dynohub charger circuit for LiFepo4 cells

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
Using linear regulator with NiMh cells on winter time is quite frustrating, battery power dies very fast.


Is anyone tried LiFepo4 cells and build charger for them what will work with dynohub. Will LifePo4 cells work better?


Current system is simple, dynohub 7v.3w, full-bridge rectifier, 78xx(fixed 11 v) with 8x 1.2 v. NiMh rechargeable cells, headlight 3x 1w and taillight 2x?red they both have own led-driver cause they are different type. It's "trash_made", only leds and veroboard are from shop and all other things came from broken electronic (and yes I know that TV receivers, accubatteries, old capasitors etc. can cause Nuclear disaster and terrorist can steal them from my bike... :huh: ). So new system might be need fresh shop hardware but testing unit can still made from recycled trash.


Maybe switching power-supply and more effective leddrivers helps, any good ideas. System size or weight is not big issue cause I never take off bike panniers but lead-acid accu system makes bike unbalanced and that way it's too difficult to carry on third floor apartment.


If my enlish is not readable or I presented myself too vague, please, brace yourself with Courage and tell me, I will try my best to make clear what i tried to tell or ask.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Using linear regulator with NiMh cells on winter time is quite frustrating, battery power dies very fast.

Is anyone tried LiFepo4 cells and build charger for them what will work with dynohub. Will LifePo4 cells work better?

A good question.
What improvement will the lithium battery offer, and what do you think is the problem with your current system?

I used to have a system that charged 5 nicad AA cells from the dynamo, and then used those cells to power the incandescent headlight. The design provided a standlight function, which was very nice. Worked great in the summer months. It didn't work well in the winter, partly because all of my commute was in the dark, so there was no time when the battery was getting charged and not providing some output to the headlight, and partly because the nicad's internal impedance increases when cold. The increased impedance was a problem because I used voltage regulation for the charging, and the increased impedance meant that the charging current was less.

Overall, it just wasn't a good design for how I wanted to use it. In my recent design....
http://www.flickr.com/photos/kurtsj00/sets/72157621965148305/
...., I still use a nicad for my standlight, but the standlight is a much lower power light than the headlight. Works great!

In my first design, the underlying problem is that the energy being put into the battery just wasn't as much as the energy being taken out of the battery over the course of daily commute. The battery itself was mostly irrelevant (although the increased impedance in cold weather was part of the problem).

I use my bike & light in very cold temperatures... down to 0 degrees F, sometimes. In those temperatures, I believe that recharging a lithium battery is not a great idea. You'd have to check with the battery manufacturer to learn what their specific recommendations are, though.

regards,
Steve K.
 

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
Well, Steve if your builds amazes people, I'm impressed your quick reply to my n00b -post and yes your reply was heplfull many way. Thanks!

A good question.
What improvement will the lithium battery offer, and what do you think is the problem with your current system?

For current system problem lies in "linear" line, system can't draw low power out from battery so when I have 3 x power in series they need power what is not available (headlight) but tail have LDO style regulator and 2x 2.2 v leds series, they light up walking speed if batteries are depleted.

LiOn batteries, hmm, I got one just today for free, no power, but loose joint in protection circuit, ;). It seems quite powerfull 4.02 v. (3.7v) and 1800mAh and my cellphone charger seems to work with it. I think LiOn gives some Boost but it lacks cold weather stability and recharge also I need nice boost-driver to run three powerled from 3.7v cell and good step-down driver to make dynohub to effective charger for LiOn.

I can use my cellphone charger too it needs 12 AC from dynamo to kickstart and then it will charge slowly, it's made 90 - 240 AC but works with dynohub.

I used to have a system that charged 5 nicad AA cells from the dynamo, and then used those cells to power the incandescent headlight. The design provided a standlight function, which was very nice. Worked great in the summer months. It didn't work well in the winter, partly because all of my commute was in the dark, so there was no time when the battery was getting charged and not providing some output to the headlight, and partly because the nicad's internal impedance increases when cold. The increased impedance was a problem because I used voltage regulation for the charging, and the increased impedance meant that the charging current was less.

That explain a lot, so current just run trough? I'm not electronic engineering, but I think I got idea what happens.

Overall, it just wasn't a good design for how I wanted to use it. In my recent design....
http://www.flickr.com/photos/kurtsj00/sets/72157621965148305/
...., I still use a nicad for my standlight, but the standlight is a much lower power light than the headlight. Works great!

In my first design, the underlying problem is that the energy being put into the battery just wasn't as much as the energy being taken out of the battery over the course of daily commute. The battery itself was mostly irrelevant (although the increased impedance in cold weather was part of the problem).

I use my bike & light in very cold temperatures... down to 0 degrees F, sometimes. In those temperatures, I believe that recharging a lithium battery is not a great idea. You'd have to check with the battery manufacturer to learn what their specific recommendations are, though.
Yep, cold is problem and NiMh are not working in cold. Here in Finland we have extra warm winter just now but if it go under zero Celsius normal - 10 to -15... batteries loose their charge what comes back when you bring bike back in.

Maybe solution is, switching buck-boost regulator, lifepo4 cells with protection circuit or self-build http://www.microchip.com/wwwproducts/Devices.aspx?dDocName=en543954, and some buck or boost led-drivers. DX might be good place to get those LiFepo4 cells for test. I readed lot off stories from zetex based led drivers, are they easy to build or is it better (cheaper) get them from DX or some similar place. :thumbsup: Last step is lead-battery but I think that my next project is order those Lifepo4 (4x to get 12v. up) cells and look how long the last.

How you make your stand-light to go off automatic? I dislike switches (Rohloff click-click is nice but ...)
 

Bandgap

Enlightened
Joined
Aug 14, 2002
Messages
284
Location
London
You have to do some calculations, on how long you are stopped against how long you are cycling.

Then you can work out how much power you have to divert into the battery when you are travelling, tosupply the lights when you are stopped.

In my experience, unless you do a lot of cycling and very little stopping, it is better to accept a dimmer headlight when you are stopped (and only need to be seen rather than needing to see 50m in the distance), as this reduces the amount of power needed into the battery.

As the wise Mr Steve K has warned me before, and said above, don't charge Liion cells below 0°C unless they are specifically designed for it.
I struggled to find a minimum charge temperature for LiFepo4 cells on the web. The only one I did find was also 0°C.

AFAIK, NiCd is best for low temperature charging (although worse for the world if you dispose of them unwisely).
Liions stor almost 100% of the charge you put into them, NiCd and NiMH only store about 70% and turn the rest to heat.

NiCds can generally withstand C/10 overcharge for ever.
some NiMH can not take over-chrge, some can take +C/10 for 1 year.
Liiion cannot take any overcharge at all for any time at all.

Liions need accurate (1%) voltage charging if charging to max rated voltage.
Charging to less voltage increases life - although even this does not allow you to charge below 0°C.

Lastly, it is far better to shunt regulate a dynamo that series regulate it, unless you always cycle slowly, even down hill.


Another Steve
 
Last edited:

Bandgap

Enlightened
Joined
Aug 14, 2002
Messages
284
Location
London
I found one other reference to minimum charging temperature for lithium iron phosphate - it said -15°C for a particular type of cell - not generally.

Steve
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Mr. Bandgap/Steve makes a good point by emphasizing the fundamental issue regarding standlights and charging batteries: You need to be sure that the dynamo can supply the needed energy. If you ride for 1 minute, and then are stopped for 5 minutes at an intersection, and do this over and over, your standlight should draw very little power!

There is also a question about how your system is configured. It sounds like you are using something like this.....
dynamo --> rectifier --> battery charge regulator --> battery --> led driver --> leds

This can work, but there are potential problems that have to be addressed.
1. can the led driver sense when the dynamo is stopped, so that it can drive the leds at a lower power when stopped?
2. can the battery charge regulator survive the range of dynamo output voltage? With a small load (i.e. the leds are off and the battery is fully charged), dynamo voltages can easily be over 50v, and I've seen 100v at high speeds.
3. can the battery survive the temperatures that it will be used at? The biggest issue is cold temperatures, but hot temperatures could be a problem for some people.
4. does the battery charger perform okay if the power is interrupted during the charging process?
5. Will the LED driver turn off at the correct voltage to avoid damaging the battery by excessive discharge?

these are just the questions I can think of right now.

There are lots of ways to have a dynamo charge a battery to provide some light while stopped. The hard part is to understand the limitations of the parts that you want to use, and be sure that everything will work before you spend the time and money to put it together. If time and money aren't a problem, then the only problem is having the light die while you are riding in the dark. :)
...that reminds me... always have a spare headlight with you! I do!

regards,
Steve K.
 

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
Lastly, it is far better to shunt regulate a dynamo that series regulate it, unless you always cycle slowly, even down hill.


Another Steve

Good points you made, but "shunt regulate.." meaning of the word, :confused: and how that kind off thing affect my dream to make cold resistant/rechargeable lighting system?
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
quick description of a shunt regulator vs a series regulator:
http://www.flickr.com/photos/kurtsj00/6686598579/in/photostream

The shunt regulator uses the internal resistance or impedance of the power source (i.e. dynamo) as a way to reduce the source's output voltage. By drawing more current from the power source, there will be more voltage dropped across the internal resistance, thusly reducing the output voltage. This is only viable when the power source has a significant internal resistance.
This is fairly common with solar panels (photovoltaics) too.

Steve K.
 

minisystem

Newly Enlightened
Joined
Jul 12, 2011
Messages
96
I used a now out of date Lithium ion charging controller for a dynamo powered standlight system. It worked well enough but I ultimately decommissioned it as some of its shortcomings (due to my own bad design) bugged me. There are a bunch of lithium ion charge controller chips available. None of the current batch out there seem as simple or straightforward as the one that I used, but alas it is no longer available. Details are here: http://minisystem.blogspot.com/2008/07/dynamo-lithium-ion-battery-charging.html
 

Bandgap

Enlightened
Joined
Aug 14, 2002
Messages
284
Location
London
Regarding shunt regulators

Just to add to what the Wise Mr K said, dynamos have got a difficult characteristic: that the output voltage rises if you take less power (given enough speed).

And that voltage can easily be high enough to blow up things like 7805 series regulators.

Far better to partially short out the dynamo and waste unwanted energy in a resitor or transistor, keeping the voltage low.

And if you do the maths, the increased waste due to partially shorting a dynamo, is generally lower than the extra waste from the higher voltage caused by the series regulator.

The-less-wise-Steve (and slower on a bike!)
 

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
Regarding shunt regulators

Just to add to what the Wise Mr K said, dynamos have got a difficult characteristic: that the output voltage rises if you take less power (given enough speed).

And that voltage can easily be high enough to blow up things like 7805 series regulators.

Far better to partially short out the dynamo and waste unwanted energy in a resitor or transistor, keeping the voltage low.

And if you do the maths, the increased waste due to partially shorting a dynamo, is generally lower than the extra waste from the higher voltage caused by the series regulator.

The-less-wise-Steve (and slower on a bike!)

Yes we can now waste some power to get power.


Well, I see my problem is Apollo 13 like, you need power to get power and when battery freezes there is no power at all, so they need to keep them warm so they feed some power to keep them alive...


Now it needs to make some testing with 3 to 5 watt ceramic resistor, must make insulated aluminium box for batteries and NTC-powertransistor based simple thermalcontroller also have to test will that circuit burn my bike, if short circuited. Maybe it works for long 3 - 5 h bicycling sessions.


ps. ordered just LiFepo4 cells from DX and waitting to get them to test (no dynamo charging for them before good research. ..)
 

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
I used a now out of date Lithium ion charging controller for a dynamo powered standlight system. It worked well enough but I ultimately decommissioned it as some of its shortcomings (due to my own bad design) bugged me. There are a bunch of lithium ion charge controller chips available. None of the current batch out there seem as simple or straightforward as the one that I used, but alas it is no longer available. Details are here: http://minisystem.blogspot.com/2008/07/dynamo-lithium-ion-battery-charging.html

Thx, I have too old light casings but I just fixed three 1 watt powerled inside. ;)

Idea is just good to see on real life!

I have to regulate my dynamo power cause "speeding up" burn system what can take under 30 volts, but thats not the case, it's winter cold and casing+warming+insulating will solve problems, (I think).
 

Bandgap

Enlightened
Joined
Aug 14, 2002
Messages
284
Location
London
Now it needs to make some testing with 3 to 5 watt ceramic resistor, must make insulated aluminium box for batteries and NTC-powertransistor based simple thermalcontroller also have to test will that circuit burn my bike, if short circuited. Maybe it works for long 3 - 5 h bicycling sessions.

I have thought about this too, but have ground to a halt each time.
- AFAIK it is hard to tell when the inside of the cell is up to temperature, unless you make some assumption about heating time and initial temperature.

Also AFAIK, the Applied Physics Laboratory of Johns Hopkins University has only just found a way to imply the internal temperature electronically, and their technique is pretty complex.

Steve
 

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
I have thought about this too, but have ground to a halt each time.
- AFAIK it is hard to tell when the inside of the cell is up to temperature, unless you make some assumption about heating time and initial temperature.

Also AFAIK, the Applied Physics Laboratory of Johns Hopkins University has only just found a way to imply the internal temperature electronically, and their technique is pretty complex.

Steve

uh, it was not rocket science, I tried insulated battery "lunch_box", and it helped. Now I have NTC resistor inside battery pack so I see better what insulation do and yes, cell temperature, car batteries have some kind build_in heater between cell laminations.

(need to find boost led driver any good ideas...)
 

Bandgap

Enlightened
Joined
Aug 14, 2002
Messages
284
Location
London
uh, it was not rocket science, I tried insulated battery "lunch_box", and it helped. Now I have NTC resistor inside battery pack so I see better what insulation do and yes, cell temperature, car batteries have some kind build_in heater between cell laminations.

You made me think Traum.

Two flat-style Li-ion cells wired in parallel could be sandwiched around a thin flat heater - maybe a meander-line of resistance wire.

A thermistor on the outside of one of them, would be able to indicate that the cells had warmed right through.

Thanks

Steve
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
just sticking my nose in here quick... are the batteries in the same housing as the LEDs? If so, it would be handy to just use the LED's waste heat to keep the batteries warm. A bit of insulation around the whole assembly could aid in keeping the batteries from getting too cold too. I have a vision of a tiny parka wrapped around the light, with a fur trimmed opening for the LEDs. :)

As far as measuring temperature, you'd definitely want to place the sensor in the middle of the battery pack somehow. Or maybe at the outside, in order to be sure that you sensed the temperature of the coldest part of the battery?? Having the battery pack well insulated would at least ensure that there wasn't much of a thermal gradient across the battery.

Interesting problem, but it does make you wonder if life wouldn't be simpler if you changed to supercaps instead of batteries, eh?

Steve K.
 

Traum

Newly Enlightened
Joined
Jan 11, 2012
Messages
10
just sticking my nose in here quick... are the batteries in the same housing as the LEDs? If so, it would be handy to just use the LED's waste heat to keep the batteries warm. A bit of insulation around the whole assembly could aid in keeping the batteries from getting too cold too. I have a vision of a tiny parka wrapped around the light, with a fur trimmed opening for the LEDs. :)

As far as measuring temperature, you'd definitely want to place the sensor in the middle of the battery pack somehow. Or maybe at the outside, in order to be sure that you sensed the temperature of the coldest part of the battery?? Having the battery pack well insulated would at least ensure that there wasn't much of a thermal gradient across the battery.

Interesting problem, but it does make you wonder if life wouldn't be simpler if you changed to supercaps instead of batteries, eh?

Steve K.
Batteries are in panniers, so cant use leds waste heat (sad). I had one battery light and it dropped always cause it was too heavy (batteries was in same casing with leds), then decision to move on with dynohub-rechargeable cells in rear panniers was (word missing) cause some idiot stole that light, I must thank that person.

Life is easy, soldergun is not warming, refrigeration unit warms up (inside) and air stays outside the rear tube (they call it "flatout?")... all has broken on same day, but batteryback worked better. Maybe I try Supercaps then when I solved NiMH and LiFepo cells problem or give up and if I want pick easy route I will pick buss-ticket, sixpack and some nachos... :thumbsup: so no free lunch, no easy route just different and after all if its broken-not_working but fixable, fix it or get new one
 
Last edited:
Top