On 2017-03-14 19:43, Jeff Liebermann wrote:
On Tue, 14 Mar 2017 07:54:26 -0700, Joerg
wrote:
On 2017-03-13 20:00, Jeff Liebermann wrote:
On Mon, 13 Mar 2017 12:38:07 -0700, Joerg
wrote:
55V at 500mA. This is encouraging.
That's 27.5 watts out of a 3 watt dynamo. I was impressed, until I
converted 136 km/hr and found that it was 84.5 mph. With a rocket
assisted bicycle, I might be able to do that.
Well, yeah, they just wanted to see where the limit is. I guess the
enameled copper wire inside would smoke out if you kept that speed for long.
Only the resistive part dissipates power in the wi
P = I^2 * R = 0.5^2 * 2 = 0.5 watts
So, it won't be the wire that gets hot. However, the cores in
saturation are going to get warm. Offhand, I don't know how to
calculate how hot.
The lowest resistance I measured is on my largest bottle dynamo, 6.2
ohms. That added to the cores also becoming hot and not being able to
sink away much heat sets it up for a burn-out.
However, this means I should be comfortably able to milk 10W or more out
of such a dynamo on a long downhill stretch and maybe north of 5W during
regular rides on flat terrain. IOW there seems to be nothing that
inherently limits things to the usual measly 3-4W.
Maybe. Let's pretend that the test demonstrates that the dynamo can
deliver 28 watts for perhaps 60 minutes before blowing up. Assuming
constant energy (Joules or watt-seconds) at any speed.
Not any speed but the usual cruise speed of a rider in decent shape.
Also assuming MPPT because it really won't be possible without that. At
least a poor man's MPPT where the input voltage is set to a fixed higher
level such as 12V.
... That would be
10 watts for 168 minutes. Assuming 10 watts output is a comfortable
ride for you,
I think I can muster a little more than 10W :-)
10W wouldn't even show up in the noise for a regular rider.
... you would have about 3 hrs of bright light before the
dynamo self-destructed. Actually, it would probably be more like 4 or
5 hrs since the system is not adiabatic and the heat radiation
efficiency of the dynamo case is time dependent.
10W should be fine considering that the dynamo will then also be cooled
by air streaming past. On hot days I have to ratchet down the front
light anyhow when riding at low speed because else the overtemp shut-off
comes on. Then the net energy draw of the power bus becomes lower and
the charge controller reduced the dynamo draw to 4-5W.
The author is mostly correct about hubs not going into saturation as
easily and bottle dynamos. However, they do go into staturation, just
at a higher RPM.
If that happens above 84mph I shall be happy :-)
Dream on. If you look at the dynamo speed vs power curves at:
http://www.myra-simon.com/bike/dynotest.html
They all begin to go into saturation at 15 to 20 km/hr. At 136 km/hr,
the dynamo will produce 27.5 watts, but the rest of the input energy
will be wasted in core saturation heating. If you paint flames on the
fenders, maybe other riders will think the burning dynamo is normal.
There is no need for 27.5W but there is for 8-10W. That should be
reasonably possible, else the dynamo would have smoked out during their
testing.
http://www.ebay.com/itm/122373782338
(Yet another project that I'll never finish).
That is a neat little instrument.
Yep. I don't have one yet, but it's a real temptation to install one
on all my various unmetered power supplies and battery packs. However,
there's a catch. The common ground is positive (+), not negative.
You can see that in the schematic:
https://img.alicdn.com/imgextra/i3/121163002/TB2NoBogpXXXXahXpXXXXXXXXXX_!!121163002.jpg
where the + leads of both the "DC in" and the load are connected
together. The only ways I could make it work in a negative ground
system was either an isolated power source, isolated load, separate
isolated power supply to run the meter, or a DC-DC inverter.
That would be a show stopper in most applications. Got to have high-side
current sensing. From a dynamo it's still feasible after things have
been turned to DC because the hub versions usually have two ground-free
terminals. Bottle dynamos unfortunately not. However, when doing MPPT
there already is a micro controller which can then also perform the job
of energy metering. I would not need that though, maybe just an
indicator of whether the dynamo is able to feed enough for a given light
setting. Even that is expendable if you have a voltmeter riding along
with the Li-Ion battery. In my case I'd just have to make sure it won't
get close to 7V where the electronics will eventually shut things off.
Anyhow, I don't
think a dynamo can easily be simulated until you start with a mechanical
model of it and that gets into COMSOL and other really expensive
modeling software. Maybe possible with MathCad which I have but rarely
use so I became rusty there.
Yep. That was pretty much my conclusion. Jim Thomson posted this
Pspice component model:
https://groups.google.com/forum/#!topic/sci.electronics.design/yPVvclIIirQ
which I have been unable to convert to LTSpice. He seems to just put
a sharp knee where the dynamo starts to go into saturation. That's
probably good enough to get started. The rest is mechanical and
rather messy if I add nonlinearities, friction, and air resistance. I
do have one advantage with a hub dynamo. I can characterize it as a
motor on the bench and then simply invert the derived functions to
produce a generator. That might be a worthy shortcut.
Nothing beats trying it on the bench, using a variable speed drill or
similar. I haven't observed a sharp knee effect on any of them so far.
Which probably explains the large number of bulbs I blew out in the 80's
when I had to get somewhere fast.
--
Regards, Joerg
http://www.analogconsultants.com/