NiCD - Charging circuit - Using PIC

[Megahertz]

Hi, I have a winch machine which I have modified to work with the remote instead of the its wired control box. I was hoping if someone can help me develop a charging circuit with battery low indicator & full charge indication circuit with or without PIC for the remote. I have couple of 12F676 for the job if needed.

I have made 3 battery packs of four 1.2V (700mAH) NiCD each, so each pack has 4.8V & total current 2.1 AH for the transmitter. I have a spare nokia charger with 5.7V @ 800mA output which I am hoping to use for this purpose. So if anyone knows or have seen any circuit which puts all together, please share it.
Thanks

[ScaleRobotics]

Here is Microchips AN1293, but it requires a MCP1631HV chip. Might give you some ideas though.

http://www.microchip.com/stellent/id...pnote=en544929

[rmteo]

How about this high efficiency charger (available in an 8-pin DIP) LT1510 - Constant-Voltage/Constant-Current Battery Charger

[prstein]

Interestingly, I'm working a similar problem right now and am considering starting my own thread...

Unless you are committed to using a PIC, take a look at the MAX 712/713. It works well enough but I think you might need a higher-voltage power supply, regardless of what you use to charge it. Several years ago (many years ago?) Melanie discouraged creating a PIC-based solution when chips such as the MAX 712 were available.

On the other hand, there is something to "rolling your own"...

If you are committed to using a PIC and want to keep it really simple, consider the humble LM317. NiCads are fairly forgiving in their charging (although mostly I use NiMH so that's really the impression I've gotten rather than experience), and you just need to pump them with a constant current until the voltage plateaus (i.e., stops increasing). The datasheet for the LM317, on page 18, shows an example 1A constant current source. Substitute your battery for the load, add another little regulator for the PIC, put an ADC on the top of the battery (through a voltage divider) and the rest is just programming... Read the MAX712 datasheet as an instruction for how it should operate.

The problem I'm having is that the LM317 is throwing more heat than I am comfortable with sealed in an aluminum box. Not horrible for a lot of situations, but more than I really want to keep in my particular application.

Hope this helps...

Best Regards,
Paul

[Megahertz]

Thanks for the heads up on this guys. I think looking at the voltage measuring problem, I will go for the PIC route finally as there is a very good article from Melanie about using comparators, that even a novice like me understood easily. Also, the LED can be controlled easily and more flexibility is there down this route.

However, this works fine when reading the voltage when it is dropping below a certain threshold.

BUT what changes can be done to measure the voltage when the battery reaches it's full charge.

The problem is that, once the charger is attached, the voltage reading is misunderstood by the PIC as the voltage it reads is equal to voltage from the charger, only when the charger is removed, correct reading is taken by the PIC. I was under the impression that when charger is plugged in, because of the current drain, voltage drop will be there & gradually as current will stop & voltage will rise and PIC will have correct reading. BUT the voltage drop is not that significant here. I have even tried it with sealed Lead Acid battery as well and its own charger but I face the same problem.

[cncmachineguy]

I'm not certain, but I think you need to measure the current doing into the battery.

[prstein]

Megahertz,

A couple of items for your consideration.

-I think you should take a hard look at using an ADC instead of a comparator. I actually had a lot more trouble figuring out how to use the comparator than the ADC.

-It would be helpful, for the purposes of this discussion, to sketch out at least a rough schematic of what you are trying to do.

-To charge NiCD or NiMH battteries, you need to supply a constant current. As I understand what you describe, you are just hooking up a regulated voltage supply. That will destroy your batteries sooner or later. The LM317 I linked earlier is the heart of a constant current supply. You also need a top voltage on the power supply of at least 1.9V per cell. That means you need at least 7.6V to charge 4 cells. The 5.7V charger is not enough.

-Lead acid batteries instead use a constant voltage to charge. The charging voltage is somewhat higher that the listed voltage, i.e. you use 14-point-something volts to charge a 12V SLA battery.

I've started my own PIC-based charger. I haven't written code yet but I can post the basic schematic tomorrow.

Best Regards,
Paul

[prstein]

The rough schematic for the battery charger I am playing with is attached. It's not the cleanest schematic ever but I trust it is sufficient to get the idea across.

D1 indicates that power is connected to the device. D2 and D3 are controlled by the PIC and I have them shown as being used to indicate that a charge is active and when it is complete.

VR1 provides the regulated +5V for the PIC. VR2 is set up to provide a constant current of ~80 mA when GP1 on the PIC is low and about 220 mA when GP1 is high. The relay switches the resistance between 16 ohms when open and 6 ohms (16 and 9 in parallel) when closed.

R7 and R8 create a voltage divider so that roughly 30V at the top of the batteries will show up as 5V at AN0.

The PIC, a 12F675, will on startup begin a "fast" charge cycle. Once it detects that (a) the voltage is too high on the cells, (b) the voltage has stopped increasing (or, for NiMH, started decreasing), or (c) too much time has elapsed it will open the relay to put it in trickle charge mode.

There are any number of items that could be changed but I think this is a decent start...

Best Regards,
Paul

[chuck]

This is a well documented about a Atmel charger, for SLA,NiMH,Nicad and Li-lon batteries you can get some very good information from it, There is code avaliable in C for Atmel but you can understand bit and may be easy converted, Just replace the the AMtel with a pic

http://www.atmel.com/dyn/resources/p...ts/doc1659.pdf

[Megahertz]

Thanks for the input guys. Prstein, you have surely got me thinking about using ADC. I have spent 2 hours today understanding it from the datasheet. It seems to me that I have an idea of what goes on (just kidding ), but still few clarifications needed.
1) I read 16F676 datasheet, it says 10 bit result - which means 1024 steps conversation. Now 1024 steps with what respect (is it the VDD)?
2) How do these steps relate to voltage supplied by two 1% precision voltage divider resistors on the ADC PIN.
3) In a nutshell, how does these 3 relate( internal VDD as Vref, 1024 steps, Voltage on ADC PIC by V. divider)?

After understanding the basics above, I will be surely be able to dig my way up and carry on with my charging solution. Having the knowledge from this thread only, I am considering using 12V Sealed lead acid battery and replacing NiCD. It seems if I can control the voltage drop of lead acid, it is more forgiving in terms of charging and it is cheaper as well + I only need one with sufficient current upto 4AH etc. Also using 12V in my TX is giving me increased range
Where as, putting two 4.8 NiCD in parallel, I have read that the current needs to be pushed in, and because the two battery packs may have different impedance(whatever that is), there is always a possibility of one charging more than the other.

So voltage drop of the 12V battery is sorted, can be done with comparator easily. Once ADC conversion is clear and the relation between 3 values mentioned above, I think I can do the charging & full charge indication as well.

[prstein]

Hi Megahertz,

The invincible Darrel Taylor has spelled it all out for us!

http://www.picbasic.co.uk/forum/content.php?r=245

Best Regards,
Paul

[Megahertz]

Darell is as always THE GURU of all this stuff. The example speaks for itself, no more questions on that front.

But if I have understood it correctly, he is measuring the voltage through a voltage divider, what can I do for current sensing? I think for that I need to add a very small value resistance in series to the current path and measure voltage drop across it, when it is nearly 0, battery charge is full, but what next? How to make a connection with PIC?

OR

Is that in his circuit he is actually doing the same thing? one set of divider on one side of the series resistor, and the other set is on the other side of the series resistor. Keep measuring both sides until they are giving nearly same readings?

[prstein]

I know almost nothing about current sensing. What little I do know I got from the MAX712/713 datasheet. I need to let others respond to this one.

Best Regards,
Paul

[pedja089]

You can use 0.1R resistor, in serial with batery. And measure voltage drop on it. I=U/R
Because you use smal value resistor( to reduce dissipation) you need to amplify resistor voltage. You can use just simple Non-Inverting Amplifier. Or something like MAX4080. Look datasheet for details. I used it, and it very simple, and it give 0-5V output, wich is perfect to use with ADC.

[ScaleRobotics]

A lot of people like the Allegro hall effect current sensing devices.

Key features:

• Low insertion loss; power dissipation much less than shunt solution
• Able to monitor both AC and DC currents
• Smaller form factor than shunt and current transformer solutions
• 5 V, single supply operation
• Voltage isolation rating certified by TUV America
• RoHS compliant

The devices for 5 up through 30 amp are bi-directional, and are on a 8 pin soic.


http://www.allegromicro.com/en/Produ...rentsensor.asp

[rmteo]

Try this 76V, High-Side, Current-Sense Amplifiers with Voltage Output



The MAX4080/MAX4081 are high-side, current-sense amplifiers with an input voltage range that extends from 4.5V to 76V making them ideal for telecom, automotive, backplane, and other systems where high-voltage current monitoring is critical. The MAX4080 is designed for unidirectional current-sense applications and the MAX4081 allows bidirectional current sensing. The MAX4081 single output pin continuously monitors the transition from charge to discharge and avoids the need for a separate polarity output. The MAX4081 requires an external reference to set the zero-current output level (VSENSE = 0V). The charging current is represented by an output voltage from VREF to VCC, while discharge current is given from VREF to GND.

For maximum versatility, the 76V input voltage range applies independently to both supply voltage (VCC) and common-mode input voltage (VRS+). High-side current monitoring does not interfere with the ground path of the load being measured, making the MAX4080/MAX4081 particularly useful in a wide range of high-voltage systems.

The combination of three gain versions (5V/V, 20V/V, 60V/V = F, T, S suffix) and a user-selectable, external sense resistor sets the full-scale current reading and its proportional output voltage. The MAX4080/MAX4081 offer a high level of integration, resulting in a simple, accurate, and compact current-sense solution.

Better still, use the LT1510 mentioned in post #3 - no PIC or programming required.

[Megahertz]

I would like to thanks everyone who has contributed above. Finally, I will see if I can find MAX4080 easily where I am.

There is one new problem, I don't know if I should start a new thread or carry on here, but I will give it a shot here first: So here I go:

I have one Doorbell, it runs on mains 230V currently, few zeners drop the voltage after the rectifier and through 78L05 the PIC & the RF receiver get their 5 V.

Now I would like to modify this circuit in such a way, that if there is a power cut, the system should switch onto two AA standby batteries for brief period of times (like appx 1 hour or so every few days) . Both PIC & RF receiver can work on appx 3V minimum (PIC can go down upto 2.5V). Now, these batteries are not chargeable, so no voltage should enter in them, I had a eureka moment and thought that it can be done with diodes, but the voltage drop across them makes them unsuitable as I only have around 0.6 V (new AA have appx 3.6V) appx to play with and I can't share it with the diode. Can someone think of a suitable circuit that will make this a possibility(with easily available components)?