LC circuit resonance?

[Eamon]

I can't find the inductance of a Schmidt hub anywhere. Does anyone know it? I'm trying to estimate resonant frequencies of dynamo lighting circuits, and I can't find that value.
Alternately, would anyone who knows how to calculate those fequencies like to share the method? Or a good source for the math?

Thx,
Eamon

 

[Steve K]

I'm pretty sure that Andreas Oehler would have the answer. He can be found at Bikecurrent, which is one of the lists on Topica.com

I'm curious as to why you'd be interested in resonant frequencies. I can understand wanting to match the load impedance to the impedance of the dynamo, but that's a different thing.

In the past, I'd raised the issue of load matching on the Bikecurrent list. Andreas didn't think it was a very practical idea. That's probably true if you are developing products for mass production. Martin's circuit does a bit of load matching, and ends up using some relatively large aluminum electrolytic capacitors.

My personal opinion is that it is more effective to adjust the load resistance with the dynamo speed if the goal is to extract the maximum power from the dynamo. There is some data that helps find the best resistances.

regards,
Steve K.

 

[Eamon]

Martin and KTonik both use capacitors to maximixe circuit current at particular speeds, and exceed the dynamo rating at those speeds. Martin said that

"The capacitor value depends a lot on the generator and the frequency where you want to have the resonance. When working with hub dynamos, the frequency is low and bipolar capacitors get really big and expensive".

If I'm reading correctly, the current in a series LC circuit can exceed the source rating at that resonant frequency. That certaily sounds like what KTronik and Martin are doing. If so, then they're using the dynamo itself as the inductor, and the capacitor ratings follow.

That's why I want the inductance value for the SON.

The basic calculation for that resonant freq is pretty simple, but all the examples I can find are classroom-limited. They're not helpful to dilettantes who want to do nonstandard things.

Eamon

 

[Steve K]

What they describe is matching the load impedance to the impedance of the power source. This is a subject that you are introduced to in the first year of EE courses. Very very basic.
Nothing to do with resonance.

The inductance is the inductance of the dynamo windings. This can be measured in various ways, both direct and indirect. The most effective is probably to pick a dynamo frequency of interest, a suitable load resistance, and adjust capacitance until you get the results you want. I'm assuming that the desired result is maximum current at the chosen frequency.

I'm sure you know this already, but when using a capacitive load in this way, you only maximize the delivered power at one speed. If you use frequency-dependant loads, then you can optimize at multiple speeds. People have put together circuits that switch in different impedances as the dynamo speed changes.

To get a more direct value of inductance, you can contact Andreas Oehler. He's one of the main engineers at Schmidt, and can be considered an expert in this area of knowledge.

regards,
Steve K.

 

[FrontRanger]

EDIT: This was supposed to be post #3, but I forgot to hit the "submit reply" button for a while and by then it was post #5.

I'm curious as to why you'd be interested in resonant frequencies. I can understand wanting to match the load impedance to the impedance of the dynamo, but that's a different thing.

That is probably what the OP means, as opposed to the possible self-resonances of the dynamo.

Martin's circuit does a bit of load matching...

Yes, Martin's "boost" capacitor does exactly that, but only in a certain frequency band. The hub's source impedance is dominated by series-RL elements. Martin's extra series capacitor provides a narrowband complex conjugate match [1]. The values that Martin lists for these series caps in his Circuits 5 and 6 place the resonant frequency in the region where the source impedance is transitioning from being RL-dominated to L-dominated. I'm too busy to work it out at the moment, but the OP could estimate the source inductance for the Shimano DH-3*71 by looking at Martin's output vs. speed graphs, his series C values, knowing the number of poles in the hub and the wheel size Martin uses, and knowing that resonance occurs at 1/sqrt(L*C) in a simple (lossless, two-element) LC resonator.

... and ends up using some relatively large aluminum electrolytic capacitors.

That's the price. On the 2-LED implementation I built, compared to just the rectifier and filter cap, it more than doubled the total volume of the circuit elements. I used them. My justification was that I occasionally have to ride slowly on multi-use paths. (But really, it's because reading CPF makes me want to squeeze out more lumens.)

[1] Some might argue about my choice of words here because the Q is not that high, but I say "narrowband" because it doesn't span the entire operating frequency range with similar magnitude.

 

[FrontRanger]

To get a more direct value of inductance, you can contact Andreas Oehler. He's one of the main engineers at Schmidt, and can be considered an expert in this area of knowledge.

In that case, that is surely the best way to find the answer!

 

[Eamon]

Y'all are better men than I. People like me drive me mad in my own expertise, and I'm not always as patient as I ought.

What the OP wants is a way to calculate where those current peaks will fall in the speed range, and choose parts accordingly. And Steve is quite right. KTronik's current measurements above 900ma gave me most furiously to think, and I was thinking about the possiblity of peaking the current across a wider speed range. Apparently people have used dynamo frequency to change impedance before. I'm curious how they did that.

What the OP needs is to read enough electronics theory to work it out himself and stop asking stupid questions. And to contact Andreas Oehler. With better questions.


Thanks for the patience,
Eamon

 

[Steve K]

Like a lot of things, there's not always an easy answer. If this was a simple thing, you'd find these in production already.

Depending on your local environment and your riding practices, aluminum electrolytic caps may not be a good choice. A lot of my riding is in the winter with painfully cold temperatures. Aluminum electrolytics can show significant reduction in capacitance and increased resistance in these temperatures, so I'm reluctant to use them. There's also issues of reliability too.

There are different ways to detect dynamo frequency and switch circuit elements based on the freq. I think Martin used a frequency detector. I've got a circuit that uses a "missing pulse detector" scheme (an old idea) in development, and used previously by a friend.

Andreas may well know the inductance, but I know that he doesn't favor this sort of idea. As such, he may not bother keeping track of what the inductance is. I think the better idea is to try different cap values and see what the effect is. This will have the benefit of including all of the parasitic effects, such as capacitor resistance or dynamo resistance. It's not that hard to ride a bike with a multimeter attached to the handlebar. I've done it downhill at 50mph.

good luck,
Steve K.