Heat Sinks - How hot is too hot?

[Dragon Tamer]

We've descused that issue before, the problem is that if the voltage drops too low, the 712 can't sense the current flowing into the battery. For my circuit, the supply voltage needs to be 1.5+(1.9x # of cells). So the voltage can't drop below 16.5V, which the current supply voltage is about 17.3V

 

[Mikebits]

We've descused that issue before, the problem is that if the voltage drops too low, the 712 can't sense the current flowing into the battery. For my circuit, the supply voltage needs to be 1.5+(1.9x # of cells). So the voltage can't drop below 16.5V, which the current supply voltage is about 17.3V

Contrary to what you might think, not everyone here stays up to date with all your post, so I do not know what previous discussion you are referring to. With that said, any battery charging circuit using the MAX712 that gets so hot that it requires a backup cooling fan is in serious need of reevaluation.

 

[audioguru]

any battery charging circuit using the MAX712 that gets so hot that it requires a backup cooling fan is in serious need of reevaluation.

The TO-220 power transistor limits the current to about 1A and dissipates 6.3W. It needs a heatsink capable of allowing the dissipation of "only" 6.3W without using a fan.
A TO220 transistor can dissipate 30W if the heatsink is big enough.

 

[Oznog]

I don't think this plan is viable. The chip is a FAST CHARGE chip, and these are NiMH batts,

The problem is that NiCd/NiMH are difficult to detect end-of-charge, because the voltage doesn't necessarily rise when full. What happens is there's a slight DROP in voltage when it reaches "full", but it's kind of a transient phenomenon.

The problem is, hooking up a load AND this charger in the same device can cause it to "just barely" drop enough to engage in fast charging again. The dv/dt drop may not show up, and it could blindly charge the batteries thinking they're low, not full, until the temp sensor realizes the batts are overheated and shuts down (hopefully). But it'll murder the batts to do that, because, starting from a cold state and using a fast-charge current, they'll take a significant overcharge before they get so hot as to trip the temp detection. It has to heat up the entire cell to trip that.

This is why you'll rarely find anything but a trickle charger integrated into an actual device which USES NiMH. I'm not gonna say it can't work- but it's risky, and this device doesn't seem to be designed for it.

 

[Dragon Tamer]

the voltage doesn't necessarily rise when full. What happens is there's a slight DROP in voltage when it reaches "full", but it's kind of a transient phenomenon.

The voltage drop is what the chip uses to sense when it should stop charging. Once the voltage drops by about a half a volt, the 712 can sense the drop and "knows" when the battery is at full cappacity.

The problem is, hooking up a load AND this charger in the same device can cause it to "just barely" drop enough to engage in fast charging again. The dv/dt drop may not show up, and it could blindly charge the batteries thinking they're low, not full, until the temp sensor realizes the batts are overheated and shuts down (hopefully). But it'll murder the batts to do that, because, starting from a cold state and using a fast-charge current, they'll take a significant overcharge before they get so hot as to trip the temp detection. It has to heat up the entire cell to trip that.

It can be a little tricky at first, when I made my circuit around the battery charger, everything that runs off of battery power is given a voltage that is slightly higher than the maximum voltage of the battery that way the circuit can function properly, and still charge the battery.