Building a MAX712 battery charger -- question about min/max values

[Jack.Straw]

Hi. I'm going to attempt to build an NiMH battery charger using the MAX712 IC. My battery pack consists of 10 AA 2000mAh cells. The datasheet for the MAX712 can be found here. I could use help determining minimum and maximum values for both my battery pack and my power supply in order to accurately select my components and settings. Any advice you can offer would be appreciated.

Choose an external DC power source (e.g., wall cube). Its minimum output voltage (including ripple) must be greater than 6V and at least 1.5V higher than the maximum battery voltage while charging.

What value do i use for "maximum battery voltage while charging"? 10 x 1.2v = 12 volts, but is that the maximum voltage while charging?

PDPNP = (maximum wall-cube voltage under load - minimum battery voltage) x (charge current in amps)

What value do i use for "maximum wall-cube voltage under load"? This obviously depends on the power source i choose (based on my first question), but does it really equal the rating listed on the wall cube? For example, if I have a power supply that says 15v DC, is 15 the value i should use?

What value do i use for "minimum battery voltage"? Again, i know 10 x 1.2v = 12v, but how do i know the minimum?

R1 = (minimum wall-cube voltage - 5V)/5mA

And finally, how do i determine "minimum wall-cube voltage"?

Thanks for your time & advice,
-Scott

 

[(*steve*)]

This should answer many of your questions.

It appears that 1.8V/cell is the highest voltage, so you need about 20V.

For answers to the other 2 problems you need to know the charge current.

Place the power supply under this load and measure the voltage. (this needs to be at least 20V -- see above). This is the max voltage under load (it is also the min voltage under load if the supply is well regulated and not variable).

The minimum voltage per cell is typically 1V, so 10V in your case.

 

[Jack.Straw]

Steve, thanks!

It appears that 1.8V/cell is the highest voltage, so you need about 20V.

Would a 24 volt DC power supply be ok? I am looking at a couple 120v AC to 24v DC 1A wall warts on ebay. Option 1 & Option 2

For answers to the other 2 problems you need to know the charge current.

If I use either of the power supplies I linked above I could do a 1A charge current, correct? C/2 would require 1A charge current if I am reading the datasheet correctly:

iFast = (capacity of battery in mAh) / (charge time in hours)

iFast = 2000mAh / 2 hours
iFast = 1000mA

Place the power supply under this load and measure the voltage. (this needs to be at least 20V -- see above). This is the max voltage under load (it is also the min voltage under load if the supply is well regulated and not variable).

Any suggestions as to how best to put the power supply under a 1A load prior to building the charger? I have a voltage tester, but I can't think of anything off-hand that i own that requires 24v 1A power.

Thanks!

 

[(*steve*)]

24V should be fine.

2 hour charge for 2000mAh battery is at approx 1A. yes

I would advise a 25V 1.5A power supply (or reduce charge to 750mA) so as not to push the power supply to its limits.

1A load for a 24V power supply would be a 24V 24W light bulb (or two 12V 12W bulbs in series)

 

[Jack.Straw]

Awesome, thanks Steve!

 

[Jack.Straw]

One more clarification if you could. I'm second guessing the way i'm supposed to calculate the value of R1. From the datasheet:

4) Choose an external DC power source (e.g., wall cube). Its minimum output voltage (including ripple) must be greater than 6V and at least 1.5V higher than the maximum battery voltage while charging. This specification is critical because normal fastcharge termination is ensured only if this requirement is maintained (see Powering the MAX712/MAX713 section for more details).

6) Limit current into V+ to between 5mA and 20mA. For a fixed or narrow-range input voltage, choose R1 in the Typical Operation Circuit using the following formula:

R1 = (minimum wall-cube voltage - 5V)/5mA

I was thinking 24v would be the "minimum wall-cube voltage" value i should use since I am using a 24v regulated supply. However, i found an example on the web from someone who has built a charger using this MAX12 chip and they used the value calculated in step #4 above (the minimum wall-cube voltage required for the supply). Which do you think it is?

R1 = (24v - 5v) / 5mA = 3.8k ohm

or

R1 = (19.5v - 5v) / 5mA = 2.9k ohm

Also, do you think it's necessary for the resistor to be exact? I mean, those are non-standard values. Should i just choose the next lower standard resistor, or should I go find a 5 band resistor of the exact value?

Thanks again,
-Scott

 

[(*steve*)]

Well, the question is, what voltage do you see when you load up the power supply?

This resistor needs actual voltage. Also you need to know something about ripple, so looking at the output on a scope would be good.

No, the resistor value doesn't need to be exact. But it shouldn't be too high.

The current needs to be less than 20mA, and greater than 5mA. Use a 2.2k or 2.7k resistor and you'll be fine -- even if your supply has *lots* of ripple.

 

[Jack.Straw]

I haven't had a chance to test the power supply yet as i'm still waiting for it to come in the mail. I was hoping not to have to do that test, but if you think it's important I will. If you're saying 2.3 to 2.7, then you're thinking my calculation for R1 should be based on the value obtained in step 4, and not the actual minimum voltage of the power supply actually used?

 

[(*steve*)]

No, the resistor value must be chosen to guarantee at least 5mA, but no more than 20mA to the IC (which appears to contain a 5V shunt regulator).

The calculations they give are for the absolute minimum 5mA. It makes sense to use a somewhat smaller resistor so that *if* your power supply turns out to be more poorly regulated, the circuit will still work.

It would make sense to use a resistor right on the bleeding edge if we were out to get the minimum possible power use, but given the charging currents, an additional 5mA won't be noticed.

 

[Jack.Straw]

That makes sense now, thanks Steve.. i really appreciate it

 

[Jack.Straw]

Thermistor

Sorry, another question. I don't know if i can trust the voltage slope to stop this thing before it heats up my batteries, so I want to add temperature control. The datasheet shows 3 thermistors used; 1 reading the batteries, and 2 reading ambient temperature. I want to get rid of the ambient temperature thermistors and replace them with standard resistors to achieve a hard cut-off temperature, regardless of what the temperature is in the room. I did some searching and came up with this diagram:

The author (from this article) says:

First choose a thermistor whose R0 is about 10kOhm, not to exceed 100kOhm (this is required by the inputs of the ‘712). It makes logical sense to pick a value of R2 that is the same resistance as the thermistor’s nominal resistance to make the voltage on the TEMP pin 1V. For illustration I’ll use the muRata NTSA0XV103, a resin coated, twin lead NTC thermistor with nominal resistance of 10k and B=3900. We’ll choose R2 to also be 10k. With these values and desired temperature limits we can determine what voltages to apply to THI and TLO.

Let’s start by finding the voltage THI equivalent to 40 degrees C (313 K). Taking room temperature to be 300 Kelvin, plugging everything in to the equation for R(T) we find that the warm thermistor resistance will be about 5.8k. A simple solution is to match R3 to T1 (when warm) and R4 to R2. Since the warm value of T1 is likely to be something obscure, you can always just match the two divider networks instead by choosing R3/R4 = T1/R2 , so long as the values of the R3 through R6 stay in the 10k to100k range. The low temperature cutoff is calculated in the similar fashion.

I have some 10k thermistors, Epcos part number B57863S0103F040 (datasheet here).

Using the Epcos NTC R/T Calculation tool for that thermistor shows that the nominal resistance at 40 degrees C is around 5.3k ohm. So if i'm going by the author's instructions, R3 would be somewhere around 5.3k ohm and R4 would be 10k ohm (or a combination that adds up to the same amount?). The authors example has him using a 5.8k for R3. However, a couple scentences later he says all the resistors need to be between 10k and 100k. The MAX712 datasheet also says those resistors need to be at least 10k. What am i missing here?

Thanks for all your help,
-Scott