Matching to the impedance of an alternator.

WHERE ARE THE PARTS VALUES?

John

Except an actual circuit. With values.

John

I guess you'd argue I hand drew all those curves ?:-)

It's nothing but an ideal buck switcher, a current source (300mA RMS @120Hz) and a 15 Ohm load (bulb)... with 6V across it, thus 400mA, so it has the requisite current gain.

From Marcel's kind German note translation I'll soon have an accurate Spice model for the dynamo, but parameterized, so users can make measurements and model their own dynamo.

When I'm happy with my behavioral modeling I'll post basically an AppNote with Spice models (the way I do things, it'll run on most Spice variants).

Values? Thanks for your efforts at displaying your misery, keeping my Schadenfreude flowing... Damn! I almost wet my pants from laughing so hard ;-) ...Jim Thompson

Stick your head up your ass and you'll see them... by LED illumination ;-) ...Jim Thompson

No, that would require thinking and understanding stuff. Futzing Spice is easier.

I suppose Vin just increases forever... but its slope doesn't look quite right to me.

What sort of mosfets did you have in mind, for infinite Vin?

Old people tend to do that.

Are you going to actually design a circuit, or just stop with an ideal behavioral model?

John

Sounds like no circuit, no values. How do you mount an ideal behavioral model on a bicycle?

John

Good question. The last time I saw a bike go by me with what could look like a model,on it, didn't behave very well at all, in fact, the tattoos really lost her some brownie points. As for as mounting them on the bike, I think they already know how to mount themselves!

Jamie

Sheeesh! If Larkin doesn't know how to "mount" a model he must be a fairy ;-) ...Jim Thompson

Hi Marcel,

Puzzling over the unloaded voltage curve... is that actually the implied voltage before the L, RFE attenuation? Otherwise you're blowing 0.5A in RFE. ...Jim Thompson

magnetic

does not

current) for

are

and

remember

operating

the

^^

That is an insupportable IF. The models resulting from test data = reported in this thread do not support the near quadrature phase angle nor the impedances necessary.

Look at the data Marcel posted, the graphs of current versus speed. There's an initial rampup region where the system is ohmic. Then the current flattens out: that's the region where the reactance dominates, and it's very wide. It's the reactance that allows the lights to work over a large speed range without blowing out.

The inflection point, where the slopey region intersects with the flat part, is the 45 degree point, which happens at a relatively low frequency. As frequency goes up from there, the current asymptotically approaches 90 degrees, relative to the excitation (open circuit voltage) phase angle. The closer to 90 degrees the current, the less real power is dissipated and the less shaft torque is needed to create it. Once the current is constant, the copper loss is constant, too.

John

In days of yore, lights were incandescent, batteries were carbon-zinc, and the only way to get reliable long-term, no hassle lighting was to use a hub dynamo. Now batteries are Li-Ion, lights are LED with buck-pucks, and for what you would pay for a bare dynamo hub, you can buy battery powered lights that will peel paint. However, for ultimate independence from stationary AC, dynamo hubs are still the way to go.

Further development of hub dynamos has actually been hampered by old German regulations, which standardized nominal voltage and wattage. It is nearly impossible to buy a hub dynamo that is rated above 3W.

With more cyclists packing cell phones, GPSs and other electrickery (as well as electric shifting), there would seem to be a niche for a dynamo/battery/lighting/intelligent electronics system that will take care of everything.

YOU are in no position to ask that, Mr. no values topology.

(experiment

Output

the

cycling

good=20

the=20

the=20

to=20

Cribbing from JT's and JF's posts now are you? Sad.

a 3W

there's 3=20

=20

And how much of that is measurement problems related to where they are placed? And how much from unit to unit calibration (if any)?

?-)

The hens are always ragging me because I post some full production schematics and some topological sketch ideas. Now JT is posting ideal behavioral models, not even circuits.

Let's see how he does.

Hey, why don't you take a shot at it?

John

Check the history.

John

Yes, I observed that also. The output is not a pure sinusoid, was going to take a picture of it. Not to say the scope is in cal though.

Resistance loading was next, but had a slight mishap ; )

No matter, a link to more representative data was posted,

The way I see it, it's the voltage measured at the terminals of the dynamo, if you are referring to page 5.

The upper graph on page 10 shows the

• calculated voltage before L, Rfe attenuation (Uq),

• measured open circuit voltage at the dynamo terminals (Messwerte)

• calculated open circuit voltage at the dynamo terminals (Modell)