5V regulated power supply for charging phones, etc...

[syc]

I stumbled across this just now Low Dropout Rectifier and Linear Regulator.

It seems interesting because he's using a mosfet rectifier and a low dropout regulator to get circuit with only about 1 V drop, and it is quite tiny (aside from the capacitor!). There was some discussions on several cycling mailing lists related to a circuit for dynamos that would allow personal gadgets to be recharged via a USB port. In the photo he has a jack for a Nokia phone, but you could put a USB port on there instead.

 

[PeaceOfMind]

The MOSFET rectifier circuit technique is sometimes called Synchronous Rectification (ex http://en.wikipedia.org/wiki/Synchronous_rectification). At low power it is much more efficient, however at higher power it will begin to be less efficient than a diode rectification circuit. The point at which one becomes more efficient than the other depends on the individual diode's forward voltage drop/IV curve and the MOSFET's "on" resistance.

This is an excellent technique for low power devices and is often used in very low voltage power supplies.

Edit: I just read the link you posted and some of this is discussed there already so sorry for any duplication

 

[Steve K]

I've been using a mosfet rectifier for many years with my dynamo powered LED headlight. Very handy circuit, but it might be hard to justify the cost of the mosfets in a commercial application.

In this application, a schottky diode has to be added because the mosfet rectifier can't tolerate having a charged cap at the output when the dynamo is stopped. I think I would have just used schottky diodes in a conventional bridge rectifier, and accepted the additional 0.4v drop while saving the cost and space of the mosfets. (just a personal choice.. not saying either one is right or wrong)

This sort of design, i.e. a series regulator, has some difficulties when used with a dynamo. As the author notes, a dynamo can put out 100v when unloaded (I've verified this with my Schmidt). When the battery is fully charged, the dynamo is effectively unloaded, so the circuit has to protect itself.

One way to protect the circuit is to design it to tolerate voltages of 100v or more. i.e. use components rated for 100v. This gets expensive and the parts get large.

The other alternative, which the author uses, is add a shunt regulator at the input of the circuit. That's what the two 1N5343 zeners are for. They will clamp the dynamo output to about 8v or so. The downside is that means that the dynamo's output of 0.5A is clamped at 8v, which means that 4 watts are being dissipated.

So... the design now consists of a shunt regulator that provides coarse regulation, followed by a series regulator that provides a fairly precise regulation.

I've set up battery charging circuits for dynamos before, and I just designed a shunt regulator that could provide a fairly good regulation. This is what I did for charging a nicad battery:
http://phred.org/~alex/pictures/bikes/bikecurrent/bbcr_schem.pdf

To generate a good 5v output, I would suggest changing the zener to a precision voltage reference such as the LM4041.

The advantages of this circuit over the circuit with the LDO regulator are reduced cost, reduced board space, and reduced power dissipation.

It does surprise me a bit that this has become a popular subject again. Must be due to so many gadgets that can be charged from a USB port. It's nice to see that sort of standardization again.

regards,
Steve K.

 

[syc]

These tiny little sot23 schottky bridge rectifiers claim very low voltage drop: http://www.centralsemi.com/PDFs/products/CMPSH1-4L.pdf (.39V at 1A) which is what you would see if you used a mosfet rectifier with a single schottky diode so that you can put a capacitor inline.

Seems like these would be a better choice than the mosfet approach, at least for these sorts of dynamo applications.

 

[Steve K]

the datasheet for these sot23 diodes indicate a max Vf of 0.27v at 100mA. I've got a datasheet from On Semi for a 1N5819 (pretty generic leaded schottky) that shows a max Vf of 0.34v at 100mA. The difference is notable, but not huge.

In this application, I would also choose the schottky diode bridge rectifier. I do wonder about the comments of the designer of the charging circuit. He mentioned something about not having enough voltage to run the regulator?? Most of the dynamos will produce plenty of voltage at moderate speeds, if the load isn't a low resistance. I wonder if he was just trying to charge a nearly dead battery?

Steve K.

 

[n4zou]

Here is a much simpler way to provide filtered and regulated power for a USB device using a bicycle dynamo.

The 4 Ni-MH batteries regulate and filter power for the USB device. Voltage is regulated at 5.2 VDC due to internal impedance of the batteries. When dynamo voltage starts exceeding 4.8 volts the batteries internal impedance starts rising as well. At 5.2 volts the dynamo can no longer overcome this impedance so voltage is limited to 5.2 volts DC. 5.2 volts DC is well within USB voltage standards. 500mA is the maximum current requirement for USB devices. Few USB devices draw the maximum allowable current. The excess current is absorbed by the batteries and used for recharging. A word of caution is necessary. The batteries must have solder tabs and connections must solid. Poor or open connections between the batteries will allow unregulated voltage and current into the USB device possibility damaging it. The circuit shown is simplified. The LED headlight is a separate circuit and is not required for proper operation of the USB power circuit. It can be replaced by any commercially available dynamo headlight and/or taillight system or not installed at all. I use a connector between the batteries and switch so the batteries may be removed and recharged away from the bicycle. This allows powering a USB device without the dynamo engaged when desired. The batteries also power the USB device when stopped in traffic or taking a break.

 

[Steve K]

this concept has been around for a long time. It works for some, but I quickly killed a set of nicads this way. Shoving a half amp through AA cells for hours is a good way to damage the cells. Add in high temperatures, and it's worse. When you make the cells NiMH instead of nicad, it's worse yet. Nicads can tolerate trickle charging fairly well, but NiMH are much less tolerant.

regards,
Steve K.

 

[syc]

the datasheet for these sot23 diodes indicate a max Vf of 0.27v at 100mA. I've got a datasheet from On Semi for a 1N5819 (pretty generic leaded schottky) that shows a max Vf of 0.34v at 100mA. The difference is notable, but not huge.

I realize now that I've been misreading these spec sheets (plus, I sent out the wrong one - I meant to send this one http://www.centralsemi.com/pdf/CBRHDSH1-40L.pdf ). I didn't catch the "per diode" in the electrical characteristics table for the CBRHDSH1-40L, and thought that they had managed to halve the voltage drop (duh!!!)

Here is a much simpler way to provide filtered and regulated power for a USB device using a bicycle dynamo....

I remember this circuit from past postings. I have a question about it though - I was under the impression that you needed fairly large batteries in order to keep from overcharging/damaging them. If you needed, for example, 4 C cells, that is pretty large. If you need 4 D cells, that starts to get very clunky? What is the appropriate sized batteries for this? Are AA's sufficient?

 

[tiny68300]

What about tapping off of the battery that runs my lights - 14.8 v LIPO. I'd love to be able to recharge my phone when I not running the lights

 

[Steve K]

I remember this circuit from past postings. I have a question about it though - I was under the impression that you needed fairly large batteries in order to keep from overcharging/damaging them. If you needed, for example, 4 C cells, that is pretty large. If you need 4 D cells, that starts to get very clunky? What is the appropriate sized batteries for this? Are AA's sufficient?

Panasonic says that the trickle charge rate for NiMH should be a max of 0.033C to 0.05C, and that charging should last no more than 10 to 20 hours.
http://www.panasonic.com/industrial/battery/oem/images/pdf/Panasonic_NiMH_ChargeMethods.pdf
So... if the charge current is 0.5A, then the cells should have a capacity of 10Ahr or more. That sounds like a D cell at least.

This is why I added a voltage limit circuit for charging my cells. It probably won't fully charge the cells, but it won't damage them either.


Steve K.

 

[n4zou]

this concept has been around for a long time. It works for some, but I quickly killed a set of nicads this way. Shoving a half amp through AA cells for hours is a good way to damage the cells. Add in high temperatures, and it's worse. When you make the cells NiMH instead of nicad, it's worse yet. Nicads can tolerate trickle charging fairly well, but NiMH are much less tolerant.

regards,
Steve K.

You are not "Shoving" 500mA through the batteries unless you do not have a USB device plugged into the USB connector. My GPS unit requires 450mA with the back light turned on. The batteries are only sinking 50mA. I've been using this system for a couple of years now with no problems at all.

 

[Steve K]

You are not "Shoving" 500mA through the batteries unless you do not have a USB device plugged into the USB connector. My GPS unit requires 450mA with the back light turned on. The batteries are only sinking 50mA. I've been using this system for a couple of years now with no problems at all.

I shoved a half amp through my nicads and killed them. Anyone who uses the circuit shown without a significant external load runs the risk of doing likewise. Perhaps it would help others to include a note on the schematic specifying the need to have a large load attached.

Maybe my biggest issue was the claim that the batteries regulate the voltage, etc. If you had a 450mA load, then that was doing the most to pull down the dynamo output voltage.

regards,
Steve K.

 

[n4zou]

I shoved a half amp through my nicads and killed them. Anyone who uses the circuit shown without a significant external load runs the risk of doing likewise. Perhaps it would help others to include a note on the schematic specifying the need to have a large load attached.

Maybe my biggest issue was the claim that the batteries regulate the voltage, etc. If you had a 450mA load, then that was doing the most to pull down the dynamo output voltage.

regards,
Steve K.

Rechargeable batteries will regulate voltage output of a standard bicycle dynamo. When charging voltage exceeds the batteries operating voltage internal impedance of the battery increases proportionally. As the impedance increases the load on the dynamo also increases. When charging voltage reaches 5.2 volts the internal impedance of the batteries loads the dynamo to the point it simply cannot overcome the impedance load so higher voltage cannot be produced. The batteries are not actually regulating voltage in the normal sense of voltage control but the effect is the same none the less.

 

[pe2er]

Maybe my biggest issue was the claim that the batteries regulate the voltage, etc.

You can look at it as a shunt regulation, like placing Zener diodes parallel to the load.

Look at this simplified schematic: The Dynamo hub and rectifier are depicted as a single ½ Amp current source (with high impedance).

The 4 NiCd batteries are shown as a voltage source (with zero impedance) and a separate Ri (Internal resistance). Ri in this example is 0.2Ω. In real life, Ri will vary with the condition of the battery.

With no load connected to the battery/ Dynamo, ½ Amp flows into the battery, charging it.
The voltage at the terminals is 4.8V + (0.5A*0.2Ω = 0.1V) = 4.9V

With a 450mA load connected to the battery/ Dynamo, voltage will be less. Only 50mA will flow into the battery (500mA – 450mA).
The voltage at the terminals is 4.8V + (0.05A*0.2Ω = 0.01V) = 4.81V
So you see; voltage is 'regulated' to a voltage a little higher that the battery voltage.

Problem is the unloaded condition. Once the battery is full, the current flowing into the battery will be converted to heat. If and for how long the battery survives depends on the battery and the duration of your ride. So I Must agree with:

Anyone who uses the circuit shown without a significant external load runs the risk of doing likewise. (Killing the batteries)

Erwin

 

[Steve K]

I do understand shunt regulation, which is why I linked to a schematic of a shunt regulator. In fact, it's a shunt regulator which controls the charge voltage of a battery.

My concern is the use of the term "regulation", especially when it is revealed that a large external load is required to protect the batteries.

I have no doubt that the circuit does work well for N4zou. My only concern is that someone will think that n4zou's circuit can be used as shown and can be relied on to not damage the battery or any arbitrary device plugged into the USB port.

regards,
Steve K.