AAA NiMH Backup Battery Pack for 12V Automotive Use

[JDW1]

JonSea, thank you for explaining that your previous post was not specific to Nigel's previous post (which has nothing to do with "NiMH in cars is not a good idea").

Please forgive me again, but I have not seen there to be a "general consensus" in our discussion thus far that "NiMH in cars is not a good idea," especially when Texas Instruments has a technical document dedicated to that very topic:

https://www.ti.com/lit/an/slua843/slua843.pdf

Not to give more credit to TI than members of this fine forum, but the above PDF really does serve to indicate that at least one person out there must have done similarly to what the project set forth in my opening post intends to accomplish. Toyota has used NiMH tech in their hybrids for over 18 years, but of course that technology is more sophisticated than the off-the-shelf AAA Panasonic NiMH Eneloops being discussed in this thread.

So why did I even post here? The answer is simple: in hopes of gathering experienced opinions. That is not to say the information so kindly provided by all of you thus far has not proven helpful. It is helpful. And I certainly appreciate every single reply. I take all information to heart, even information about the merits of alarm systems in general. I am simply curious about how similar projects have panned out for others, and perhaps what charging and discharging design methodology was used in those real world applications.

 

[Pommie]

Looking at that datasheet, it doesn't state a max temperature but the temperature range that the parameter is valid for. Why don't you build an oven and try charging/discharging at elevated temperatures to find out what the capacity is? I suspect that at 80C it will be greatly reduced but still usable.

Mike.

 

[JDW1]

Why don't you build an oven and try charging/discharging at elevated temperatures to find out what the capacity is? I suspect that at 80C it will be greatly reduced but still usable.
Mike.

That's precisely what will need to be done, Mike. Thank you for the suggestion. Even knowing that, I still wanted to post in this forum to see if anyone had practical experience doing something similar to my project, perhaps providing data or insight into how such devices pan out, not merely for short temperature tests, but over the long haul (months of use and repeated exposure to hot and cold extremes).

Again, thank you.

 

[rjenkinsgb]

rjenkinsgb, I am not sure why you cannot Google up any NiMH battery datasheets, but here is one of many that are conveniently available, which shows exactly what I said previously. More specifically, there are temperature ranges for Charging and Discharging, storage, etc. on the following datasheet. Keep in mind this datasheet is for the INDUSTRIAL spec (OE sales) of the Panasonic NiMH battery line, which has wider acceptable temperature specifications than your commonly available consumer-grade Eneloops.

**broken link removed**

In response to what you wrote, whether "charging occurs when the vehicle is running" depends completely on the design of whatever it is we are talking about. Indeed, I could design a system that charges with the ignition ON but the engine off, or I could design a system that charges with both the Ignition and engine off, although in that case I would need to take care to keep average current consumption low, or charge at a few hundred milliamperes for only a short duration and then stop.

I've seen those, they are a very abbreviated outline spec and not what I consider a fit datasheet to base a serious design on. No characteristic graphs and nothing that can be used to extrapolate characteristics over a range.

it doesn't state a max temperature but the temperature range that the parameter is valid for.

Exactly what I'm saying.


See this info - much more useful and practical:
https://data.energizer.com/pdfs/nickelmetalhydride_appman.pdf

I based my passive charge design on that; from the info there you are far more likely to have problems at _low_ temperatures than high.

At temperatures where the cells themselves can no longer function, high or low, the charge circuit becomes irrelevant. Those conditions must be avoided by overall system design or you need a different battery typo.


Re. charge control, I'm still referring to the characteristics of my passive three diode plus one resistor method. That is self limiting and self regulating, if you consider variations in vehicle battery voltage under stopped and running conditions.

KISS principle - there is no need for anything more complex.

Plus, practical tests are essential. Trying to make a design for production from data sheets alone without functional tests is a recipe for disaster (or bankruptcy).
Get some cells and _test_it_ before dismissing.

 

[JDW1]

rjenkinsgb, thank you for the useful PDF. It confirms what I have found that your average consumer grade NiMH AAA or AA battery is best used at the temperature range of 0°C to 50°C (page 13 of that PDF, Device Design Considerations, Operating). Of course, that is merely their suggested general rule.

With regard to your 3-diode + 1 resistor passive charger circuit, what if you have 1 out of 10 (AAA NiMH battery) that has a problem during charging? Maybe it is already full whereas the other 9 are depleted. Maybe that 1 AAA simply won't charge for some reason and gets hot during charging. How does your circuit deal with those not-so-rare cases? (I don't have a schematic in front of me with your design, so I am trying to figure out if that KISS approach is really best in this case.

We must also bear in mind that this battery backup device must not only charge but it must also discharge at times when the vehicle battery voltage falls below a predetermined threshold. And in that case discharge, you don't want any 1 of the 10 batteries to be depleted below a certain predetermined low threshold point, lest that overly depleted AAA battery go into reverse polarity (which is often the case with over-depleted NiMH cells). So to prevent that it would seem that a super-simple passive circuit may not be the complete solution here. Even so, I am very open to hearing your thoughts on this.

Once again, thank you for continuing to share your opinions and ideas.

 

[Pommie]

Every nimh power tool ever made had cells connected in series without any kind of balancing circuitry. I think you're over thinking this.

Mike.

 

[rjenkinsgb]

Under trickle charge conditions, NiMH cell batteries are self-balancing. The charge current has no effect on any cell that's already fully charged so only affects those still needing charge.
With the pack fully charged the trickle current should be a few tens of milliamps so power dissipated as heat per cell a few tens of milliwatts; eg. 25mA trickle means around 40mW heat per cell, not enough cause any significant heating - imbalance and overheating are "fast charge" problems.

A permanent (or near permanent) low trickle setup avoids all the common problems with NiMH (or NiCD) cells and actually extends their life over cyclic use.

As mentioned earlier, the load current would be drawn from whichever battery has higher voltage at any instant.

As the NiMH battery voltage can never drop significantly below the vehicle battery voltage, over discharge should be totally impossible.
Under normal conditions, drops below 10V should only ever occur during starting and that's what the battery pack is intended to ride through.

The only event that can cause that over-discharge is the vehicle battery going flat or being disconnected, for long enough to drain the pack via the alarm standby current or if the alarm is sounding.

That is a condition where the owner should be advised to replace the backup cells as a matter of course.

For production "belt & braces" safety, I'd add a couple of polyswitches (PTC fuses), one in series with the cell pack to prevent excess current in case of a wiring fault and one in the diode+resistor charge path to limit charge current in case the pack has been totally discharged. Put a higher value fixed resistor across that to provide some charge to gradually bring the pack back to normal voltage.

For battery condition:
A load resistor (to put say half an amp load on) in tandem with a LED plus resistor and zener, so it only lights above 12V, plus a momentary push button to connect the circuit directly across the battery pack as a manual "battery test".

That should show all cells are functional and can provide load current, so a freak cell failure can be detected.

You could also add status LEDS or a bar graph, operating only while the ignition is on, for visual feedback and to increase the perceived value of the unit - though is that a good idea with something that's a part of a security system? It shows a thief what to disconnect...