Storage for my VAWT

[taylorp035]

For the people that have not been following my treadmill windmill, I am building a 1.2v machine that is designed to cut in at 7 mph or so.  Output should be at a steady rate of 3 amps or less.  The 1.2v pack would end up being 108 ah, so a rate of 1/30 C or less would be normal.

My plan is after the initial testing stages, I was thinking of using 12 nimh c cells rated at 9ah each.  But after doing some heavy reading ( maybe 6 hrs worth ), it seems like nimh would be a bad choice for me.

My research came from this well made website:
http://www.batteryuniversity.com/partone-13.htm

Things that I find disturbing include:
Nimh cells don't like to be overcharged
High discharge rate ( may be hard to keep up with the self discharge rate )
I don't have a nimh charger.  I would never know if the cells were charged.
Nimh cells don't like to be discharged all of the way.


Pro's:
high capacity
low cost
1.2v compared to 12v


A Nicd would have better overcharge characteristics and a higher tolerance abuse, but at a much higher cost.

I was wondering if I could tell the SoC of a nimh cell based on voltage.

[zap]

I was wondering if I could tell the SoC of a nimh cell based on voltage.

Unless you monitor amps in/amps out(along with voltage)... voltage is the only practical way you can tell SOC in nimh.

As for your cons there are circuits you can find online for building your own charger although I don't know how well these would work being powered by a vawt.

For the high self discharge rate problem, look towards the lsd (low self discharge) batteries such as the Sanyo Eneloop, Maha Imedion, etc.
http://en.wikipedia.org/wiki/Low_self-discharge_NiMH_battery

[taylorp035]

Thanks zap.

I think I might even make my own SoC table for nimh cells with my AA's.

[ghurd]

The fancy chargers do some really fancy stuff now.  Reverse slope and such.

I have had perfectly good results just charging nimh and nicd based on voltage.
Basically just floating them about 1.40V each.
I believe most of the 'power' goes in at about 1.35V or so.  Takes a long time to get them past about 1.38V.

They may not get to 100% charge.  Maybe they only get to 90%?  I don't know.
But I do know they are not getting the electrolyte boiled out.  And I would rather have batteries that last for years instead of worrying about that last 10% of charge.

Most of mine are Well over 5 years old.  Some are showing signs of age, some are toast, etc, but overall that's still pretty good for their age (IMHO).
A lot better life than B&D VersaPack batteries of either type, in the factory charger!

We are still using the ones I posted pics of back in 2005.  New camera has an internal lion bat, so not using AAs as much now, but they still work.
http://www.otherpower.com/images/scimages/2050/PV612.jpg

Stray thought.
Given the % of diode loss in a 1.4V system, have you considered going up to 2.8V?
It would only take about a 2/3rd increase in RPM.
G-

[taylorp035]

I have thought about doing 2.8v if the 1.4v test is successful.  2.8v would probably yield a 150% increase in in power generated as the motor is more efficient at higher voltages.

As for the batteries, I have several nimh cells that are 5-6 years old and some nicd cells that are 18-20 years old that I use every day.

How many volts will I loose through my bridge rectifier?  I guess I could ask you guys, or I could go test it myself....

[DamonHD]

Standard silicon bridge rectifier will loose 2x0.7V, ie 1.4V typically, rising to 2V under heavy load, I think.

You could build a bridge with Schottky diodes dropping about 2x0.4V-ish, ie 0.8V total.

But again, someone should leap in and correct me.

Rgds

Damon

[ghurd]

DC treadmill motor, right?
Then if the bridge is wired correctly for use as a DC blocking diode, loss is 0.7V.

Schottkys would be better.  One for each cell.
Bigger amp rated Schottkys would tend to have less drop at a lower charging current.
G-

[taylorp035]

Then if the bridge is wired correctly for use as a DC blocking diode, loss is 0.7V.

Ok, what is this special way of wiring it up?  I currently just run the power into the AC terminals as normal.
I have made bridge rectifiers out of 4 diodes before, but this one you cant really tell how it is wired up.  I guess if I compared the 2 diagrams...

This would be good to know since I am rewiring it.

[RP]

Since the motor is already putting out DC anyway you only need a single diode in-line with it to prevent "motoring" the mill.

Easiest way is to remove the battery negative wire from the bridge and also the negative wire from the treadmill motor and connect the two wires together.

NOTE: it will be important to know what polarity the DC is coming from the motor.  If you get it wrong, the diode will prevent any power from getting through to the battery.

[TomW]

NOTE: it will be important to know what polarity the DC is coming from the motor.  If you get it wrong, the diode will prevent any power from getting through to the battery.

A simple way to ensure proper connection without a meter is:

Connect the diode so the motor turns by experimenting reversing it. When the motor runs just reverse the diode. This is how it should be connected for charging batteries This also ensures your diode "works".

Note. This WILL NOT work with a bridge rectifier just a single diode!

Tom

[ghurd]

If the Pos from the motor output is connected to both AC terminals,
Pos bridge output to the Pos battery terminal,
Then it uses 2 parallel internal diodes. 
That way only drops about 0.7V. (Its the same as a single diode)
G-

[taylorp035]

Thanks guys.  I don't have a large diode, so I use my 25A bridge rectifier for everything.  Hopefully I can have results tomorrow.  All three blades are made and the hub is ready and mounted.