Seeking advice about design/engineering issues when making an LED flashlight

[bpierce]

Hi -- I'm a newbie who is joining these forums to learn more about making LED flashlights. I created a prototype of a new flashlight product, but the characteristics of it may create some interesting design challenges as I look to create a real/final design spec ...and I'm a software engineer, not an electrical engineer. The product is an LED flashlight that docks in a table lamp so that it is always charged (I call it a "Flashlamp" ...not to be confused with a photographic flash-lamp).

A few of the potential design challenges:

(1) Since it's a lamp, it could sit on somebody's end table for many years. Are there particular battery technologies that are recommended for this type of application ...especially considering it will always be plugged in and charging? Is an 18650 Li-ion battery fine in that context?

(2) The flashlight must be bright since it may be used when the power is out ...and because I think that will it more marketable... but I also need to keep the cost down to make the product viable. Should I be looking at alternatives to the most powerful CREE LEDs?

(3) The body of the flashlight has a non-standard shape because it needs to dock into the base of the lamp. Will I have any concerns about being able to dissipate heat with such a body design? The internal side of the flashlight (that is hidden when the flashlight is docked) may be plastic instead of metal - would that create any new issues with regard to dissipating heat?

Maybe these aren't really concerns at all. I am working with an industrial design firm outside of Chicago, but of course they will tell me it's all doable since I'm paying them. Looking for 2nd opinions here, and also suggestions for resources that can best help me climb this learning curve quickly. Pics of the product are below, and a video of the product in action is available at my Kickstarter here: http://kck.st/1uZGjux.

Thanks in advance for any advice!

Brian

 

[Steve K]

that's interesting.... you've started a kickstarter project before you had anything more than a general idea of what you want the product to be.

As an experienced designer, I would start with a specification... how bright should it be, how long should it run, how big should it be, how much should it cost, what will the production volume be, etc. Then start working on the concept for the subassemblies.. what LEDs will meet the spec, what electronics are needed, what batteries will meet the spec, what does the housing need to be made of, etc.

Once you've got a spec and an idea of what the subassemblies should be, you start prototyping and seeing if your first ideas are any good. You'll find out that there were a lot of things you hadn't thought about or were ignorant of. This is typical. You'll go through a few versions in the process of getting the details worked out.

Of course, since this is a consumer product, there are safety regulation as well as FCC regulations, and some EU regulations (such as elimination of hazardous substances) that you'll need to comply with. You're familiar with all of these, right? You don't want to have the product kept off of the market because it doesn't comply with the FCC part 15 regulations for electromagnetic emissions.

Out of curiosity, why does a person want to get into the business of manufacturing a product that they don't know anything about?

 

[bpierce]

that's interesting.... you've started a kickstarter project before you had anything more than a general idea of what you want the product to be.

As an experienced designer, I would start with a specification... how bright should it be, how long should it run, how big should it be, how much should it cost, what will the production volume be, etc. Then start working on the concept for the subassemblies.. what LEDs will meet the spec, what electronics are needed, what batteries will meet the spec, what does the housing need to be made of, etc.

Once you've got a spec and an idea of what the subassemblies should be, you start prototyping and seeing if your first ideas are any good. You'll find out that there were a lot of things you hadn't thought about or were ignorant of. This is typical. You'll go through a few versions in the process of getting the details worked out.

Of course, since this is a consumer product, there are safety regulation as well as FCC regulations, and some EU regulations (such as elimination of hazardous substances) that you'll need to comply with. You're familiar with all of these, right? You don't want to have the product kept off of the market because it doesn't comply with the FCC part 15 regulations for electromagnetic emissions.

Out of curiosity, why does a person want to get into the business of manufacturing a product that they don't know anything about?

Thanks for the feedback! You are correct that I am not an expert in all of those areas, nor am I an expert in patent law (another important aspect of this project). I need to rely on the experts I engage -- the engineers and lawyers. But I believe in a "trust but verify" model, whereby I try to learn as much as I reasonably can about the subject matter; that way I can make informed decisions given the input of experts. The engineers did a great job on the prototype, and these forums seem like a great place to get objective feedback as we move toward a final design spec.

I am also a fan of the lean approach to starting a business. The beauty of Kickstarter is you can get market feedback on a product concept before investing too aggressively in design and prototyping. I could choose to first invest heavily in acquiring expertise in flashlight design, but that effort may be largely wasted if there is no demand for the product I envision. Besides, the feedback of my potential customers (especially very knowledgeable ones) is much more valuable than any specific ideas I may have.

We will have language in our contract with our manufacturer about satisfying any regulatory requirements in our target markets, and that language will be reviewed by our attorneys ...and I'll do my own research on that during that process, so that I feel good about the whole thing. One good thing about this particular project is that all of the electrical components will be off-the-shelf, and what we are providing is ...in effect... a novel form factor for the docking station. For that reason, we do not anticipate we'll have too much difficulty in selling Flashlamp legally.

Please let me know any other thoughts you may have. All feedback is helpful feedback!

Brian

 

[DIWdiver]

1. How long do you expect the battery to last? 5 years, 10 years, more? NiMH and Li-Ion are probably reasonable choices, depending on your requirements. For longest lifetime, a Li-Ion charger should have a lower termination voltage, and allow the voltage to drop some due to self-discharge before topping off. NiMH should probably do something like that too, but the charging is more complicated. The charging circuit should be able to track this even over power failures and if the light is removed and used.

2. Everyone's idea of 'bright' is different. My 400 lumen dive light seems pretty bright to me if all the lights are off, and I would not be at all surprised if it's actually only doing 300-350 emitter lumens, 300 or less OTF. Cree's highest power LEDs are definitely overkill. Remember this is going to be used in close quarters (in the house) when it's pretty dark. I would definitely consider any and all LEDs that can achieve your performance levels. Given that the total lifetime usage may be in the tens of hours or even less, lumen maintenance is not much of an issue.

3. 400 lumens can be had today for around 3W. About 30-35% of that should be light out the front, the rest is dissipated as heat in the LED, reflector, lens, etc. That gives at worst 2.1W of heat, most of which is in the LED. This is a pretty small number, and while it will take some attention to ensure it's handled correctly, it should not be at all hard to do. A metal body is certainly helpful, but at that power level not necessary.

I would say there's no reason to think (from what you've told us; I haven't perused your web site carefully or watched the video) that your engineering firm is blowing smoke. IMHO it's very doable.

Since you opened the door to other advice, I think it should automatically light up when the power goes out, so you can find it. Nothing major, just a small white LED that can illuminate the end table where the lamp is resting.

The biggest problem users may have is that they use it for something then don't place it back in the charger, then next time it's either dead or nowhere to be found. Perhaps when the power is on and the flash is not in the base, it should chirp (like a smoke alarm low battery indicator) to remind them to replace the flash in the lamp.

I think it's a great idea and I have little doubt that at $60-80 you could sell a pile of them. But you probably wouldn't sell me any. In my opinion, an emergency backup light should be replicated in numerous places throughout the house (especially if there are more than 2 people living there) and should have primary cells. They should be tested when you test your smoke alarms, and have batteries replaced at appropriate intervals regardless of whether they are dead (use up replaced ones in other devices). Not to mention, you've made that one so inconspicuous, I'd probably forget it's there (especially if it doesn't have a light that comes on automatically during power failure).

 

[Steve K]

in reply to the question about charging 18650 batteries indefinitely... this is the battery type used in laptops, and a lot of them get charged quite a bit. There are charger circuits that will maintain them at a charged level without too much trouble. I'd recommend looking at the websites of the battery manufacturer. The manufacturer will tell you how to properly charge it.

regarding LED choice: You'll want to generate a spec for the LED (how many lumens, beam pattern and suitable optics, etc.), and review what LEDs are available that meet the spec. Suppliers like Digi-key and Mouser are handy as a way to quickly review what is on the market. Your contract manufacturer may want to work with a distributor like Avnet to arrange for better prices, screening, etc. Distributors can be your best friend, especially when they are the ones who can tell you in advance that the part that is critical to your design is going obsolete in 6 months and you need to figure out how you will adapt your design to handle this issue.

Regarding thermal design: a metal surface that is exposed to air ought to provide decent heatsinking for most things. The power dissipation during battery charging should be limited. This is what the prototype will tell you, though.

 

[Steve K]

How does it work with a contract manufacturer? I'm assuming that you provide the design documentation (prints, bill of materials, etc), and they figure out what the processes need to be to build and test it. Do they really do all of the work to ensure that it complies with the applicable regulations? That's more than the manufacturers I've worked with.

Does the contract manufacturer also work with all of the suppliers? There will be someplace needed to make the unique bits for the housing.. the injection molded plastic parts, the aluminum parts that need to be cast/extruded/machined, the printed circuit boards, etc. Dealing with suppliers can be a bit of work. Presumably these will be suppliers that this manufacturer has dealt with in the past.

Does the manufacturer do any of the design work? I know that I can design a circuit, lay out a circuit board, etc., but I also know that the mechanical packaging details of electronics takes a lot of specialized knowledge too. I've seen a lot of weird and strange things happen as the result of poor selection of plastic materials.. nylon can absorb moisture and swell up, some plastics are sensitive to threadlock material and can crack after a while, dimensional stability is a concern..... lots of little details that require a lot of validation testing to detect, and may require the recall of the parts if a problem shows up. Another concern is tolerances.. mechanical tolerances have to be chosen so that everything will fit together in all instances. You don't want to have to scrap parts because stuff doesn't quite fit. Prototypes will only show if this batch of parts fit together. Only a proper understanding of tolerances will ensure that all batches of parts fit together.

Personally, I love designing stuff. I don't love having my designs go into production because it means that I have to support the product. This means that I have to figure out why half the parts aren't passing the end-of-line test, or figure out what to do when that unique I.C. is no longer in production and no one makes a replacement, or what to do when you start getting a lot of failed parts back from the customers. I've had some fun research jobs where I was challenged and learning, and eventually it all got thrown away when funding disappeared. No production problems, no reliability issues.. it was great!
Why do you want to get involved in these sorts of headaches?

 

[Norm]

:welcome: To CPFs consultant services, we will be passing the hat at the end of the session.

 

[bpierce]

...deleting. meant to respond with quotes...

 

[bpierce]

1. How long do you expect the battery to last? 5 years, 10 years, more? NiMH and Li-Ion are probably reasonable choices, depending on your requirements. For longest lifetime, a Li-Ion charger should have a lower termination voltage, and allow the voltage to drop some due to self-discharge before topping off. NiMH should probably do something like that too, but the charging is more complicated. The charging circuit should be able to track this even over power failures and if the light is removed and used.

2. Everyone's idea of 'bright' is different. My 400 lumen dive light seems pretty bright to me if all the lights are off, and I would not be at all surprised if it's actually only doing 300-350 emitter lumens, 300 or less OTF. Cree's highest power LEDs are definitely overkill. Remember this is going to be used in close quarters (in the house) when it's pretty dark. I would definitely consider any and all LEDs that can achieve your performance levels. Given that the total lifetime usage may be in the tens of hours or even less, lumen maintenance is not much of an issue.

3. 400 lumens can be had today for around 3W. About 30-35% of that should be light out the front, the rest is dissipated as heat in the LED, reflector, lens, etc. That gives at worst 2.1W of heat, most of which is in the LED. This is a pretty small number, and while it will take some attention to ensure it's handled correctly, it should not be at all hard to do. A metal body is certainly helpful, but at that power level not necessary.

I would say there's no reason to think (from what you've told us; I haven't perused your web site carefully or watched the video) that your engineering firm is blowing smoke. IMHO it's very doable.

Since you opened the door to other advice, I think it should automatically light up when the power goes out, so you can find it. Nothing major, just a small white LED that can illuminate the end table where the lamp is resting.

The biggest problem users may have is that they use it for something then don't place it back in the charger, then next time it's either dead or nowhere to be found. Perhaps when the power is on and the flash is not in the base, it should chirp (like a smoke alarm low battery indicator) to remind them to replace the flash in the lamp.

I think it's a great idea and I have little doubt that at $60-80 you could sell a pile of them. But you probably wouldn't sell me any. In my opinion, an emergency backup light should be replicated in numerous places throughout the house (especially if there are more than 2 people living there) and should have primary cells. They should be tested when you test your smoke alarms, and have batteries replaced at appropriate intervals regardless of whether they are dead (use up replaced ones in other devices). Not to mention, you've made that one so inconspicuous, I'd probably forget it's there (especially if it doesn't have a light that comes on automatically during power failure).

Thanks for the great thoughts! The battery lifetime is a tough one. Realistically, a user will probably almost forget it's there (as you mentioned) and not replace the battery for years. I am going to do some more research on the points you raise, and also discuss them with my electrical engineer. My main concern is minimizing the chance of corrosion that destroys the flashlight before they do.

I was also thinking the highest power CREEs may be overkill for this application, so nice to have that validated. Also reassuring to get a second opinion that the heat will likely be manageable.

On the emergency light, that's something I've heard a lot. I am going to try and figure out how to best work something like that in, as I agree it adds a lot of value ...especially since the safety aspect is part of how I'm selling it.

And on users not putting it back in the charger... I did consider having some mechanism whereby maybe the user could push a button on the base and the flashlight makes a sound (to help them find it). But making the flashlight, itself, just chirp after a while is better since it really prompts them to replace it. I'm hopeful that even without that feature, users will be more inclined to replace the flashlight for 2 reasons: 1) there's a natural place for it, and 2) the lamp looks silly when it's not replaced. It's going to be a balancing act between providing these great features and keeping the price at a consumer-friendly level.

Over time I hope I can implement features that may make you a customer! Lamps are interesting since they are relatively bulky, conveniently located throughout a house and always plugged in. I could use ethernet over power to let the lamps communicate, and if one is "smarter" than the others then perhaps the battery status of all could be available via the Web (with alerts on a mobile app, so you'll know when to change the batteries). Of course that won't be there day one .

Thanks again for the great thoughts, and please don't hesitate to post any other insights you may have.

Brian

 

[bpierce]

in reply to the question about charging 18650 batteries indefinitely... this is the battery type used in laptops, and a lot of them get charged quite a bit. There are charger circuits that will maintain them at a charged level without too much trouble. I'd recommend looking at the websites of the battery manufacturer. The manufacturer will tell you how to properly charge it.

regarding LED choice: You'll want to generate a spec for the LED (how many lumens, beam pattern and suitable optics, etc.), and review what LEDs are available that meet the spec. Suppliers like Digi-key and Mouser are handy as a way to quickly review what is on the market. Your contract manufacturer may want to work with a distributor like Avnet to arrange for better prices, screening, etc. Distributors can be your best friend, especially when they are the ones who can tell you in advance that the part that is critical to your design is going obsolete in 6 months and you need to figure out how you will adapt your design to handle this issue.

Regarding thermal design: a metal surface that is exposed to air ought to provide decent heatsinking for most things. The power dissipation during battery charging should be limited. This is what the prototype will tell you, though.

Thanks again Steve. I'll definitely checkout those resources. I'm much more likely to keep the price down if I have the relationship with the component suppliers and distributors directly, rather than letting a manufacturer handle that for me.

Part of the reason I raised particular concern about the power dissipation is because, in the prototype, some of the internal parts were 3D printed. The team that built it for me warned that the material wouldn't stand up to the heat the way a prod unit would, so I should be careful to leave it on too long. I think they are being straightforward, but seeking reassurance that it probably is only a potential issue with the printable plastic. I feel like I'm already getting some of that reassurance from these responses.

 

[bpierce]

How does it work with a contract manufacturer? I'm assuming that you provide the design documentation (prints, bill of materials, etc), and they figure out what the processes need to be to build and test it. Do they really do all of the work to ensure that it complies with the applicable regulations? That's more than the manufacturers I've worked with.

Does the contract manufacturer also work with all of the suppliers? There will be someplace needed to make the unique bits for the housing.. the injection molded plastic parts, the aluminum parts that need to be cast/extruded/machined, the printed circuit boards, etc. Dealing with suppliers can be a bit of work. Presumably these will be suppliers that this manufacturer has dealt with in the past.

Does the manufacturer do any of the design work? I know that I can design a circuit, lay out a circuit board, etc., but I also know that the mechanical packaging details of electronics takes a lot of specialized knowledge too. I've seen a lot of weird and strange things happen as the result of poor selection of plastic materials.. nylon can absorb moisture and swell up, some plastics are sensitive to threadlock material and can crack after a while, dimensional stability is a concern..... lots of little details that require a lot of validation testing to detect, and may require the recall of the parts if a problem shows up. Another concern is tolerances.. mechanical tolerances have to be chosen so that everything will fit together in all instances. You don't want to have to scrap parts because stuff doesn't quite fit. Prototypes will only show if this batch of parts fit together. Only a proper understanding of tolerances will ensure that all batches of parts fit together.

Personally, I love designing stuff. I don't love having my designs go into production because it means that I have to support the product. This means that I have to figure out why half the parts aren't passing the end-of-line test, or figure out what to do when that unique I.C. is no longer in production and no one makes a replacement, or what to do when you start getting a lot of failed parts back from the customers. I've had some fun research jobs where I was challenged and learning, and eventually it all got thrown away when funding disappeared. No production problems, no reliability issues.. it was great!
Why do you want to get involved in these sorts of headaches?

I'll be evaluating manufacturers more over the next couple of months, and to be honest I don't know exactly what to expect until I speak with them. I do know that some have their own brands, so may presumably know how to ensure regulatory compliance. Some have in-house design teams, so again should be able to design for regulatory compliance. I've debated whether to try and go with a one-stop-shop like that for the final design spec and manufacturing, but I'm concerned about the conflict of interest (I sort of want my design guys to care only about a great design, not about maximizing manufacturing profits). In any event, it's a question both I and my counsel will ask.

I've heard lots of horror stories around the issues you describe. The engineers I'm working with have also been warning me about some of these concerns. For example, I originally said I wanted the flashlight to fit so tightly you couldn't notice a seam ...and they warned that trying to keep the tolerances that tight would be a nightmare, and the consistency of the quality would be terrible (I believed them). I don't pretend I'll be able to avoid these issues altogether, so I'm going to try and keep the investment as small as possible as the issues are ironed out.

As a software engineer, I can empathize with you. I love designing software, but making systems work for real-world use cases always gets messy ...and supporting the systems after they are deployed can be a nightmare, even if everything was done "right" in terms of testing/monitoring/etc. But ultimately that delivery to the end user ...and the risk that goes along with it... needs to happen if we want to improve people's lives, so I almost feel like I'd be remiss not to try. Not to imply my Flashlamp will necessarily do a lot to improve lives, but it's a small step and could maybe grow to something more meaningful over time.

Again, I appreciate all of the insights. Please keep any other thoughts coming!

 

[bpierce]

:welcome: To CPFs consultant services, we will be passing the hat at the end of the session.

I'd be glad to throw in!

 

[The_Driver]

Please take a look at this flashlight from LedLenser called the M17R. It uses a modern LED (Cree XM-L2), has regulation circuity, uses LiFePO4 batteries and has a charging cradle. The intended users are law enforcement agencies (and firemen etc.) that will uses it regularily for many, many years and will always have it sitting in the charging base ready-to-go.

LiFePO4 is a special type of rechargeable lithium battery. They can be recharged thousands of times (compared to ~500 cycles of normal Li-Ion batteries), but have a lower energy density compared to the standard Li-Ion batteries. For your application they make a lot of sense. They are also safer (they can't cause fires and explode).

When the power goes out the M17R automatically turns on so that the user can find it in the dark. This is a very useful functionality.

 

[bpierce]

Please take a look at this flashlight from LedLenser called the M17R. It uses a modern LED (Cree XM-L2), has regulation circuity, uses LiFePO4 batteries and has a charging cradle. The intended users are law enforcement agencies (and firemen etc.) that will uses it regularily for many, many years and will always have it sitting in the charging base ready-to-go.

LiFePO4 is a special type of rechargeable lithium battery. They can be recharged thousands of times (compared to ~500 cycles of normal Li-Ion batteries), but have a lower energy density compared to the standard Li-Ion batteries. For your application they make a lot of sense. They are also safer (they can't cause fires and explode).

When the power goes out the M17R automatically turns on so that the user can find it in the dark. This is a very useful functionality.

Wow - the LiFePO4 battery seems like a no-brainer. Thanks!

I'll have to work in an emergency light that turns on when the power is cut ...possibly doubling as a night light for those who want it on when dark, power or no power. That feature has just come up too many times.

Brian

 

[Steve K]

regarding batteries... the rechargable battery will eventually die. Is there a mechanism for signalling that the battery is dead, and is the battery replaceable by the consumer? If it is replaceable, is it a common type, and is there a way to detect if the wrong battery type is installed? i.e. it would be nice if it was a common, cheap battery, such as a couple of NiMH AA's. The charging system would need to detect if the user put alkaline cells in the light instead of NiMH.

Properly used, I know that nicads can last over a decade in a system (based on my involvement with satellite power systems). Cordless phones seem to do well with NiMH AA's too. Simply not trying to get to 100% capacity will be kinder to the battery. 80% or 90% is better for the battery. You might consider some sort of periodic partial charge/discharge cycle too.

You could make an argument for using supercaps too. There are some impressively large capacities available nowadays. The cost per amp-hour is rather high, but at least you wouldn't have to make it replaceable. That will save a little bit of cost in the housing.

 

[bpierce]

regarding batteries... the rechargable battery will eventually die. Is there a mechanism for signalling that the battery is dead, and is the battery replaceable by the consumer? If it is replaceable, is it a common type, and is there a way to detect if the wrong battery type is installed? i.e. it would be nice if it was a common, cheap battery, such as a couple of NiMH AA's. The charging system would need to detect if the user put alkaline cells in the light instead of NiMH.

Properly used, I know that nicads can last over a decade in a system (based on my involvement with satellite power systems). Cordless phones seem to do well with NiMH AA's too. Simply not trying to get to 100% capacity will be kinder to the battery. 80% or 90% is better for the battery. You might consider some sort of periodic partial charge/discharge cycle too.

You could make an argument for using supercaps too. There are some impressively large capacities available nowadays. The cost per amp-hour is rather high, but at least you wouldn't have to make it replaceable. That will save a little bit of cost in the housing.

regarding batteries... the rechargable battery will eventually die. Is there a mechanism for signalling that the battery is dead, and is the battery replaceable by the consumer? If it is replaceable, is it a common type, and is there a way to detect if the wrong battery type is installed? i.e. it would be nice if it was a common, cheap battery, such as a couple of NiMH AA's. The charging system would need to detect if the user put alkaline cells in the light instead of NiMH.

Properly used, I know that nicads can last over a decade in a system (based on my involvement with satellite power systems). Cordless phones seem to do well with NiMH AA's too. Simply not trying to get to 100% capacity will be kinder to the battery. 80% or 90% is better for the battery. You might consider some sort of periodic partial charge/discharge cycle too.

You could make an argument for using supercaps too. There are some impressively large capacities available nowadays. The cost per amp-hour is rather high, but at least you wouldn't have to make it replaceable. That will save a little bit of cost in the housing.

Using a common type of battery that can be easily replaced by the consumer is a core requirement, so that will be there. Having the charging system detect a wrong battery ...or alert when the ability to hold a charge diminishes... is a great idea and something I'll investigate. Ditto on the circuitry to prevent it from charging to 100% and supporting the charge/discharge cycling. Are there flashlights on the market now that do this?

I hadn't even thought of supercaps, but would be interested to see what that option may cost. At some point I may be looking at distinct flavors of the product: a high end one that comes at a higher price point, and a more "consumer grade" one that comes at a Target-friendly price point (which may not realistically have a lot of the nicer features). This continues to be great info, so at the risk of wearing "thanks" thin ...thanks.

 

[The_Driver]

Properly used, I know that nicads can last over a decade in a system (based on my involvement with satellite power systems). Cordless phones seem to do well with NiMH AA's too. Simply not trying to get to 100% capacity will be kinder to the battery. 80% or 90% is better for the battery. You might consider some sort of periodic partial charge/discharge cycle too.

Nicads are really old, outdated technology. To my knowledge thay have no benefits over LiFePO4 batteries, but a lot of downsides (poisonous Cadmium inside, high self discharge rate etc.). The idea of not using all of the capacity is a very good one though. Electric cars use this method to raise the cycle count of the batteries to an acceptable level. One method is only charging the batteries to 90% and discharging to only 10%. The problem is that one has less and less capacity to actually work with.

The problem with supercaps is that they are huuuuuge. Their energy density is very low compared to rechargeable batteries. Here you can find a review of a supercap containing a similar amount of energy (Wh) as a NimH AA battery including some size comparions. You can basically make a light as bright as a Zebralight h52 with similar runtimes that is larger than a 2D Maglite (60mm diameter!!!!).

 

[Steve K]

Nicads are really old, outdated technology. To my knowledge thay have no benefits over LiFePO4 batteries, but a lot of downsides (poisonous Cadmium inside, high self discharge rate etc.). The idea of not using all of the capacity is a very good one though. Electric cars use this method to raise the cycle count of the batteries to an acceptable level. One method is only charging the batteries to 90% and discharging to only 10%. The problem is that one has less and less capacity to actually work with.

The problem with supercaps is that they are huuuuuge. Their energy density is very low compared to rechargeable batteries. Here you can find a review of a supercap containing a similar amount of energy (Wh) as a NimH AA battery including some size comparions. You can basically make a light as bright as a Zebralight h52 with similar runtimes that is larger than a 2D Maglite (60mm diameter!!!!).

I mentioned NiCads only because that is what was in the satellite power supplies that I had been involved with. This was 20 years ago. Interestingly, a satellite that was launched with our power system came back to earth a couple of years ago. I think they got over 15 years of operation from those nicads. Not bad for going through a charge/discharge cycle every 90 minutes! It wasn't a full charge/discharge, but that's still a lot of use.

Supercaps are pretty big, but I used a nice supercap from Bussman-Cooper in a recent bicycle light design. 100F at 2.7v. Not too bad at all. Larger than a AA, but will run a LED at a decent level for 10 minutes using a little boost converter from Zetex (now Diodes Inc.). This gets back to a fundamental question: what is the spec for run time and for output level? Is this light supposed to let you go find a real flashlight that is in the next room, or is it supposed to run at 3 watts for a few hours?

 

[Fireclaw18]

Very interesting idea.

I'll leave the serious consulting to the experienced designers who previously posted. However, I do have some observations:

1. Unless you're running an extremely low output LED at very low power, heat is going to be an issue. In a typical flashlight the LED is mounted on a small circuitboard called a "star". The star then sits on either a metal pill or a metal shelf that in turn sits against the outside of the light, which is also metal. This provides a direct metal-to-metal connection from the LED to the exterior of the light.

In your light, you have a plastic upper section with a metal lower section. A simple way to heatsink your light is to machine a shelf into the metal section for the star to sit on. That would give you the proper metal-to-metal connection and should provide sufficient heat-sinking for your use.

2. Output. With a partially plastic body and using LiPo batteries at 90% charge, you're not going to compete with hot-rod 1000 lumen all-metal flashlights. And given the use you're planning I don't think it's necessary to try. I suggest an output range of maybe 200-300 lumens. For the emitter maybe a CREE XPG2 at 1 amp or maybe 1.5 amps. Sufficiently low power that heat is not an issue, but bright and modern enough to hopefully sell. If you want to save cash you could try using an older emitter, perhaps a CREE XPG or XML. The older emitter would give 20% less output, but should be cheaper. I'd steer clear of the really cheap emitters like CREE XPE or most of the non-CREE ones. They aren't bright enough to impress.

3. Tint / CRI. Your goal is to produce a flashlight that hides in a table-lamp. It's probably not necessary to have anything more than an LED that lights up. Cool white light in the 6000K+ range is probably sufficient. However, be aware that LEDs come in a wide arrray of tints. Many people find neutral white tints (4200-5000K) to be much more pleasing to the eye. It is also possible to buy LEDs that have "High CRI", which basically means they render colors better. High CRI LEDs tend to be more expensive and have less output though.

4. Modes. Most LED flashlights today come with multiple modes. A lower-power mode may be better for close up use, or going to the bathroom without disturbing your partner. It also prolongs battery life. You should consider whether you want your light to have multiple modes.... and if so, what the user interface will be (such as single-click for low power, double-click for max power, etc.).

5. Turbo mode. Some lights have what I call a "turbo" mode. In turbo mode the light can run well beyond what the heatsinking would ordinarily sustain. But because of power concerns, the circuit ramps it back to a lower power mode. Usually this is done with a turbo timer that reduces current after maybe 1-2 minutes. But it can be done with a thermal sensor that ramps back current if it gets too hot. If you include a turbo mode, you might be able to advertise a higher top output... maybe 500 lumens.

6. Battery life. It is critical to decide how long you want the light to run on a full charge. Do you want it to run 1 hour at max power? Do you want 45 minutes at max power, with the addition of turbo mode with a 1 minute turbo timer, etc. I think the best way to handle this is probably to design it in the following order:

(a) choose the battery chemistry and select batteries that will fit into the light. Once you have done this you'll then know how much power you have to work with.
(b) decide what modes you want and how long you want the light to run in each mode.
(c) choose the appropriate LED. Steps (b) and (c) will be somewhat interchangeable.

 

[bpierce]

Very interesting idea.

I'll leave the serious consulting to the experienced designers who previously posted. However, I do have some observations:

1. Unless you're running an extremely low output LED at very low power, heat is going to be an issue. In a typical flashlight the LED is mounted on a small circuitboard called a "star". The star then sits on either a metal pill or a metal shelf that in turn sits against the outside of the light, which is also metal. This provides a direct metal-to-metal connection from the LED to the exterior of the light.

In your light, you have a plastic upper section with a metal lower section. A simple way to heatsink your light is to machine a shelf into the metal section for the star to sit on. That would give you the proper metal-to-metal connection and should provide sufficient heat-sinking for your use.

2. Output. With a partially plastic body and using LiPo batteries at 90% charge, you're not going to compete with hot-rod 1000 lumen all-metal flashlights. And given the use you're planning I don't think it's necessary to try. I suggest an output range of maybe 200-300 lumens. For the emitter maybe a CREE XPG2 at 1 amp or maybe 1.5 amps. Sufficiently low power that heat is not an issue, but bright and modern enough to hopefully sell. If you want to save cash you could try using an older emitter, perhaps a CREE XPG or XML. The older emitter would give 20% less output, but should be cheaper. I'd steer clear of the really cheap emitters like CREE XPE or most of the non-CREE ones. They aren't bright enough to impress.

3. Tint / CRI. Your goal is to produce a flashlight that hides in a table-lamp. It's probably not necessary to have anything more than an LED that lights up. Cool white light in the 6000K+ range is probably sufficient. However, be aware that LEDs come in a wide arrray of tints. Many people find neutral white tints (4200-5000K) to be much more pleasing to the eye. It is also possible to buy LEDs that have "High CRI", which basically means they render colors better. High CRI LEDs tend to be more expensive and have less output though.

4. Modes. Most LED flashlights today come with multiple modes. A lower-power mode may be better for close up use, or going to the bathroom without disturbing your partner. It also prolongs battery life. You should consider whether you want your light to have multiple modes.... and if so, what the user interface will be (such as single-click for low power, double-click for max power, etc.).

5. Turbo mode. Some lights have what I call a "turbo" mode. In turbo mode the light can run well beyond what the heatsinking would ordinarily sustain. But because of power concerns, the circuit ramps it back to a lower power mode. Usually this is done with a turbo timer that reduces current after maybe 1-2 minutes. But it can be done with a thermal sensor that ramps back current if it gets too hot. If you include a turbo mode, you might be able to advertise a higher top output... maybe 500 lumens.

6. Battery life. It is critical to decide how long you want the light to run on a full charge. Do you want it to run 1 hour at max power? Do you want 45 minutes at max power, with the addition of turbo mode with a 1 minute turbo timer, etc. I think the best way to handle this is probably to design it in the following order:

(a) choose the battery chemistry and select batteries that will fit into the light. Once you have done this you'll then know how much power you have to work with.
(b) decide what modes you want and how long you want the light to run in each mode.
(c) choose the appropriate LED. Steps (b) and (c) will be somewhat interchangeable.

Fantastic feedback -- thanks! Based on what I'd seen from relatively cheap flashlights at big box retailers (ones that also sell lamps), I was also thinking that something in the 250 lumens range may suffice for an initial version of the product. Both you and Steve K have mentioned the importance of stating a desired run time. Realistically, I think 45 minutes to an hour is likely to suffice initially. I'd love to be able to crank that way up over time, but TBD how realistic that is in an initial release. I wasn't necessarily planning on having multiple modes on the flashlight, but providing a low output mode could make a lot of sense if it helps extend the run time in an emergency.

Eventually I'd love to offer both premium products that target a smaller market and also a mass market product at a lower price point. Based on the price in my Kickstarter (between $40 and $50) you can see that I decided to start with the mass market version. So I'll probably have to have a few omissions and compromises; it's likely, for example, that a turbo mode is out for now. I definitely like the idea of the LiFePO4 batteries -- they're safer and more environmentally friendly, which is inherently good and nice from a marketing angle -- and coupled with a roughly 250 lumen output for 45 mins it seems I should be able to back into the right CREE as you suggest.

The feedback about specific components to consider (the specific CREEs that could make sense) is particularly helpful. Thanks again!