Request Guidance: NiMH Low-Voltage Cut-Off

Thread Starter

phototron

Joined Nov 22, 2013
22
Hi everyone. I'm new here, but optimistic. I need guidance for a project that requires a low voltage cut-off for a NiMH battery.

The project uses a NiMH battery of 4 AA cells in series to power a simple load of resistors to produce around 1.5 - 2 watts of heat. The heat will help prevent a hummingbird feeder from freezing in temperatures as low as -5°C. Here in Victoria, BC, and in other parts of the Pacific Northwest, the Anna's hummingbird doesn't migrate. Because they are at their limit in the winter nights, at the first light before sunrise they go to their usual feeders and recover from the torpor.

In the past years I had used 5 watt night lights and a layer of aluminum foil to keep the feeders from freezing, but where I live now there are no outdoor outlets. I would make an inductor to transfer electricity through the single-pane basement window, but I don't yet know how to do something like that. I plan to go to college here for an electrical engineering program, but in the mien time there are cold hummingbirds.

I figure with a modest amount of heat from resistors (designed to dissipate the heat) and a foil insulation wrap, I should get a few hours of operation on one set of batteries before I switch them for freshly charged ones.

Back to the task at hand, I need a way of preventing the batteries from over-discharging, as I don't want them to be damaged. I saw this thread but I am unsure which resistor values to change to set the cut-off points. I also saw this forum thread where an IC is mentioned, the MAX8211, however when I look at the spec sheet it looks like it will detect the voltage, and will include hysteresis, but I don't know how to use its output to disconnect the load. Does it simply send a digital signal to be used by a processor? I can't find what NMI means on the output.

If anyone could help me better understand the circuits I've seen I would be grateful. Thanks in advance for any guidance.
 
1.0 V is the NiMH Low-voltage Cut Off for a single cell. For a pack of 6 cells in series, the NiMH Low-voltage Cut Off is 6.0V.
Incidentally, the NiCd Low-voltage Cut Off is 0.85 V/cell
 

wayneh

Joined Sep 9, 2010
17,496
... but I don't know how to use its output to disconnect the load. Does it simply send a digital signal to be used by a processor?
The control system will generate a voltage, either high or low, when the load needs to be disconnected from the battery. This signal will control the voltage on the gate of a MOSFET acting as a switch for your heater. For an N-type, high voltage is switch on, low voltage is switch off. It may be necessary to invert the logic of the signal, or use a P-type, to get the MOSFET to behave the way you want.

Does that help?
 

Thread Starter

phototron

Joined Nov 22, 2013
22
The control system will generate a voltage, either high or low, when the load needs to be disconnected from the battery. This signal will control the voltage on the gate of a MOSFET acting as a switch for your heater. For an N-type, high voltage is switch on, low voltage is switch off. It may be necessary to invert the logic of the signal, or use a P-type, to get the MOSFET to behave the way you want.

Does that help?
Thanks Wayneh, that makes sense. I have not used a transistor as a switch before. When I look at the specifications of transistors, do they describe the voltage required to switch them? Is that control voltage supposed to connect to the collector or the base? The emitter is for the load, yes? Does the MAX IC spec sheet describe what voltage it sends as a control? I will look further.

I will try to find information about using a transistor as a switch. Things like control voltage, current drain / efficiency (not a matter in a circuit designed to be a heater), and what connections are to go to which pins.

Thanks for the help!
 

Thread Starter

phototron

Joined Nov 22, 2013
22
1.0 V is the NiMH Low-voltage Cut Off for a single cell. For a pack of 6 cells in series, the NiMH Low-voltage Cut Off is 6.0V.
Incidentally, the NiCd Low-voltage Cut Off is 0.85 V/cell
Thanks Inspironator. I thought that NiCd cells were safe to go to 0 volts without damage. Is that not the case? Or is that cut-off the point at which they simply produce too weak a current to be useful?

I decided against NiCd due to the much lower energy density.

Thanks for your help!
 

wayneh

Joined Sep 9, 2010
17,496
I will try to find information about using a transistor as a switch. Things like control voltage, current drain / efficiency (not a matter in a circuit designed to be a heater), and what connections are to go to which pins.
Look first at a MOSFET, a special transistor that is excellent for switching big current loads easily and cheaply. The only reason to NOT choose a MOSFET is that they require a higher voltage on their gate (analogous to the base of a BJT) to turn fully on. A logic level MOSFET might need 3V and a regular MOSFET needs ~10V. A BJT needs only ~0.6V and a darlington (2 BJTs) needs ~1.2V.

Why not just use the BJT? Because they require a base current of ~10% of the load current. A MOSFET requires none, only the voltage. This makes them so very useful in amplifying a control voltage at a tiny current into a large current for your load.
 

MikeML

Joined Oct 2, 2009
5,444
Here is something I posted (somewhere) a few years ago. It has a fewer parts count than most other circuits I have seen. It includes a high-side switch, in this case a PNP, but if the Vce(sat) voltage drop is too high, it could be replaced with an PFET. This one turns on automatically, but a toggle switch in the Bat+ line would fix that.
 

Attachments

Thread Starter

phototron

Joined Nov 22, 2013
22
Here is something I posted (somewhere) a few years ago. It has a fewer parts count than most other circuits I have seen. It includes a high-side switch, in this case a PNP, but if the Vce(sat) voltage drop is too high, it could be replaced with an PFET. This one turns on automatically, but a toggle switch in the Bat+ line would fix that.
Thanks for the example circuit, MikeML. Is the TL431 an SCR? I am looking at specifications for it. The 2N3906 transistor seems to have a current limit of 100mA, is that right? In this circuit, is the load passing through the transistor?

Does an increase in R3 result in wider hysteresis?

Thanks for the suggestions.
 

Thread Starter

phototron

Joined Nov 22, 2013
22
I found a PFET, part no. IRLML2060TRPbF. The data sheet says the Gate Threshold Voltage is min. 1 and max. 2.5. It also says the Continuous Source Current is max. 1.2.

If I were to dissipate 2 watts with a 4.8v battery the current would be about 416mA, so I should be safe there.

There were many other PFETs listed, I simply selected the top one in the list on the Digikey search. I have used Newark's Canadian site in the past but they are down for upgrades this weekend.

With a PFET I would be able to eliminate R5 from MikeML's circuit, correct?
 

MikeML

Joined Oct 2, 2009
5,444
The PFET should be a "logic level" type. I revised the circuit to utilize a PFET, increased the load current, tweaked the voltage up, and show the effect of varying the feedback resistor R3. Note that the hysteresis adjustment and voltage adjustment will interact, so you might have to do a couple of iterations until you get it right.

The LM431 (TL431) is a very useful part. You might as well as buy 10...
 

Attachments

Thread Starter

phototron

Joined Nov 22, 2013
22
Thank you for the modified circuit, MikeML. I'm looking at the spec sheet for the FDS9933A and am puzzled by one of the specifications. It says the total power dissipation for single-unit operation is a max of 1.6W. In this circuit, is the current drawn by the load going through the PFET?

I am a little hazy on how the circuit cuts power to the load. It looks like there is always a connection between battery and load over the variable 50k resistor and the fixed R3. Please forgive my ignorance. I have a steep learning curve while not yet in school.
 
Back to the task at hand, I need a way of preventing the batteries from over-discharging, as I don't want them to be damaged.
Fully discharging NiMH cells, even all the way down to zero volts, doesn't hurt them. It's reverse charging them that causes damage. See:

http://www.camlight.com/techinfo/whydischarge_4.html

You don't need any circuitry to turn off your load. Just connect the 4 AA cells in parallel instead of in series. Then you can use a lower resistance load, with value chosen to give the desired heating.

With the cells in parallel, any cell of lower capacity than the others is not in danger of being reverse charged. You can allow them to fully discharge, which won't damage the cells.
 

Thread Starter

phototron

Joined Nov 22, 2013
22
Ok, one more try. This one has uses a momentary SPST button to start it. Current after it cuts off is leakage only. Since it has a manual start, no hysteresis is needed.
Thank you for the new circuit. I like it's simplicity, but I am interested in trying the first version as well. Would it be wise to use a pot for R3 to adjust the hysteresis?

Can you offer insight into The Electrician's opinion that NiMH is safe to drain to 0 volts, and that as long as the pack of cells is in parallel configuration, the cells won't be reverse-charged and will remain undamaged?

For the initial circuit, just to get the heater installed while I work on a low-voltage cut-off as per your suggestions, it would be nice if the parallel configuration was safe.

I definitely want to learn how to make the cut-off though, as having a higher voltage and a lower current would be useful for this project and would give me the option of a way to monitor charge level with an LED. I would also find the cut-off very useful for other projects that I want to use multiple NiMH cells on, such as LED lanterns. I have 4 new LSD NIMH Sanyo Eneloops that I'm itching to try out! I have tested them with my La Crosse Tech BC1000 charger and they appear to have the stated capacity.
 

Thread Starter

phototron

Joined Nov 22, 2013
22
Fully discharging NiMH cells, even all the way down to zero volts, doesn't hurt them. It's reverse charging them that causes damage. See:

http://www.camlight.com/techinfo/whydischarge_4.html

You don't need any circuitry to turn off your load. Just connect the 4 AA cells in parallel instead of in series. Then you can use a lower resistance load, with value chosen to give the desired heating.

With the cells in parallel, any cell of lower capacity than the others is not in danger of being reverse charged. You can allow them to fully discharge, which won't damage the cells.
Thank you for the information. Does this mean all Nickel cells are safe to discharge to 0 volts? I'm thinking of NiFe, but maybe NiZn as well?

A parallel set-up might be the ticket to getting the heater running as soon as possible while I learn how to build a cut-off circuit. A cut-off would be very useful for other projects, plus it would be useful to have a high-enough voltage on the heater circuit to power a simple voltage monitor, such as an LED.

I am also concerned about AA holders and current-capacity, as a 1.5 watt heater will require 1.25A at 1.2v. It does make me think that, if it turns out a higher wattage is required to keep the feeder from freezing, using 8 cells of the same capacity in parallel would be possible, if not too bulky.

Thanks for the idea!
 

Thread Starter

phototron

Joined Nov 22, 2013
22
Also, it occurred to me that - at the risk of over-complicating an already challenging project, a temperature-based control might be good, to prolong the limited life of the 4AA battery by only turning the heat on when the temperature is low enough, such as when it reaches 1 degree.

KISS, right?
 
Thank you for the information. Does this mean all Nickel cells are safe to discharge to 0 volts? I'm thinking of NiFe, but maybe NiZn as well?
If by NiFe you mean the old Edison cells, I believe they are quite tolerant of abuse, so a deep discharge may be ok. NiZn cells? I don't know about them.
 

MikeML

Joined Oct 2, 2009
5,444
With respect to discharging a pack consisting of several series cells, if the criteria is to prevent the weakest cell from reversing polarity due to the load current, then one way is to monitor the pack voltage, and stop discharging when the total pack voltage drops. Otherwise, you would need as many voltage monitors as there are cells in the pack.

Here is the discharge curve for 2000mAh NiMh cells, reproduced below.

In a four-cell pack, if you use a cutoff voltage of 0.95V per cell, that would mean a pack cutoff voltage of 3.8V. If the pack contained one weaker cell and three healthy cells that still had 1.1V per cell, the load cuttoff would occur just as the weak cell reaches 3.8-3*1.1 = 0.5V, well above "reversing". This has worked for years...
 

Attachments

Thread Starter

phototron

Joined Nov 22, 2013
22
That discharge graph looks like the West Mountain Radio load tester software. I used to have one of those. I was amazed at how much current a AA NiMH cell would deliver - over an amp for over an hour.

I will try to build the circuit you described and showed me, MikeML. Newark's site is back up now so I should be able to order the parts I need from them. I don't know who is the best supplier of parts, but the local shop here is expensive by comparison so I must order from abroad.

I will try the manual start circuit first on my breadboard, then try the one with hysteresis. You recommended a 50kΩ pot for the cut-off adjustment. Would you recommend a value for a pot for the hysteresis adjustment?

Another question. It looks like the load goes through the PFET. The spec sheet says a max of 1.6 watts for single operation. Does that mean it can only handle 1.6 watts running through it? There is a chance I will need to use a load of 2 watts or slightly more.

Thank you again for all your help!
 
Top