Adapting battery packs from Cree T6 light to USB charger

[DoctorRad]

The voltage sag will basically be due to an impedance matching issue between the dynamo and the regulator. When it's trying to deliver a high current (>=1A) the regulator will show a low input impedance to the dynamo, which in turn simply cannot supply enough current to prevent voltage sag at the output of the regulator.

Better impedance matching can be provided by what is often referred to as a tuning capacitor. This acts to provide a more optimal input impedance to the relatively inductive output impedance of the dynamo. For maximum AC power transfer, the (complex) input impedance of the load circuit should be the complex conjugate of the output impedance of the generator at the frequency of power being generated. The issue with a dynamo is that the output frequency varies with speed, so it is not possible to correctly match the output impedance at all speeds unless someone invents a very clever frequency dependent capacitor.

What is therefore usually done is to choose a capacitor value which impedance matches well at low speed where the dynamo needs more help in getting its output to the world at large. Some example circuits and component values for driving LEDs are given here:

http://www.pilom.com/BicycleElectronics ... .htm#Boost

YMMV for driving a regulator, but it's a good starting point. You'll either need a bipolar capacitor or back-to-back electrolytics, but that's all covered in the web page above.

I'm waiting on our house extension being finished when I will get some significant desk space so I can start building this stuff for real.

[edocaster]

Thanks for the reply. I've used the Pilom tuning capacitor before (although not for this circuit), but without delving into the theory. I don't think I'm quite ready for imaginary numbers and all that stuff!

I didn't use the Pilom tuning capacitor on this circuit as I thought it only really came into its own when operating the dynamo at higher voltages (12V, 15V, etc), where people have used a conservative tuning capacitor to obtain decent current at lower speeds. Conversely, some use 'too small' capacitors to obtain a peaky curve, with a lot more than 550mA at the peak. The effect seems to be primarily on current.

But if a tuning capacitor can help prevent voltage sag at lower voltages, it may be worth investigating here too.

In practice, the circuit seems to be able to charge a phone and USB power pack quite well, so I'll probably leave this incarnation as is.

[RichardPH]

Coming a bit late to this discussion, but it seems to embody the cutting edge of the forum knowledge, certainly way beyond my own. I'm putting together a package to free me from the tyranny of having to find mains for my USB devices, common issue.

So far I have a wheel and Shimano DH-2N70 and have also ordered four of these from AliExpress, intent to package into a four USB output box.

https://www.aliexpress.com/item/DC-6-24 ... 0.0.p4X3wk

The output from the alternator needs to converted from AC to DC, and some overvoltage protection embodied. So far I've grown to understand that a full wave rectifier bridge and Zener diode plus a load resistor may be sufficient, but the Shimano based circuit discussed in this thread promises a better result.

Short question first, how did it go in practice?

Please bear in mind that whilst I'm a qualified mechanical engineer, electronics are not usually my thing, so you'll not insult me by spelling it out in very simple terms.

[busb]

I find this rechargeable battery resource useful. This link is specific to Li Ion cells:
http://batteryuniversity.com/learn/arti ... _batteries
Other pages cover other types. As a general rule, charging cells in parallel is not ideal, especially types that require constant current.