building a NiMH charger

[degarb]

I am thinking of just building my own 12 volt aa pack and charging all cells at once or even several packs at once. 30 watt hour pack, times number of workers. Too many chargers, being too expensive and too impractical, unless I can build my own home made a pack charger.

What are the needed tools to build own charger? Like, "14 volt dc output at 1 amp out put for 2 hours on timer, and canabalis xx to get voltage cut off switch". How important would a voltage cut off be? How important for all cells to be of equal stages of fatigue.

 

[MrAl]

Hi there,


There is one way around building a sophisticated battery or
pack charger. The way to do this is to estimate the charge
taken out of the cell or pack while it's being used, then
put it on the charger for the correct amount of time to just
replenish that amount, plus a small percentage like 10 percent.

An example makes this easy...

Lets say you have 2000mAh cells, and your device (load)
draws 2 amps. Now lets say you use your device for 1/2
hour. 2 amps times 1/2 hour is 1 Ahr, or 1000mAh, so you
need to replenish 1000mAh plus another 10 percent (for
a fast charger). Now if your charger puts out 1 amp,
this means you have to charge for 1.1 hours. At the end
of that time your cells will be approximately charged up
again.
It gets a little more difficult however when you let the pack
sit for some time, as then you have to estimate the self
discharge, which could be 2 percent per day. This means
you have to charge a little longer than the above calculation
based on 2 percent per day loss of charge.

Another example...
2000mAh cells, 1 amp load for 1 hour after sitting for one
month, and a 1 amp charger.

The charge time consists of two parts:
1. In use discharge
2. Self discharge

The 'in use discharge' was 1Ahr.
The self discharge for 30 days is:
2000*(1-(1-0.02)^30)=909mAh, or 0.9 Ahr.

The total is
1+0.9=1.9Ahr
Multiply that by 1.1, we get 2.09 Ahr,
so for a charger that puts out 1 amp we have to charge
for 2.1 hours to replenish the charge.

It helps to monitor temperature too.

 

[qwertyydude]

Timer charging a nimh pack is a no no. They heat up far too much and too fast at the end of the charge when they're at full capacity. Best to get a hobby charger which can charge fast and automatically detect a full pack.

 

[MrAl]

It does help to monitor temperature. This tells you if the pack
is getting too hot. This can be done with a thermistor or two
stuck inside the pack so it touches at least two cells.
Measure the resistance and do a little calculation and you
know the temperature. Some thermistors are as cheap
as 50 cents each, unless you go SM when they could be
even cheaper for the 10k resistance units.

 

[qwertyydude]

The sharp tempertature rise is kinda unreliable as you may be overcharging and is usually an emergency backup for more advanced chargers. So just monitoring temperature for charge termination shortens cycle life of the packs.

 

[Mr Happy]

The sharp tempertature rise is kinda unreliable as you may be overcharging and is usually an emergency backup for more advanced chargers. So just monitoring temperature for charge termination shortens cycle life of the packs.

This is not strictly true (in fact it is incorrect).

The minus delta V algorithm used by most consumer chargers relies on the temperature rise of the cell to produce the minus delta V signal -- the minus delta V signal will not and cannot be produced until the cells start to get hot. It is therefore an indirect method of detecting the end of charge, and it explains why fast chargers that use the minus delta V method produce hot cells at the end of charge. The faster the charge rate the hotter the cells tend to get.

A more reliable way to detect end of charge is to measure the cell temperature directly and avoid the indirect use of minus delta V. However, to do this well requires temperature sensors that are well integrated having good thermal connection to the cells, and is most suited for dedicated battery packs rather than removable consumer cells.

When it is possible, the use of temperature sensors is the preferred way of detecting the end of charge condition for NiMH cells.

 

[MrAl]

The sharp tempertature rise is kinda unreliable as you may be overcharging and is usually an emergency backup for more advanced chargers. So just monitoring temperature for charge termination shortens cycle life of the packs.

Hi again,


As MrHappy mentioned, that's not quite true im afraid, and the proof is in the testing:



As this and many other tests like this show, if i had used temperature as the end of
charge termination method i could have actually terminated the charging sooner.
Sometimes the difference is a little more, sometimes a little less, but the
temperature profile shows the end of charge in better detail than the
voltage profile does and so allows its determination sooner.

 

[AlexLED]

... the temperature profile shows the end of charge in better detail than the voltage profile does and so allows its determination sooner.

I second that.

If you look carefully at the graphs on page 5 of the Twicell NiMH technical documentation, significant rise of the temperature occurs slightly sooner than the peak voltage or even -dV:

http://us.sanyo.com/batteries/pdfs/twicellT_E.pdf

(Of course, you will need a current with at least around 0.7 x capacity for any reliable sign of full-charge-state.)

@degarb
Of course, you could use of the common charge-ic's which use the -dV as full-charge-state and build you own charger. But common chargers used for rc-cars or similar should do the job quite well and are available from around 70 USD. Some examples are even listed here in the forum.

However, keeping the cells balanced in a 10 cell pack will be a challenge. Try to manually select the cells by actual capacity from the beginning, avoid too deep discharge, monitor the voltages of the individual cells every 4 weeks or so and plan to re-pack the cells after some 100 cycles.

Keep in mind, an inferior cell will degrade even faster and already one degraded cell will significantly lower the power of the pack.