Wheel hub dynamo delima

[jkrienert]

Thanks Flux.
The dual coils are perminet in their distance from one another, but the other coil does have some marginal movment.
I have tried it with a little to the left-and a little to the right- but my meter really doesnt get specific enough to clarify.
What I did gather, was a couple photos to show where it is at currently.
Im not an ace photographer. Just so you know...



The output of this is suprising.  It seems like its more than it should be.  If it is worth someones time - I hope I can share this.

If you celebrate it, Thanks<im>giving!

[Bruce S]

Free capacitors?
What planet are you from?!
Do you welcome earthlings?

The LED I am planing on are
        (x1-2) 3.7v at 700mah max [front]
        (x1) 3v at 500mah max [rear]

Of course, these are subject to change, as I found a source for CREE emitters for really good prices nearby.

but please, Gurd if you can - wait for a road-stop
cheers.

My BAD. I was speaking of common Caps.
I am from earth  , but there were times !!

Since this is going to be for a bike 2-wheeled? right? I would lean towards the NiCd/NiMh. The NiMh would be easier to purchase in a higher density for lower price, unless you are getting NiCds free  .
The charging is similar enough that and the discharge rate is similar enough that you really shouldn't worry too much about them.
A few other questions until GHURD chimes in with his vast knowledge  .
Are you wanting to light these at full brightness? 700mA is going to be a lot and could shorten the lifespan of these , though not by much.
the difference between say 600mA and 700 may not be noticeable amount of light output. Even then it is still over 1A at 3.7Vdc.
You could use a few different resistors for the front and since a tail light only needs to be seen, you can drop its rating to 1/2 and still be seen easily.
Going one further, since the back is gonna be red, go for the red LED and cut the mA rating in 1/2 too.
More to come once you have the charging setup values known.

Bruce S

 

[jkrienert]

im starting to have some growing concerns about my rectification of each coil, then combining them in parrallel Dc output.
From what I have been reading, DC outputs when combined, must be identical-or there is a possibilty for some major problems.
Am I avoiding these problems with the shunting of the diodes in this diagram?

[TomW]

From what I have been reading, DC outputs when combined, must be identical-or there is a possibilty for some major problems.

I have to disagree with this. My solar does about 30 amps on good days. My turbines can pump well over 100 amps all into one bus on one battery bank at nominal 24 volts. Absolutely no problems.  Volts can vary wildly, too.

Sounds like some huckster justifying some "box" to fix this?

Just from here.

Tom

[jkrienert]

Well that is awesome news tom!
I guess I should have looked no further then those with experiance under their belt.

Thanks for the uplift.
Back to sleep now.

[Flux]

Your method as shown is fine and may be the only sensible way in this case, at least you don't have to think about it.

If your coils were mounted on a precision E core with correct span then the flux in the centre limb would be twice that in the outer ones. That would imply that the outer coils would need twice as many turns as the middle one for the same voltage.

You could try this, it might work better but I doubt it in your case. You could keep the middle coil to its rectifier as now, but you could connect the outer coils in series (ac wise but in the correct phase) and connect the series pair to a second rectifier.

In theory you would get similar results with the rectifiers of the outer coils connected in series and the series pair in parallel with the middle one. The extra diode drops are against you at such low levels and anything could happen in real life.

Regarding connecting dc sources in parallel, for anything supplied by a current limited source or fed from diodes so there can't be a back feed then there is no problem. That means solar and wind can feed together to a single battery ( Tom's case)

For stiff dc supplies that can feed either way then yes there is a serious issue, Connect a 6v battery in parallel with a 12v one and you have a disaster. Similar issues with dc generators in parallel or feeding batteries. The issue is also true with ac and you have an extra problem of phase in addition to voltage to equalise before you parallel.  If you have read something on power engineering then yes you will have found this information.  The old problem of using the right information for the job in hand. At your power level nothing bad will happen ( or you won't notice it).

Flux

[jkrienert]

You cats are sharply honed in your info, and the manner of voicing it.  Thank you.

I just setup the 3 separate rectified version with two 10F 2.5 volt caps to directly lite 7 5mm white leds (1.2v at 60mah?).
Took forever to balance the load of the leds, and to fill out the s.caps-but one it did, it holds light like a battery with no flickering at all.
After the wheel stops, it illuminates for about two Minuit's solidly, then starts to decay smoothly.
The charge time makes their size kinda unfeasible unless I'm doing laps before I actually go on a ride, so I am wondering if half that farad rating might be better
in parallel (x2 5f at 2.5 volts)...

On the notion of connecting the two outside coils together am going to try that today and see the levels in comparison.
Even if there is a marginal increase, it seems worth it on such a simple setup.

Is it likely that with these two coils in series/phase I will see a raise in current or voltage (or both )?

Also going back to the origins of this thread, I had a go at mounting a perpendicular generator on top of all these other bits, and it works with much success!
Actually, it outputs almost the exact same rating as the three coils circuit (in both current and voltage).
I will try and get a photo of it up today.
It might give me the ability to run a rear and front light on their own source. Convenient.
{its basically a very low ohm 24v output 4 phase centertap stepper motor. It is mounted on a L-bracket and the shaft has a
diametrically magnetized tube neo with two poles. Hello magnetic gearing!}

Happy sunrise (from central US)

[Bruce S]

Some of the issues you're seeing with getting the LEDs lit is the voltage. The Super-Caps are at 2.5Volts and the Vf for the 5mm LEDs are the same. by the time you get the circuit completed you're already below the Vf ( Voltage forward or turn on voltage  ).
The added issue can be that Caps really really don't like being over-Voltage!
Try one of the 5.5V Caps and see how well it works out, just make sure to use a 100Ohm (or Higher) resistor for every 3 LEDs to keep the 5.5V from popping the LEDs.
Have a little fun while you're at it 

Where is Central US?

 

[jkrienert]

Southern Illinois!

[Bruce S]

St. Louis close enough?

[jkrienert]

only an hour away!

Ok, I started testing per advice on sync-ing the two outside coils. Here are my results...
(all numbers are after rectification)

     Baseline for each coil alone:
              (red)-low: 5v at 30mah
                      high: 10v at 60mah

              (green)-low: 5v at 30mah
                          high:10v at 60mah
  
              (violet)-low: 5v at 90mah [diff. coil type then red and green (more wraps)]
                         high: 10v at 180mah

Now for the outputs while in the varrying wiring patterns...
    (a)  Individual outputs in parralel:
               Low: 8v at 170mah  (1.36 watts)
               High: 15v at (cheap meter wont read high enough) but looks like at least 225mah (3-3.5 watts)

    (b)  Outer coils parralel(red&violet) + center individual:
               Low: 10v at 90mah  (.9 watts)
               High: 20v at 150mah  (3 watts)
  
    (c)   Outer coils series(red&violet) + center individual:
               Low: 6v at 30mah  (.18 watts)
               High: 12v at 60mah  (.72 watts)

Far as I can tell, the worst output is c.  
Thus I am left leaning towards A.
Which one would be best considering the application of (2-3) 1 watt LEDs?

-keep in mind, the multimeter I have is a cheap one, so their is probably a 10-20mah varriance-

[Flux]

If the coils are not identical then it is back to experiment.  Possibly the odd coil might be better in the middle (green ) position but the iron circuit linking the coil backs may not be good enough to make any of this work.

I suspect the coil that gives more output has similar turns but thicker wire. Resistance will be the factor here in all probability.

Winding the coils with less turns of thicker wire would most likely get you a better match to the low impedance of the LED. This is where the buck converter would most likely have helped, but in the end it will pay to keep things simple.

All your indications that parallel is better than series seems to indicate that you need lower resistance coils with less turns.

Diode losses may also be a factor, using bridges of 1A schottky diodes may be a lot better with half the volt drop.

Similarly I think it would be better to use a string of smaller LEDs in series rather than a single big one. If you had a higher supply voltage with less current for the same wattage it would match your coils better. You would need to raise the capacitor voltage a lot to do that.

Lots of room for experiment, like wind power it's very much a case of matching all the components to get the best result.

Flux

[Bruce S]

Flux is light years ahead of me when it comes to coils and such So I'll defer to his recommendations.
 
Is there a reason you are going with the blinding light these 1 watters give out will give out?
An hour from ST.L. means you are possibly around the Mount Vernon area. I could then understand wanting to be seen  by the cars and Semi's flying through there.
However if its just the make sure you can see the roads and be seen/safe then it might be a good idea to re-think the amount of lighting you want.
For the back a couple really good red LEDs that rate in the 20 - 50mA range will be more than enough to me seen. They have a Vf of 2.5Volts.
There are some CREE and 5mm Superbrights than have a Vf of 3.5 but go full bright at 20mA which would save you from needing such a large bank of either batteries or Super-Caps.
IF you're using LEDs you already have then I understand, how well did those 7 LEDs look? At full out power 7 5mm at 3.5Vdc will only be using 140mA.
Bruce S

[jkrienert]

yea.  close to mt. v. about 30 min. away.

well, I upped the count of the 5mm to 9, and get a really high level of light.
I really dont think I need any of those fancy CREE lights, and this will be good enough.
My concern about the level, stems from partialy wanting to be sceen, but also to see the road as the first part of my daily round trip is at about 5:30 and pitch black.

I feel I have pretty much hammered out all I can do with the mechanical end of things.
I am sure it could be more refined and efficant, but that might be better left on the next one I make for someone else.
The stage this setup is at (already encased in silicone) makes anything more then ultra-fine tuning a rain check.

All things considered, it seems to me, that Method 'A' discussed prior is going to be the way to go.
No for the circuit-which I have even less skil in :O.

[Bruce S]

OKAY;
 Now onto the LED circuit design.
Have you decided to go with the Super-Caps or Batteries?
For a good reading here's a link that has several grouped together.
http://fieldlines.com/board/index.php/topic,143569.0.html

Read on on these and see if it helps with the design.
Bruce S

[jkrienert]

Thanks!
some good reading tonight.
Apreciate it.

[jkrienert]

Wow.  The content of these articals is a bit over my head.
Not stoping me from learning as much as I can.
Applying this to a circuit is going to be tough.
challenges are good though.

Thanks again.

[Bruce S]

Wow.  The content of these articles is a bit over my head.
Not stopping me from learning as much as I can.
Applying this to a circuit is going to be tough.
challenges are good though.

Thanks again.

Try not to worry about the overall scope too much , just do the numbers as Commanda has them and plug in the values for your LEDs, it'll all fall into place.
Might look daunting at first, but her classes on this subject are right on the money.
Also , IF you have them post the values of those LEDs you're looking to use. 9-superbright , Vf=? current=?
Battery or SuperCaps? battery bank voltage=?

[jkrienert]

Sounds good. But I am at a bit of a hiccup.
Plans have changed, and this 3-coil genny will be powering a rear red/stand-light.
This is because I recently came to acquire a f-dynamo hub for the headlight-and have not even started working that end out.

My first though is to use some of the LEDs I already have, and just mount them inside an enclosure with a red reflector screen.
The current blue-ish white LEDs I have been testing with are unknown as far as specs are concerned. Yet they seem to be plenty for
illuminating the inside of a reflector box.

They are from the cheap $2 9LED flashlights you see in the dollar bin at hardware stores.
Any clue on the numbers for these such lights?

Cheers!

[Bruce S]

Plans have changed, and this 3-coil genny will be powering a rear red/stand-light.

My first though is to use some of the LEDs I already have, and just mount them inside an enclosure with a red reflector screen.
The current blue-ish white LEDs I have been testing with are unknown as far as specs are concerned. Yet they seem to be plenty for
illuminating the inside of a reflector box.

They are from the cheap $2 9LED flashlights you see in the dollar bin at hardware stores.
Any clue on the numbers for these such lights?

Well, if they are the bulb looking type then they are probably the 3mm 2.5V 20mA type that I've seen (abused  ).
If those are the same then the light uses either 2 AA or has an insert that uses 3 AAAs, which gives the same voltage output.
IF this is true, then using that part of the light is pretty easy, AND there's an even easier workaround :-) but let me know what size batteries it uses just in case.
Norm rebuilt one using some NiCd AAs to make a nice light sticks OR was it Sub-Cs  .
Come back with that info

[jkrienert]

It is the setup with a flimsy 3 AAA insert.
Right now I am planing on hard-wiring the 9 LED string in the housing - and having at least the capacitors in there.
Then sealing it up with a link out the rear.  In line switch, then the actual rectification near the rear hub.
Although... I am wondering - even though a short distance, is it better to have ac travel as line voltage or dc to the light?

[light,capacitors]-------------------------switch---------------------------rectifiers[hub]
or...
[light,capacitors,rectifiers]--------------switch---------------------------[hub]
?

The goal is for ultra capacitors - instead of batteries.
I will see how long they last-then maybe down the things may change.
The capacitors I have on hand are 2.7v at 4f (i have a handful, so some in series or parallel is an option).

And yes, those LEDs do take some abuse. I purchased a 10 pack of the flashlights at a local bargain store for $5.
Running one directly to 12v at 200mah and it peaked pretty high in brightness with no visible damage to the diodes.
Although, I will put this particular one aside for the circuit design - because its likely there is some wear due to experimenting.
Going to try some math this afternoon (not my expertise by any means - in application or outright understanding!)

Thanks!

[ghurd]

The "magnetic circuit" is a mess.
There is no reason to have a laminate core on a single coil unless it is magnetically linked to another pole.  (Mr. Flux?)
If the laminates are improperly aligned to the magnet motion, then it is an electric bike brake.
The solid steel behind the core is certainly causing drag
(eddy current)

=

"My first though is to use some of the LEDs I already have, and just mount them inside an enclosure with a red reflector screen.
The current blue-ish white LEDs I have been testing with are unknown as far as specs are concerned. Yet they seem to be plenty for
illuminating the inside of a reflector box."

LEDs are spectrum specific.  (they make One color of light)
RED lenses abosorb all the colors of light, except red.
"white" LEDs are really Blue LEDs with a phosphour to glow Yellow, so basically 'white' LEDs only make Blue and Yellow light.
Meaning a 'white' LED behind a red lense puts out very little light.
There are many more details beyond the scope of this discussion, but that is it in a nutshell.

Use only RED LEDs behind a RED lense.
Or Red LEDs behind a clear lense (even better).

=

"Running one directly to 12v at 200mah and it peaked pretty high in brightness with no visible damage to the diodes."
Or find the specs for those LEDs ("100,000 hours before they 'wear out' {become half as bright}) and see if they last 12 hours at 12V.
In another nutshell, Don't do that to them.
'it peaked pretty high in brightness'
Measured with a meter, or human eyes?  Human eyes do not register brightness linearly, and instead it is more logrythmically sensed.
This LED situation is far beyond what most people understand.

=

I do not understand what LED lights you are using.
Red, or white.  10-pack for $5, or $3 each.  Etc.

You will be better off building you own LED circuit with your own LEDs.
Use a big pile of LEDs driven at low current each.

=

Semi-off-topic,
I could argue the best is
[hub, Schotty rectifiers, caps]-----------[light]

G-