Bike HUB motor (as generator):why a power output difference between LED array and Supercap charger

[Sean D]

Hello,

A newbie with a question here, please be gentle.

I am tryng to understand why I get such different power outputs when I swap between an array of LED's at 12V, vs no LED array connected, but charging a supercap at 2.5v DC.

The rough circuit is below.

I am using a hub motor (250watt, 24v) as a hand crank generator. The hub motor has internal planetary gearing so it's spinning about 300 rpm.

I rectify the three phase output and I can see with an inline power meter that I get between 0 and 24v depending how vigorous I am feeling. The smooth-ish DC goes to a step down converter board which gives me 12v, which I can see with another meter (not shown in pic).

When I connect only a 14watt 12v array of LED's I can get amazing output, in fact I can easily output 20 to 30 watts, way too much for the array, and about the limit that little DC converter (it maxes out at 3A)

If I then disconnect the LED's and use another DC -DC point of load converter to drop the 12v to a stable 2.5 volts out ( which I see with another meter - not shown in pic) then the generator spins freely and I am only able to output about 6 watts. It varies a little depending on how charged the supercap is, but I thought the supercap acts like a dead short when empty (so I manualy add a few ohm resistor, then remove it at around .5v) and up to about 13watts when I have about 1.3v in the cap.

Q1: For my understanding. What do you think is the reason for that? Why would the supercap not appear as an equivalent load?

Q2: Advice: I want to charge the cap as quickly as possible using my vigour while it lasts.. Is there a better way to dump all this power into a single 2.7v cap?


Many thanks

 

[pgib8]

Well, you say power output, but maybe you should think of it as power-draw by the components.
When the crank has very little physical load, it means no or little power is being drawn.
The capacitor probably charges up in no-time. So you might think you get little output but really, you have way more output than what's being used up.
Using the resistor in series with the cap will probably make for a smoother ride for the DC-DC converter(s) as you are limiting how much current can be drawn.

So let's do a little math, the 1st DC/DC can do 3A. Let me pretend it's 2A for extra margin.

Your max voltage for the capacitor is 2.5V.
So 2.5V / 2A = 1.25 Ohm. (minimum 5 Watt rated resistor)

I don't know much about supercaps but I would imagine it would charge up pretty quickly, I mean how big is this thing?

 

[Sean D]

Hello pgib8, thank you for the reply.

Well, you say power output, but maybe you should think of it as power-draw by the components.
When the crank has very little physical load, it means no or little power is being drawn.
The capacitor probably charges up in no-time.
.
.I don't know much about supercaps but I would imagine it would charge up pretty quickly, I mean how big is this thing?

The supercap is quite large, it is 400Farad rated at 2.7v (2.85v surge) so when it is charged it contains 1458 Joules of energy (Energy = ½*C*V² 0.5*400*2.7² )

Power in Watts = Energy in Joules/Time, so if I want to charge it in 60 seconds I need a continuous 24.3 watts of power (1458J/60sec) at 12V it's 2.025 amps at the first DC-DC converter output, but then so I need to be pumping in 9 Amps at 2.7 volts to get that power from the final stage to the supercap. (which is within spec for that dc-dc converter and the cap)

If I use a dc power brick instead of the hand crank generator the POL DC-DC converter can send though that sort of power, so I suspect it is because the input is repeatedly dropping below whatever the final POL DC-DC converter can work with and I can't see that with a multimeter, or more likly, I just don't understand enough about the mysteries of analog circuits with all those inductances, reactances, impedances.

Thanks

 

[pgib8]

ah i see. you know quite a lot about the super cap and the circuit. you may also be onto something. you said the dc is smooth-ish. Without an oscilloscope it may be hard to tell. Can you try your meter on the AC setting? Maybe it will give you some info.
I would suspect that both your DC/DC converters already have input and output capacitors. It wouldn't hurt to add some extras. I would particularly add a capacitor before your first DC/DC converter. What you said is quite possible. The voltage briefly drops too low and then the DC/DC converter needs some time to get back to work. If this happens continuously, then your dc/dc converter may spend most of its time "resetting".
I'm not sure you need to worry about impedance since it's based on DC. There is a slight possibility though that you are having trouble between the generator and the first dc/dc converter, with reactive power. Here again, a capacitor of sufficient size would compensate for that.

 

[Sean D]

pgib8,

Thanks again. I'll try a few addtional capacitors and see.

The diode rectifier section has a 50v 1000uF cap on the output, hopefully smoothing the DC a bit.
I've just googled "reactive power".. and now I have a bit more reading to be done it seems.

For other capacitors on the inputs of sufficent size: I guess it is to keep the voltage high during the variations in cranking, but are you meaning largish electrolytics, such as a 1000uF, or 10000uF ? or something else?
I have seen tantalum 100uF caps on the inputs of these POL DC-DC convereters and sometimes a small inductor (1uH) in series with the positive input, but I am not sure of the reasons behind it, maybe ripple reduction?

 

[Sean D]

I tried the multimeter on the AC setting on the output of the diode bridge and see a varying result between 0 and 3V and and about 0.1V on the output of the first dc-dc converter, But both seem to spike up to 10V when I first start cranking and then settle down.

 

[Sean D]

Some more very interesting information.

Brief version : I replaced the super capacitor with a 25watt 0.5Ohm resistor and I am able to produce 25-35 watts of power while cranking. My understanding of an empty supercapacitor is that it is like a short when empty and accepts as much current as it can until it fills up to the input voltage and the current drops off to zero when that happens.


Longer Version:
I tried adding a 22000uF 16V cap to the output of the first dc-dc converter.
It certainly allows the 12V out to be more stable.. great, so I thought.

When cranking while the super capacitor was under 1.5 volts then it made things worse, while cranking I'd attached and remove the capacitor and I could immediately feel the differece, cap connected = free spinning low power, cap off = harder to crank, more power.

When the super cap is over 2 volts or so, the 22000uF cap seems to improve things, cap connected = harder to crank, more power, cap off = free spinning, low power.

I futzed around for a bit and decided to swap out the supercap for a 25W 0.5Ohm restistor. Surely enough, once I do that , the cranking is hard and I can produce 25-35 watts, and the dc-dc pol output is around 2.2V while doing this, so all good. I wonder why the difference.

 

[pgib8]

Ok next thought: We haven't talked much about the generator. I don't know too much about generators but I think some (maybe most) need reactive power. It probably has something to do with the fact that they don't have a permanent magnet but rely on some current going through the stator to produce the magnetic field. Can you try adding 3 capacitors between the 3 phases, from phase to phase (delta) or perhaps all from phase to a central point (star)?
I don't know about what sizes to use, maybe just try something. I also read somewhere that you may need to start a generator without load at first, so that it can build up a voltage (presumably for the stator's magnetic field).

I almost think you ought to post this question in a forum for electric generation and distribution (power grids). That's exactly what you're dealing with, just on a small scale.

 

[pgib8]

Oh and a small conductor in series to me sounds like to reduce current spikes. Inductors like to keep the current going through them the same and induce a voltage to oppose any change.
Capacitors pretty much do the opposite, they like to keep the voltage on them the same and will induce a current to counteract any voltage changes.

 

[pgib8]

One more thing, just did some reading about it. This may be your missing link, pun intended.

http://www.electway.net/news/difference_between_Asynchronous_and_synchronous_generator.html

"However, an asynchronous generator in a stand alone system can be used if it is provided with capacitors which supply the necessary magnetisation current."

 

[Sean D]

Thanks again! I will be able to do more testing tomorrow, but will read up on it tonight.
One thing I noticed is that the length of the cables/type that I am using between dc-dc pol and super capacitor seem to make a difference, the longest bit is about 70cm, and beefy dc high wire strand stuff, oddly the shorter test leads seem to allow more current to flow. perhaps I have a dodgy cable.. I will check as well.

Thanks for all your help.

 

[Sean D]

One last thing before I get some sleep.. The motor/generator is a brushless DC type, so it does have permanent magnets in there.

 

[pgib8]

well then i'm out of ideas and you're welcome.

 

[Sean D]

Hi,

Just as a final update, I found that if I have a direct cable connection, and by that I mean very low resistance from the output of the POL DC-DC converter to the supercap then it only would allow me to crank out a few watts. but if I connected the supercap via a thin testlead, the increased resistance of the lead was enough to allow me to pump out 25 or more watts in to that supercap. I really dont know why that tiny increase in resistance was causing such a large effect, but to confirm I put in a 25watt 0.1 Ohm resistor in series with the beefy DC cable and that seems to have done the trick. This is not an elegant solution at all, as the resistor certainly wastes power by getting hot when charging the supercap, but it is at least working much better than it was before. If you can think of another way to achieve this resistance, or the effect that the resistance causes, I am all ears.

Thanks for all your help on this, I do appreciate it. The additional cap you suggested makes it more stable, so that is a good thing, and the new reading material about reactive power and async/sync generators too.

Sean