[Dougigs]

There's no reason for this otherwise useful discussion to descend into flaming. Morsel's information was solid -- a) because it followed the essential linear-design principle of 'think of the worst possible signal conditions and add 100 per cent' (one of the many reasons why her massively over-specd M3 amp is such a great cirtuit), and b) because, remember, this thread opened with my interest in hauling batteries out of cell phones and playing with them -- which, if it were done with the assumption that they behave like ordinary batteries, would quickly produce something unpleasant (the worst-case scenario of 1A going into an already over- or under-charged cell is very, very easy to produce if you're working under such misconceptions).

However, as I'll point out in my next post, there are some safe and well-tested DIY solutions...

 

[Dougigs]

Everyone with any interest in these batteries should immediately download the project "simple li-poly charger" from the April 2005 issue of Elektor Electronics. It manages charging for a 3-cell, approx. 12-volt li-poly pack (2300MaH is widely available, they claim, and it's extremely small). They have a PCB available for 14.60 euros, and its very safety-conscious design uses readily available, non-SMD parts.

Their website: www.elektor-electronics.co.uk ... you want to search for U050460.pdf , the download of which will cost you one pound (it uses the secure BT payment site).

 

[Dougigs]

Ok, it turns out that the folks who are into radio-controlled helicopters are well ahead of us on this.

If you search EBay for "li-poly" or "li-polymer," you'll find a good selection of 11.1-v cells of up to 2200 MaH -- just right for our needs -- in the 23-pound (50 buck) range. (there are even cheaper ones, but they seem to have strange connectors).

These are the cells meant to be used by the elektor circuit above. They typically measure about 2.5 inch by 1 inch by 1.5 inch. Some even come with wall-wart chargers....

 

[KZEE]

Quote:

Originally Posted by morsel ...but this forum is full of people without the engineering background or common sense to work with hazardous technology. Many have never seen an oscilloscope in real life.

Full of people with no common sense? Ouch, you think very highly of yourself don't you. I've only been participating in the Head-Fi DIY forum for a couple of years or so, but unlike you my observation has been that the vast majority of Head-Fiers participating in this forum are very thoughtful and knowlegable about electronics in general, and the vast majority of those Head-Fiers would seem to have a good measure of common sense coursing through their viens.
This ain't rocket science, and I would direct your attention to the radio controlled model car and airplane hobbies - they've been routinely working with and using lithium technology for years, and you don't hear horror stories coming from them about lithuim technology biting them in the buttocks. Do you think many radio control hobbiests have ever seen an oscilloscope in real life?

 

[Dougigs]

Really, let's stick to the technology and put the ad homenims aside.

And on that note, I should point this out: The Elektor circuit seems good, but its chip is a bit hard to source at reasonable prices.

It turns out that there's a good market in tiny SMD-based li-poly chargers built for 3-cell units -- just visit the sites that sell stuff for rc helicopters. I just ordered one for US$16; it uses the same, safe technology as that Elektor circuit (and is much smaller, once its pcb is removed from the plastic case and mounted in a little amp box). These things typically use 12 to 18 volts DC input, so you could easily rig up a headphone amp with a jack on the back that you could plug your laptop's adaptor into for charging. So if you're on the road (as I am a lot), you only need to bring one ac adaptor...

 

[tangent]

Quote:

Originally Posted by Frenchman These companies then buy cheap batteries from second rate manufacturers

If you think some DIYers won't do the same, you haven't been paying attention.

Quote:

Unless you plan on buying thousands of them, you won't deal with these companies.

That's right. We will instead be dealing with a company two or three removed from these distributors, but still in the same chain.

Quote:

The world is switching to Lithium based rechargeable chemistries

"If your friends jumped off a bridge..."

Quote:

I am currently using 4 high discharge Lipo packs

I never said it couldn't be done. I simply wanted to impress on people that the dangers are higher than with nickel-based chemistries.

Quote:

Why they have a unnatural fear of these batteries is quite beyond me.

I dunno...perhaps it's because fear of fire is wired into every mammal's hindbrain. Does that mean we do not use things that can cause fire? Of course not. It does mean that we have a moral obligation to pass on lore that may prevent people from getting burned, though.

Quote:

I believe this means electrolytic capacitors are unsuitable for hobbiests!

A bad electrolytic can't burn your house down.

 

[tangent]

I've been thinking more about this, Frenchman, and I think there's a key difference between people like you and people like Morsel and I. We are the primary point on contact on some very popular projects. We see questions and disasters that you don't.

I have received -- multiple times -- questions from people asking what the battery life of DIY amp X is. No details about the circuit. It's like they think they're asking a marketing department somewhere for a spec missing from the "product" page. When you reply with an answer that boils down to "read the datasheet", many are offended, as though this is an unreasonable answer. There are self-styled DIYers out there who cannot or will not read a datasheet.

If you answer enough of these questions, and see the disaster that sometimes results, you come to a point where you start dividing people into two classes: those who could figure out the answer to their own question if you prod them in the right direction, and those who cannot or will not. You start answering questions as though everyone is from the first class (give 'em the benefit of the doubt) and if it becomes clear that they're from the second class, you take evasive action.

You are of the first class, Frenchman. You figured out how to build a LiPo charger from reading the datasheet. I'm happy for you, truly. Not everyone can do that, but there are still people from the second class who will try. I don't want responsibility for those people.

 

[Frenchman]

Selectivly screening the google results? I took the link you provided, popped it into the IE address bar and read DOWN THE LIST.

You are not required to be a primary point of contact. If you assume that position, then that's fine. In this thread, you are providing information only. The dangers may very well be highly. But any battery can be abused to a condition where it will be destroyed. It just so happens that the circuits to charge more simple chemistries are more easily accessable. Modern circuits for charging and protecting Lithium rechargable chemistries are DEAD EASY.

You say you are merely trying to impress the dangers upon people.That's fine. I've personally never seen a lithium based pack explode. From what I understand, such instances are extremely rare! Apples defective battery recall was prompted after SIX cases. Six. Out of... how many thousands of iBooks? Impress all you like, but you arn't just impressing. You're flat out saying they are UNSAFE and shouldn't be used. The poster seems very interested in using these packs, so if you want to impress, you've done so. Now he wants help figuring out how to use them.

To be honest, most of the chargers I use are the very nice chargers that you can buy from Kokam (the manufacturer I use and trust), I have only built a single charger. Besides, what is the point of showing that you can blow up a cell if you put a huge reverse voltage on it? My favorite part of that site you showed is this line from the forum post where he talks about that setup:

"NOTE:
This demonstration was performed only to show what a single lipo cell looks like IF it vents and ignites from an overcharge. No conclusions should be drawn from the information given, as other cells or packs may not perform in the same manner."

It must be noted that he was grossly abusing the cell. Really there isn't much to charging them. You charge at 0.5C until the voltage hits 4.2V and stop. Discharge until you drop to 3.0 and cut off.

Instead of railing on about what could happen if you grossly abuse a cell, why put out the right information and let the people design the circuits. The kind of catastrophic battery failure is no more dangerous or likely to happen then overcharging a lead acid for a long period. Lets face it. All batteries have dangerous harmful chemicals in them. If you don't know what you're doing, you can damage any cell made of any chemistry. Lithium Ion/polymer is DIFFERENT but not impossible or even especially dangerous. Also, let's note that there have only been one or two instances of "exploding" batteries. And these weren't even actually flaming ones, they were cell phone batteries that vented gas until the case popped apart. The apple recall is for batteries that can OVERHEAT and cause a firehazard, they don't light on fire.

Anyway, I'm not saying any of the stuff you talk about can't happen. But I am saying that these batteries arn't "little bombs, waiting to blow up in the hands of unqualified hobbyists." There are many far less qualified airplane people using these batteries and now they have all switched to this technology. If these batteries were so difficult to use and so dangerous, they wouldn't be the highest selling rechargable technology out there.

 

[Dougigs]

Oh boy. Flame wars like these (invariably between hobbyists and engineers) are why I stopped frequenting several other DIY forums, and their relative absence is why I decided to start visiting this relatively harmonious and very productive community this summer. Frenchman: vous ne devriez pas prendre ces personnes tellement au sérieux. Engineers sometimes sound haughty and dismissive, but we sound like bumbling idiots poised on the edge of catastrophe to them. Both views have some foundation in reality (theirs is rooted in the stunted social development of their educational milieux, ours in our decision to toy with technologies whose actual workings are beyond our knowledge, leading to a sort of superstitious naivete.)

The best approach is politely to ignore the worst excesses in either direction, and focus on the actual information that might help you.

So, Frenchman, why don't you share some of your actual experiences with li-poly cells with us? What circuits have worked, and what haven't? How do you keep your cells from discharging below an undesirable level? Do you have your cells hooked up to a comparator circuit that provides an indicator light, or cuts them off? Or do you keep them hooked up full-time to a monolithic charger that monitors their levels even when DC isn't connected?

What are your experiences with full discharges? This is a weakness of headphone amps for these types of cells: We tend to listen until we're completely out of juice. Well, li-polys aren't supposed to undergo this -- so what are you doing to prevent it?

I'm interested in hearing how these things bahave at different charging levels... One reason why laptop batteries only last a year is that most people keep them charged at 100% most of the time (i.e. plugged into DC), and this condition is actually bad for li-ion batteries. Now, li-poly datasheets indicate that this is less of a problem, but I'd like to hear first-hand.

Pictures would also be nice.

 

[Frenchman]

The majority of what I use are premade chargers. They just work and there really isn't a reason to build a charger for the packs that we use. The charger that has been designed and built is somewhat interesting. It's for a Cube Satellite, a 10cm cube which we design and build and Stanford launches for us. It has solar panels that generate electricity to recharge a set of Lipoly batteries. On this, we simply use comparators to drive the enable pins on various switching regulators. The high side comparator drives the solar panel converter enable pin so that when the batteries are fully charged (aka the voltage is up to 4. whatever it is in the datasheet), the regulator is turned off so it no longer charges the batteries. The low side comparitor drives buck switching regulators that drop the battery voltage down to 5 and 2.8V. When battery voltage drops too low, the enable pin is cut off and the regulator stops sourcing current.

These circuits work very well. Using a 10 ohm resistor, we've tested discharge conditions under several temperatures and it cuts off every time. The charger also works perfectly. Most of the smart chargers simply have the charging curves programmed in. In our particular application, we can't charge the batteries at 0.5C so we actually have the upper charge voltage set a little different because we charge at like 0.2-0.3C. We've left these circuits charging and discharging in the sun with various loads on and we've never had even a slightly unsafe condition arise.

If you want the full range of your battery, it's best to get a smart charger. It can charge the battery much easier then you can with a much much lower part count. The majority of these are just controllers, they drive an external fet or bjt. They are VERY easy to use. Supply them DC and they work, don't supply them and they don't work. They will charge the battery as long as DC is hooked up and will charge the battery up to it's safe upper limit. Some of these can even safely charge cells that have been discharged somewhat below the safe levels. Protection circuits are also pretty simple to construct. They protect against under discharge, short circuit and some of the nice ones even protect against over charging!

Really, you just need to look through datasheets from Maxim. Their products are very nice and very easy to use. The application you're looking for is exactly what their typical apps are, so the stuff they put in there is right up your alley.

To your question about discharging. As I said previously, multiple small discharges extend the life of the battery considerably, full discharges are especially hard on them. But on the flip side, people who use them constantly plugged into the AC use up all their many short charges very quickly. Use it every day and plug it in every two days to recharge and you'll get several nice years out of a modern pack. Unfortunatly we don't have any serious long term data for our regular discharge conditions, we are simply going off manufacturer info. Ours haven't been run enough times to come close to hitting even the lower limit on cycles.

 

[Dougigs]

That's quite useful, Frenchman, thanks -- there are some interesting solutions out there.

The charging aspect is pretty easy, and there are lots of solid, safe options that can fit into a tiny headphone-amp case, both in terms of dedicated chips and pre-made charger pcbs (often using SMDs) out there (I'll report on my results).

But the one aspect of this application that I need to get right is discharge protection. It turns out that a lot of these li-po cells sold for model-airplane use do not contain any undercharge protection circuitry. This is of course quite serious: You really don't want to let these things discharge below 2.8v per cell (or 8.4v for a 3-cell pile -- maybe 9v to be safe) If that happens, at best you might lose your $70 battery; at worst, you get a movie like the one Morsel showed us.

Just to stress this for those who may have just come into this: YOU CAN'T PUT LI-POLYMER BATTERIES IN A HEADPHONE AMP UNLESS YOU HAVE A CIRCUIT IN THERE THAT DISCONNECTS THE BATTERIES COMPLETELY THE INSTANT THEIR VOLTAGE DROPS BELOW 2.8V PER CELL.

As it turns out, there are some great options here. You can simply use a comparator attached to a small voltage reference and driving a p-channel MOSFET that switches the battery in or out... (you'd also want to have a dpdt swich that flicks the battery over to the charger... this would function nicely as an on-off as well)

But as it happens, there's a great little chip called the MAX1614 that does all this quite neatly -- it is an undervoltage detector/battery shutoff chip, containing a comparator, a reference, some hysterisis circuitry and MOSFETs -- it also contains a little flip-flop so you can use a momentary pushbutton switch to turn on and off your amp.

It's cheap and widely available, uses supplies up to 16v, and is in the teeny tiny SSOP-8 package (SSOP to DIP adaptors are available from Browndog in USA or RS in Britain). Its datasheet is very clear on how to set it up and calculate everything, and it only needs about 4 passive components.

I will let you know how it goes. Anyone have any reverse-engineering notes on the handful of headphone amps that do use these batteries?

 

[mono]

If the cells are to be disconnected 2 8V, then what is the point of them? I'm not suggesting that we should run them down further, but this idea about the benefit of Li-Poly cells is marginal at best.

IF one had a randomly-selected case that was just barely too small for AAA cells, yes they might squeeze Li-Poly in it... except you need protection circuitry too. You need the expensive charger, manditory. More expensive cells. Inherant risk aside, Li-Poly just doesn't look attractive. It might be different if there were some small deck-of-cards-sized 500mAH 18V cell pack, but that's not what's being considered here.

 

[Dougigs]

Mono -- "Full discharge" on a li-poly cell is considered the 2.8-volt point (or whatever the danger cutoff point is) on any datasheet. So if it's a 2400mah cell, that means it delivers 2.4 amp hours from full charge to the 2.8-volt point (check the datasheets for some cells to confirm this).

The benefits for headphone amps are real. An 11.1v li-poly cell occupies slightly less space than 2 9V batteries; it offers 2400 mah of current whereas the 2x 9v, if series-connected, give you just over 200 mah. They also weigh less than one of the 9v cells. If the 9v into a 40ma headphone amp gives you 5 hours of play time, the li-poly will give you 50 hours.

That 11.1 cell will cost you US$50. I've used that much worth of copper-top 9V in the last month, even though I recharge my copper-tops.

And the battery-life figures for these things are much better than for li-ion cells or many other rechargables: After 500 full discharges on a typical li-poly cell, full capacity drops only to 80 per cent.

The charging and discharge-cutoff circuitry is cheap, easy to design, and extremely small. I'm working on a prototype the size of half a MINT circuit board that contains a) a 14-pin SMD li-poly smart-charging chip that cuts off at full charge etc; b) an 8-pin SMD discharge-protection chip that cuts the battery out at 2.8v (or slightly higher, for a safety margin) and provides a power-switching circuit; c) a single small MOSFET; d) about 5 resistors and 3 other passive components, all of which can be SMD.

Using such a unit would be pure simplicity: It has a 2.5mm socket in the back. When the battery discharges and the audio cuts off, you plug your laptop-computer power pack into the back for 2 hours, until the charge-done LED lights, and then you've got 4 more days of dawn-to-dusk listening on a PIMETA or a loaded MINT.

Total cost for all these parts is about US$20. Or you could get a pre-made charger board, about 2 inches square and fully SMD, for US$16 from various radio-controlled hobby shops, and build a discharge-cutoff board -- same cost, maybe a bit easier.

It won't quite fit into an Altoids tin but it will fit into one of the smallest Hammond cases. Using the flattest, smallest mobile-phone cells could get the whole thing into a mint tin, but the control PCB and extra LEDs etc would make it a very tight squeeze indeed. I'll leave that experiment to some non-drinker with tiny fingers.

You're right that there are easier battery approaches. The nimh fast-charge board for the PPA is a great design that can easily give you the same amount of charge capacity with a dozen nimh AA cells. I'm interested in experimenting with these because they provide the same capacity and rechargeability in a truly pocket-sized enclosure.

NOTE AGAIN: YOU MUST NOT USE THESE WITHOUT A RELIABLE CIRCUITS THAT CUT THE BATTERIES OUT WHEN THEY'VE CHARGED UP TO PEAK POWER OR DISCHARGED BELOW 2.8V PER CELL. ANY OTHER CONFIGURATION IS POTENTIALLY VERY DANGEROUS.

 

[Dougigs]

I wonder if this is worth worrying about:

The MAX1614 seems like the perfect little chip for handling li-poly cells in a small headphone amp -- tiny, simple, few external components, reliable.

However, it is able to drive its switching MOSFETS on battery voltages because it uses a charge-pump circuit. This is built around a 50khz oscillator inside the chip.

Obviously I get a bit uncomfortable thinking about a 50khz oscillator sitting in an audio circuit. Even if it is just providing the current thad drives the gate of the MOSFET that turns on the power-supply voltage, which is subsequently filtered somewhat.... Is this likely to be put any artifacts into the supply? Or interfere in any other way? In other words, am I better sticking to an olf-fashioned comparator?

 

[mono]

Quote:

Originally Posted by Dougigs The benefits for headphone amps are real. An 11.1v li-poly cell occupies slightly less space than 2 9V batteries; it offers 2400 mah of current whereas the 2x 9v, if series-connected, give you just over 200 mah.

Who said you had to use only these or 2 x 9V? We'd have to count the extra space to end up with 18V of Li-Poly cells if you're going to compare these two alternatives, else a declaration about what kind of cans someone will need to have per a particular gain scenario, as well as which amp.


Quote:

They also weigh less than one of the 9v cells. If the 9v into a 40ma headphone amp gives you 5 hours of play time, the li-poly will give you 50 hours.

The thread opened with mention of 900mAH cells. I may have overlooked these 2400mAH cells? I don't think weight is an issue though, we're talking about a few ounces as a percentage of total amp weight.

On the one hand you're trying to argue Li-Poly is smaller, then here you're arguing about better runtime when the 9V is smaller. Do we really need 50 hours of runtime? AT what cost and time to implement?

Quote:

That 11.1 cell will cost you US$50. I've used that much worth of copper-top 9V in the last month, even though I recharge my copper-tops.

We both know $50 is significantly more expensive than one or two 9V NiMH, put in the context of price comparisons. IF you find a particular application where Li-Poly works well, you might justify it, but that's a completely different matter than trying to argue outside of that context.

Quote:

And the battery-life figures for these things are much better than for li-ion cells or many other rechargables: After 500 full discharges on a typical li-poly cell, full capacity drops only to 80 per cent.

Considering their cost, AND the additional cost to implement them, that's pretty much irrelevant. It makes them seem marginally better but is one of many factors.

Quote:

The charging and discharge-cutoff circuitry is cheap, easy to design, and extremely small.

Compared to a car, yes all of the above is true. Compared to already employed means of powering headamps, no, it'll be a little smaller and that will only be of benefit in particular cases, with particular amps where you'd run out of room otherwise.


Quote:

I'm working on a prototype the size of half a MINT circuit board that contains a) a 14-pin SMD li-poly smart-charging chip that cuts off at full charge etc; b) an 8-pin SMD discharge-protection chip that cuts the battery out at 2.8v (or slightly higher, for a safety margin) and provides a power-switching circuit; c) a single small MOSFET; d) about 5 resistors and 3 other passive components, all of which can be SMD.

I don't doubt that it's nice and that you should be complimented on your work when it's all done. However, you seem more interested in swaying others towarsd this goal you haven't even realized yet. After countless hours you will have educated yourself well on these issues but unless you hide away your design, it's pretty inherant that others will be less educated about it, else they too will spend quite a bit of time reinventing the wheel so to speak, at great cost, again only benefitting with particular case, amp, and useage or recharge pattern (intervals). What you end up with might be a godsend to someone out backpacking for the weekend but I don't see where 50 hours between charges is worth the time or $ to "most" typical uses.

Quote:

Using such a unit would be pure simplicity: It has a 2.5mm socket in the back. When the battery discharges and the audio cuts off, you plug your laptop-computer power pack into the back for 2 hours, until the charge-done LED lights, and then you've got 4 more days of dawn-to-dusk listening on a PIMETA or a loaded MINT.

So you're going to be lugging around a laptop battery to recharge a 2nd, lower capacity but just as expensive (when all is said and done) battery? I would tend to think there's plenty of time to recharge batteries while the owner sleeps unless you have some kind of special mission you're not telling us about.

Quote:

Total cost for all these parts is about US$20. Or you could get a pre-made charger board, about 2 inches square and fully SMD, for US$16 from various radio-controlled hobby shops, and build a discharge-cutoff board -- same cost, maybe a bit easier.

So it's about $70 minimum to lose a few ounces and "maybe" a few cubic CM of space... or maybe not, if this control board doesn't integrate into any particular case well it may end up being larger. 2 sq. inches is a lot of space for batteries.

Quote:

It won't quite fit into an Altoids tin but it will fit into one of the smallest Hammond cases. Using the flattest, smallest mobile-phone cells could get the whole thing into a mint tin, but the control PCB and extra LEDs etc would make it a very tight squeeze indeed. I'll leave that experiment to some non-drinker with tiny fingers.

What about impact protection? It's not going to be good to squeeze any such cells right up against a thin-walled metal (container). Are you considering a Mint or Pimeta? $70 for powering a Mint seems like a very luxurious endevour to save a few cu. cm of space considering there's still the source and laptop - unless the laptop IS the source, but then you completely avoid the whole issue and could've just ran the amp from the laptop battery and completely avoided the losses from recharging the Li-Poly cells. You must have some specific goal a bit more involved than mentioned thus far.

Quote:

You're right that there are easier battery approaches. The nimh fast-charge board for the PPA is a great design that can easily give you the same amount of charge capacity with a dozen nimh AA cells. I'm interested in experimenting with these because they provide the same capacity and rechargeability in a truly pocket-sized enclosure.

Contrary to what my devil's-advocate post would seem to imply, I'm not dead-set against using Li-Poly cells given the need... but it seems like like a "I can therefore I will" rather than an expressed need, at a substantially higher cost, complexity, and potential danager. Accidents tend to happen because of the things one doesn't forsee.

Quote:

NOTE AGAIN: YOU MUST NOT USE THESE WITHOUT A RELIABLE CIRCUITS THAT CUT THE BATTERIES OUT WHEN THEY'VE CHARGED UP TO PEAK POWER OR DISCHARGED BELOW 2.8V PER CELL. ANY OTHER CONFIGURATION IS POTENTIALLY VERY DANGEROUS.

I think it's good and useful that you're being a pioneer here, but along with the benefits also come the drawbacks. It would only be fair to point those out to keep the discussion on an even keel.