NiCad Charger - Need some hlp on components :D

[pittuck]

Ok here is the 'circuit' with some missing values!

Basic opperation goes like this; the first LM317 is used to create a contant voltage, this is chosen using a digial (i2c or spi) potentiometer and a PIC microcontroller. The minimum voltage out is 1.25V, so i have put a diode in series to get a voltage drop to allow for lower voltages. This is the first 'problem', which value of diode will work. I know the voltage drop changes with temp, but the output voltage after the diode is checked using a ADC on the PIC and so i can adjust to get the correct value. But the diode should be able to handle up to 0.5A.

Next is another LM317 which is used as a constant current source, now i want 4 different charging rates, 8mA, 27mA, 150mA and 300mA. The resistors are to be 0.5W Metal film type (the largest calculated power on the resistor is 400mW, but to get the resistance from standard resistor values i need to put 2 8.2ohm resistors in [edit]parallel[/edit]).

So to choose the output current i am using PNP transistors. A) will this work, B) is there a better way and c) and components you can advise?

There will also be current sensing and battery voltage sensing to make sure things are going ok.

Further more, this circuit should prevent the battery from powering the circuit, but i would like a discharge option...

perhaps at 20 - 30mA?

REgards

- Martyn

 

[Nigel Goodwin]

First thing! - why the voltage regulator at all?.

It's totally superfluous, and would probably stop the second LM317 working, if you didn't supply it enough volts.

I'm dubious about the transistors switching the current as well, FET's would work better as a 'switch', but in either case you need to have the biasing correct for it to work.

A more common way, if you're using a PIC, is to have a preset constant current source (at your maximum setting), and pulse it ON and OFF with PWM to give your lower curretn settings.

For a PIC based charger, have a look at https://www.angelfire.com/electronic/hayles/charge1.html, which should give you more ideas - notice how it uses an LM317 to give a constant current, and then uses PWM to adjust it.

 

[pittuck]

ok thanks,

i had not realised the batterys will charge at any voltage... Would it cause problems if u tried to charge a 1.2V cell with a 12V source???

If so then perhaps i could just use the first regulator to bring down the voltage to an acceptable level for charging (2x the cell voltage?)

The PWM thing would be much better, and allow more scope for current values.

Just to add a note, this is meant for the small NiCad solderable cells at 3.6V & 170mA, its for a kind of kit type shop, but the problem is that all the chargers avaliable commercially are fairly overkill for charging this type of cell!

IE most of the ones i have seen start at £20-30 and a fairly big. Considering the cells cost $0.69 (at solarbotics.net), the charger costs quite a bit more!

 

[audioguru]

Hi Martyn,
1) A Ni-Cad cell charges at nearly 1.4V, not only 1.25V which is its discharging voltage. Read about them at www.energizer.com and click on their Technical Info. They still post info about their obsolete Ni-Cads. Look at their new very high capacity Ni-MH rechargable cells too.
2) You are planning to "quick-charge" the cells but don't have a method to detect and shut-off when fully charged. The Energizer site describes the nice explosion that you'll have when all the charging power is converted to heat when the cell becomes over-charged!
3) Your diode is backwards and should be at the output.
4) Don't you understand? You don't need to "bring the voltage down". I agree with Nigel, you don't need a voltage regulator since a Ni-Cad or Ni-MH cell limits the voltage anyway. They are charged with current. You could charge them from a few million volts if you limit the current to their rating.
4) Your supply voltage is way too low for a single LM317 current regulator. It must be at least 3V higher than its output voltage to regulate properly as shown on its datasheet.
5) How can you have 4.1 ohms with two 8.2 resistors in series? Put the resistors in parallel and measure the result to make sure.
I don't think the resistor values are correct anyway.
Your schematic shows mW which is power, instead of mA which is current.
6) I agree with Nigel, the transistors make lousy switches, use Mosfets.

 

[pittuck]

haha, making circuits with a hangover @ 9ish on a bank holiday morning is never good.

LOL

1) Thought u can get lower voltage cells than 1.2V (1.4V)

2)

There will also be current sensing and battery voltage sensing to make sure things are going ok.

3) my bad

4a) okies

4b) Supply is 5V - 12V on the schematic, 1.4 + 3 = 4.4V.... Only when charging a 9.6V pack would u have touble, but u can always use a different power source...

5) my bad lol.

6) okies, i will actually use the PWM instead. Probably make the max current output 0.3A and go from there.

 

[pittuck]

I don't think the resistor values are correct anyway.
Your schematic shows mW which is power, instead of mA which is current.

Take 150Ohm resistor, 1.25/150 = 0.00833 = 8.33mA (which is what i labeled the 'transistor' to use that resistor.

0.00833^2 * 150 = 0.010Watt = 10mW

I was just using the schematic to make sure the resistors chosen would have the correct power ratings.

On a side note, i have a nicad charger that requires the voltage to be selected for it to work, would this just be for comparing and identifying discharge levels / charge peaks?

 

[audioguru]

Hi Martyn,
Now you are on the right track. As Energizer says, fast chargers use a thermal sensor and/or timer as a backup to shut down the charger in case the controller fails or misses the voltage dip when fully charged. You don't want a big BANG!

I am sorry when I said your resistors are incorrect about the charging current. They are correct. I mixed-up your mA with your mW.

Are the cells 370mA/hr capacity, or 370mA of charging current for 14 hours?

 

[pittuck]

well i use 2x 3.6V, 170mA cells in parallel, although i might go for series and a switching regualtor (long story).

Anyhow, the numbers on the side relate the the capacity, rather than charging rate.

I will work out a new circuit later, i need to do some work on a online Design Rule Checker i have been making for sparkfun.com, and do some revision for college.

 

[pittuck]

Ok, here is a pre-lim, first thoughts circuit

Erm, the constant current source is pretty much the same as the one in the link nigel gave. I am not 100% sure how it works though.

The one difference is the 4.1ohm resistor instead of a 1ohm one. There is also the voltage and current sense resistors there.

The 0.1ohm resistor should be about 15ohm which should be rated at > 1.35Watts (it should give a 5V output at 330mA).

Any problems? The spare pic pins will be used for a LCD where voltage (for peak drop), charge current and cut off times can be set.

Just to make sure i get it right, a 50% duty on the transistor base will provide half the current of a 100% duty? and what range of frequencies are to be used???

Finally, any ideas on discharging? perhaps a seperate circuit swtiched to with a relay and monitered with the PIC?

 

[Nigel Goodwin]

ok thanks,

i had not realised the batterys will charge at any voltage... Would it cause problems if u tried to charge a 1.2V cell with a 12V source???

No it wouldn't, you don't charge them with VOLTAGE, you charge them with CURRENT - you either have a constant current source, or a simple resistor, to set the current to what you want - the voltage is fairly irrelevent, as long as it's high enough to allow the correct current to flow, and not so high the current limiting overheats.

If so then perhaps i could just use the first regulator to bring down the voltage to an acceptable level for charging (2x the cell voltage?)

WHY??? - all you are doing is adding extra stages which will only reduce the effectiveness of the charger, perhaps even stopping it working all together!. It will only provide disadvantages, not a single advantage at all!.

 

[Nigel Goodwin]

Ok, here is a pre-lim, first thoughts circuit

Erm, the constant current source is pretty much the same as the one in the link nigel gave. I am not 100% sure how it works though.

When the transistor turns ON it turns the current source OFF, so it can be switch ON and OFF by the PWM, presumably you've connected the base of the transistor to a hardware PWM output?.

The one difference is the 4.1ohm resistor instead of a 1ohm one. There is also the voltage and current sense resistors there.

You don't need to sense current, it's a contant current source! - what point would measuring it achieve?.

The 0.1ohm resistor should be about 15ohm which should be rated at > 1.35Watts (it should give a 5V output at 330mA).

What is that resistor supposed to be doing?, it looks like it's there solely to destroy the PIC?. It's certainly not monitoring anything - which would be a waste of time anyway!.

Any problems? The spare pic pins will be used for a LCD where voltage (for peak drop), charge current and cut off times can be set.

Just to make sure i get it right, a 50% duty on the transistor base will provide half the current of a 100% duty? and what range of frequencies are to be used???

Yes, as simple as that - frequency isn't critical, perhaps the website I posted previously may have some thoughts on it?.

Finally, any ideas on discharging? perhaps a seperate circuit swtiched to with a relay and monitered with the PIC?

A simple NPN transistor switching a resistor across the battery is all that's required - you should also move R5 to the battery, NOT before the reverse blocking diode.

 

[audioguru]

Hi Martyn,
In your new schematic, you don't need D1. It just reduces the input voltage of the 7805 to lower than the 8V minimum for it to work properly.
With your 7V minimum input voltage without D1, some 7805 regulators won't regulate because the datasheet shows that 7V is its guaranteed dropout voltage where the weakest ones are dropped-out.
You also might as well add some additional input voltage to allow for times when the mains voltage drops (the 'fridge turns on, your stereo or PC are turned on).

 

[pittuck]

there is a specific power supply we plan on supplying / using which it is possible to plug in backwards.

I plan to use a LDO 5V regulator, but it depends on the current needs of the pic side. I have some SMT regulators from TI which should fit the bill, but are a bit too over speced (1.5A). But i will make sure the regulator has enough overhead.

 

[pittuck]

haha, ok current sense is gone (i had missed a impotant bit out), i will just have to remember some terminals for testing