (the upper) 4 LEDs. If instead you short the upper 4, the string now looks like (the lower) 3. For whatever period neither section is shorted, the string looks like 7. There are 3-position toggle switches that could provide all three options. Some battery/driver combinations will support this behavior, some won't, depending on driver type and battery voltage.

    Quote Originally Posted by YZed View Post
    One more thought. Would this thread be better off in the Light Modding forum?
    If you mean "Homemade and Modified Flashlights Discussion", then yes, you would get a lot more eyes there. Just post here asking a mod to move it if you want to.

  5. #5

    Default Re: Canister light attempt. Zero electronics knowledge:/

    If a Mod didn't mind moving this to the "Homemade and Modified Flashlights Discussion", it wouldn't hurt. I'm liking the 3*XHP50.2s with the H6Flex, but have no doubt I'll be asking more questions as I go.

    I think my confusion with the shorting is due to lack of knowing how the LEDs are wired. I assumed if you shorted anywhere in the circuit, they'd all go out. I wonder if 2 drivers would be a better/easier way? Have the power going through a 2 output relay to switch between drivers...but that's not a consideration for the 3LED light. It also requires the finding of a multi-optics that has 2 different beam angles.

    Thanks for the help.

  6. #6
    Moderator
    nbp's Avatar
    Join Date
    Dec 2007
    Location
    Wisconsin
    Posts
    9,400

    Default Re: Canister light attempt. Zero electronics knowledge:/

    Thread moved.

  7. #7

    Default Re: Canister light attempt. Zero electronics knowledge:/

    For consideration, I have a cheap generic Chinese headlamp with 3 LEDs, one central thrower and two side wide angle lenses similar to;
    https://www.aliexpress.com/item/T6-X...707181660.html
    The driver board had two separate + power outputs for two different LED circuits and the ability to power them independently in order to select between:
    1) just the thrower,
    2) just the two wide-angles which are in parallel with each other, and
    3) both circuits 1 and 2 together.

    This is a fairly cheaply made headlamp, runs just a 3.6v output battery pack of 2 parallel cells so it is nothing high-power or directly usable for your project. I only mention it as an example of a switchable dual-circuit output board that perhaps might be possible to copy, cannibalize or heavily modify to use its programming in order to control two separate high power circuits?

    As for the shorting question above, imagine a string of 7 series LEDs connected to ground at the end to complete the circuit. Add to that a three-way switch that is connected at the middle of the string. In one position, the switch just connects the front and back half of the strings together so they are all in series. Switching to add a short to ground in the middle of the string would provide a shorter, lower resistance path to ground so effectively bypass the last 1/2 of the string and only keep the front 1/2 of the string powered. Lastly, switching to connect the (ungrounded) middle of the string to the front of the string would just bypass the front and power only the rear 1/2.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •