NiMH Battery Charger With MAX712

[waffle]

Hey guys

I was planning in using the MAX712 chip to charge a 9.6v 2500mah NiMH battery pack. After reading through the datasheet a few times, this is what ive come up with.

I wanted to use a laptop power supply for power (**broken link removed**)

I wanted to charge at 1C (2500 milliamps). I dont want to use the temperature sensing feature.

Here is my schematic.
R1 = 2340 ohms (or closest resistor)
Rsense = .1 ohms (ten 1 ohms in parallel?)
**broken link removed**

Here is the datasheet for the chip
http://datasheets.maxim-ic.com/en/ds/MAX712-MAX713.pdf

I have a few questions regarding my circuit:

1) I notice the diode used in the Typical Operating Circuit is a 1N4001. Digikey says this can handle up to 1 amp. Since I want to charge at 2.5 amps, this is a problem, right? I was looking around on Digikey and i found this beast (Digi-Key - 641-1317-1-ND (Manufacturer - 6A4-G)). Supposedly it can handle 6 amps, with almost every other detail the same as the 1N4001. Would this be a good substitute? I get the sense that this isnt necessary, since this diode is huge, and it looks like one of those super specialized things. 6 amps isnt even that much...

2) With the current circuit, the temperature sensing feature wouldnt interfere would it? (keep in mind, i dont want to use it. I want it to essentially be disabled)

3) The LED should go ON when fast charging and OFF when trickle charging. Is that setup correct?

4) Would ceramic capacitors work for C1 and C2?



Finally,
Do you see any problems with this circuit?

Thanks!

 

[Diver300]

Page 12 of the Max712-713 datasheet says how to use, and how to disable, the temperature sensing circuit. The resistors in the diagram that you have used are what they recommend to disable it.

Beware of cheap supplies like that one on Ebay. They are generally not that well made.

The power supply voltage is nearly twice the battery voltage. You will be dropping around 8 V, which at 2.5 A means 20 W of heat in the transistor. You need a heatsink of no more than 2.5 deg C per Watt. If you found a lower voltage supply, the heatsink would be much smaller. Also, that transistor is only just rated to what you are doing. A larger one would have a bigger safety factor, but reducing the supply voltage would be better.

 

[waffle]

What rating of the transistor is borderline? The reason I picked such a high voltage is that the datasheet says to multiply the number of cells by the maximum possible voltage for each ( it says to use 1.9 to be safe).and add 1.5.
(8*1.9)+1.5 = 16.7
Like you said, this is probably a low quality power supply, so I expect a significant voltage drop if I draw lots of current. My hope was that drawing 2.5A would drop the voltage to a more reasonable range.


Can you comment on the diode?

Thanks!

(and sorry for the language, I edited it out.)

 

[Diver300]

Those power supplies won't drop much at all at 2.5 A.

I see what you mean about the voltage, although 1.9 would imply batteries in quite poor condition. You do need a voltage margin or the peak is impossible to detect when you hit the voltage limit.

The transistor is rated to have a minimum gain of 30, and the Max712 / 3 can drive 30 mA. That gives you a guarantee of only 900 mA, although you will probably get more. Fig 5 of https://www.onsemi.com/pub_link/Collateral/2N6107-D.PDF has where you want to run on the thermal limit. That is why I think that the 2N6109 is marginal.

 

[waffle]

Upon looking around on Digikey, there arent a whole lot of high gain transistors that suit my needs. Could i just use two 2n6109s in a darlington pair to get
30ma * 30 = 900ma * 30 = 27,000 ma

anyone on the diode?

 

[Blueteeth]

I'm sure you could use them as a darlington pair. The extra 0.6V drop won't make that much of a difference. For 9.6v nicads, the maximum voltage I would say is 1.6V per cell, = 12.8V. 1.9V per cell seems awfully high to me, even when charging at 2C I've never seen the voltage spike over 1.6v, (but different batteries have different specs), so 15v seems appropriate. You may be able to modify your PSU, as many smaller laptop PSU's are flyback converters, with a resistor divider in the feedback path controlling the output voltage. The efficiency may suffer though. This would minimize the heat disappated in the transistor, a heatsink would obviously still be required.

For charging currents above around 1.5A's the industry generally uses switched mode power supplies, to deliver constant current. But that opens up another can of worms which you don't need.

About the diode, for testing I would get a TO220 version. These can take serious current and are easy to heatsink, generally don't cost that much. At 2.5A a silicon is likely to drop >1V = 2.5W. The maximum a to-220 can kick out on its own. A shottky diode would have a lower drop, but not really necessary in this circuit, it would just slightly increase the drop across your transistor. Maybe this?
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=497-5285-5-ND

As always, its best to knock up a quick and dirty prototype and get your multimeter out. See how hot things get. You could even test the constant current supply by using a resistor to set the base current of your driving transistor/darlington pair and checking the current into a load resistor. Also remember, you can use transistors in parallel in you're careful.

 

[waffle]

I have added a second 2n6901 transistor in a darlington array. (ive never used pnps in a darlington, is this right?) And i replaced the diode with the one blueteeth suggested.

**broken link removed**




Any last bits of advice before i build this?

 

[Blueteeth]

Looks good to me!

Well only advice I can give would be generic prototpying Stripboard/vero board is greart stuff, easy to solder/desolder parts to, to cut, and add test pins to. I cannot tell you how often I wish I had done this before starting a a final version. A socket for your Max712 (assuming its dip) as well. Perhaps if you have some handy - a fuse. Maybe its because I'm a miliamp kind of guy (anything over 1.5Amps scares me) but its a nice simple over current protection to save your components.

If you're unsure about the darlington, knocking that up too wouldn't take long, good for measuring input/output currents/voltages.

Maybe I'm just the cautious type, but prototyping in 'modules' is great. Allows ones to tweak each part of the circuit so it does its job greatly increasing the chances of it working off the bat first time.

Good luck, and let us know how it goes

 

[waffle]

Great!

Thanks both of you for all your help. I will go ahead and order the parts today, and then I will wait!

The main thing that scares me about prototyping is i dont have a huge amount of parts on hand, and have to order them all. I dont want to wait 2 weeks to get the stuff just to find out i needed a different kind of transistor. I plan on getting a large assortment of resistors so i can mess with charge currents and such.

Again, thanks for your help!

 

[Blueteeth]

Carbon resistor packs are cheap and will last for many many projects. I would check ebay for bumper packs, not great for 'high quality' parts, but cheap and cheerful for your 'bits box'

 

[Diver300]

On the darlington pair, it could be easier to buy it as one component. https://www.onsemi.com/pub_link/Collateral/BD676-D.PDF

If you do use two transistors, only one will generate more than a tiny bit of heat. The extra resistors as shown in the BD676 datasheet are a good idea so that the leakage through one transistor isn't amplified by the second transistor.