Probably Fried The Charging Circuit

[arzgi]

I've used my Zaurus 3100 as a gps device on long bicycle trips.

The gps-card I have (Billionton) sucks battery quite soon empty, so
decided to do a travel charger. My bike's hub dynamo gives 6 V / 2,4 W,
and thus 0,4 amperes, as I learned at school many years ago.

Added a diodebridge to smoothen voltage to dc, and a 5 V regulator
to ensure voltage won't get too high.

Then rode a test trip voltage meter mounted on handle bar. Voltage
newer rised over 4,7 volts even I pedaled 50 km/h (31 mph). I almost
never exceed that speed.

I thought it was safe to use the system to charge zaurus.

Soon on first trip I almost fell, fastening strap twisted connector
quite strongly. After that I could not charge even using the regular
plug charger, led wont lit up, bottom gets hot under power connector.
Took Zaurus to be repaired, has been there almost 10 weeks, not
very promising. First they said it can be fixed, last time a called
they were still waiting for spare parts.

After that secured the fastening belts, triple checked polarity.
Did a test drive at very low speed, but seems I've burn't also
my second Z's charging circuit. Sometimes yellow led lits, usually
just blinks when charging with plug charger.

Does not hamper my usual usage, I have an external charger.

Seems it's time to forget dynamo-charger, but I wonder what wen't
wrong. The plug charger gives 5 V and 1 A (readings on sticker, I also
measured 5 volts).

I newer measured the amperes of my bike charger.
I've seen graphs of hub-dynamo's speed vs. voltage curve, and it is quite
even. So thought it would never exeed 1 A.

Can someone who has deeper knowledge of electronics tell what was wrong?

[gromituk]

Can someone who has deeper knowledge of electronics tell what was wrong?

A bicycle dynamo is a current source rather than a voltage source.  Its current is proportional to its speed, rather than its voltage.  This means that as its speed rises, it will want to push out more and more current (until it self-limits due to its impedance increasing with frequency, as it is inductive), and will produce more and more volts to do this.  You therefore must ensure that the current it wants to produce is always absorbed.  Unless your Zaurus was trying to take all the current that was available, the voltage at the input side of the voltage regulator could rise to a level at which the voltage regulator would fail, and these things tend to fail short circuit, dumping that high voltage straight into the Zaurus.  What you should have installed was a shunt regulator such as a zener diode of, say, 10V across the input to the regulator (which would have to be quite a high powered zener to dissipate 3W).  Some bicycle lamps have two back-to-back zener diodes in them to prevent blowing the bulb at high speed.

Did you fit a smoothing capacitor before the regulator?  If not, you meter may have misread completely because it was trying to measure a relatively high frequency pulsating DC, rather than a smooth DC.  Who knows how the voltage regulator might have performed under those circumstances, too.  And it needs a non-electrolytic capacitor on input and output to prevent oscillations.

[arzgi]

Thanks for a quick and very thorough reply gromituk!

Yepp, I considered zener-diodes, but I've read many succes stories of using dynamos
to charge cell phones and laptops even by just 4 diodes. Perhaps zaurus is more picky,
have to try that, thanks.

I had a halogen bicycle light more than 10 years ago, which had two zeners.
I now use a regular light bulb, no zeners. After two years of hub dynamo usage
only one lamp has burned out. That was quite normal on bottle dynamo too. Burning
did not happen at high speed. And reason why I still use this old lamp is it's great
reflector. I have heard people thought there was a car coming, when they saw
my bicyle light  

About smoothing capacitor, no I did not use any. I have an impression that ic-diodebridges
do contain a smoothing capacitor. But as I have said, my knowledge is thin.
Have to try that too, I probably have suitable one in my component box.

Almost every time I've soldered during past years (only a few times) I do miss an oscillocope at
some point.  A 5 euros digital multimeter is quite limited.  Luckily this time found digital
arm dso oscillocopes which do not cost more than a few dozen euros/dollars. Ordered one
instantly.

And lastly good news: I erased nand and reflashed it, and now charging works again.
Already charged an empty battery to full, no problems. Did not try that on my zaurus
I took to be repaired, I need to visit there soon.

[cybersphinx]

Perhaps zaurus is more picky,
have to try that, thanks.

Reportedly the Zaurus is very picky regarding its input voltage.

[rlubikey]

Bad luck having to go through all this grief. At least you tried but, as gromituk says, a bike dynamo is a complex source of power. You would most definitely have needed a smoothing capacitor, quite a large value, to supplement your rectifier pack as the frequency would be quite low at low speeds. Especially so as you have a hub dynamo!

The fact that you measured up to 4.7V on your DVM is consistent with the regulated 5V being periodically "interrupted" as the rectified, un-smoothed voltage fell to zero every half-cycle. I should think it was this that confused the charging circuit and brought about the failure.

I seem to recall that the Z needs a 1-amp power supply - perhaps someone can confirm this. Even so, all bike dynamos I have ever seen only produce 0.5-amps. (Even the 6-watt hub dynamo my Nearest & Dearest has on her bike is 12V @ 0.5A) You *could* make a suitable power supply using a switch-mode regulator to translate (in excess of) 10V @ 0.5A down to 5V @ 1A when you're pedaling fast enough, but I suspect this is quite a challenging project.

Richard

[arzgi]

Bad luck having to go through all this grief. At least you tried but, as gromituk says, a bike dynamo is a complex source of power. You would most definitely have needed a smoothing capacitor, quite a large value, to supplement your rectifier pack as the frequency would be quite low at low speeds. Especially so as you have a hub dynamo!

The fact that you measured up to 4.7V on your DVM is consistent with the regulated 5V being periodically "interrupted" as the rectified, un-smoothed voltage fell to zero every half-cycle. I should think it was this that confused the charging circuit and brought about the failure.

I seem to recall that the Z needs a 1-amp power supply - perhaps someone can confirm this. Even so, all bike dynamos I have ever seen only produce 0.5-amps. (Even the 6-watt hub dynamo my Nearest & Dearest has on her bike is 12V @ 0.5A) You *could* make a suitable power supply using a switch-mode regulator to translate (in excess of) 10V @ 0.5A down to 5V @ 1A when you're pedaling fast enough, but I suspect this is quite a challenging project.

Richard

Yepp, thanks to gromituk seems to work now. I made some modifications as gromituk kindly suggested and  Zaurus accepts now the hub dynamo charger (yellow charging led lits and does not hamper charging by the plug charger). I have not tested on the road yet, it's been very rainy here past few days, and the water shield I've planned is still in proggress.

[arzgi]

Road test done, works perfectly, thanks again gromituk!

I tried with totally empty battery. Took about two minutes before charging led lit, but after that no problems at all.
Led starts to blink when speed is 4 km/h (2,5 mph), but already at walking speed 7 km/h (4,35 mph) stays on
constantly. Max speed I tried was 48 km/h (29,8 mph), still nothing to worry, yellow led still shines. At home plugged zaurus to its own charger, and charging continues.

I also tried the head light at the same time, but as I suspected, the light was very dim, and zaurus stopped charging.

Goal reached, but I wonder if it is even possible to use those two at the same time without second dynamo at
rear wheel? Bulb "eats" 2,4 watts, so no much left to zaurus. Perhaps then, if I change to led lamp?

[rlubikey]

Excellent! Which hub dynamo do you have and what value capacitor did you use? Do you measure 5V now, instead of the 4.7V you had before?

For headlamp operation, LED headlamps are *very* good at low speed. One of these may give you enough light when your Z is charging - no guarantees. But, as they give much more light than a halogen (x2 light over x2 area - I've measured my B&M IQ Fly) you may consider it a useful upgrade, even if Z-charging doesn't work

Richard

[arzgi]

Excellent! Which hub dynamo do you have and what value capacitor did you use? Do you measure 5V now, instead of the 4.7V you had before?

For headlamp operation, LED headlamps are *very* good at low speed. One of these may give you enough light when your Z is charging - no guarantees. But, as they give much more light than a halogen (x2 light over x2 area - I've measured my B&M IQ Fly) you may consider it a useful upgrade, even if Z-charging doesn't work

Richard

Hub dynamo is Shimano DH-2N30-E/HB-NX22 (what a code!), 6 V / 2.4 W.

Smoothing capacitor for the rectifier is 470 uF electrolytic. I don't know how many magnets there are in the hub,
and speed is variable anyhow, so could not calculate any precise value. Seems to do its' job.

Also added a 5 W 9.1 V zener diode as a shunt voltage stablizer. And 330 nF and 100 nF ceramic capacitors to
reduce ripple voltage of the regulator.

Measured voltage was always above 4.9 V. Something like 4.97 V at lower speeds, closer to 4.9 V when I
pedaled faster.

I forgot I have a battery led light too. Has been in my bicycle jacket's pocket since last winter. Idea behind that
was, that it is unpleasent to change burnt bulb on the road when temperature is below freezing. Has happened
only once since I installed hub dynamo though. Light was still bright, batteries ok    I can use that, if I need
gps and headlight at the same time.

Next head light I'll buy for the hub dynamo will definetely be a led light.