I have a device which is rated at 5V 300mA and power is supplied via a USB cable.
I want to add a battery backup using 4 X AA niMH rechargeble batteries (ie 4.8V).
I have a simple circuit I have used for another application however the supply voltage for that was much higher than the rechargeable battery (see attached).
Is it ok to connect the 5V from USB directly to the rechargeable battery (as there is almost a perfect voltage match) or do I need a resister (as per the attached circuit) and if so what value.
Thanks

 

Update on circuit for clarification

 

Your main problem is that NiMH batteries absolutely do not like being left to trickle charge. Even if you get the current right and don‘t over heat them, you will kil the battery by leaving it to charge continuously. If you want to use NiMH you will have ot use a lot more sophisticated approach that includes separate charging, monitoring (for self-discharge, which MiMH will do), and of course switching on power outage.

A small board like this one could be used with a lithium chemistry battery but I am not aware of any modules designed to do the same thing with NiMH which are inherently problematic for the application.

​

 

Also, even a Schottky diode will likely drop the voltage about a half volt, so the load voltage would be no more than about 4.3V.
Your load may not operate properly as that voltage.

 

The diode from the 5V USB reduces the voltage to maybe 4.3V. Then each NiMH battery cell gets only (4.3V/4= 1.075V) which is dead. A Ni-MH battery cell charges to at least 1.4V then the charging must be turned off.

A Ni-MH cell voltage averages 1.25V during a discharge. 1.1V is dead and 1.4V is fully charged.

 

Welcome to AAC!

I have a device which is rated at 5V 300mA and power is supplied via a USB cable.
I want to add a battery backup using 4 X AA niMH rechargeble batteries (ie 4.8V).

What is the voltage tolerance of this device?

USB voltage is specified to be 4.5-5.5V. The nominal voltage of 4 NiMH batteries in series is 4.8V, but the fully charged voltage will be closer to 1.4V. Will your device tolerate 5.6V?

 

Thanks everyone.

I did not know that niMH will die when trickle charged although I had a 9V (8.4V) niMH battery (single unit) on trickle charge for at least a year (before other parts of the device being supplied stopped working for other reasons) and the battery still had 8.4V when I removed it so maybe it takes a long time to kill it? It doesn't matter to me if I kill the 1.2V AA rechargeable batteries over time as the ones I will use are as cheap as alkaline.

The device is marked 5/6V 300mAh and will also run on 4 X alkaline batteries so the voltage tolerance should handle niMH fully charged.

Yes I will use schottke diodes for a lower voltage drop.

I haven't measured the voltage from the usb (plug pack) charger. Maybe it is a little higher than 5V?

Would I be better off using alkaline batteries but how to I isolate them from the normal USB supplied 5V (and so they will automatically supply power in the case of a usb power outage?

The battery backup will (hopefully) only be required once every 2 or 3 weeks and usually only for a matter of a few minutes (we get regular overnight power outages where I live lasting usually only for a few minutes)?

 

It doesn't matter to me if I kill the 1.2V AA rechargeable batteries over time as the ones I will use are as cheap as alkaline.

Buy cheap, get cheap.
High quality Ni-MH cells cost much more than alkaline cells.
Except, cheap Chinese Ni-MH cells have 1/4 of the capacity or less of high quality Ni-MH cells. They also rust away soon.
Some cheap Chinese Ni-MH cells might be old Ni-Cad cells.

Energizer battery company have high quality Ni-MH cells made for them in Japan and they have the same modern features as Japanese Eneloop cells.
Duracell has some Ni-MH cells made in China and their capacity is half what the Energizer cells have. They cost a little less.


Energizer says you can charge Ni-MH cells over night (9 hours) at 1/10th their mAh rate or you can trickle charge continuously at 1/40th their mAh rate.
Here is what they say about overcharging:

 

Here's a circuit to isolate the battery using a MOSFET and a rail-rail op amp, that causes no significant drop when the battery is powering the load.
It may be more complex then you like, but it should allow you to use NiMH batteries with the load voltage essentially equal to the battery voltage.

In the simulation below, the op amp turns the P-MOSFET ON (gate voltage to 0V, blue trace) when the USB voltage is disconnected by the switch (green trace) and the output voltage drops below the battery voltage.
When the USB is then reconnected at 2 seconds, the MOSFET turns off and the load is again powered from the USB voltage.

The transistor can be just about any logic-level (Vgs(th) <2V) P-MOSFET with an Ron of <100mΩ.

 

Thanks
I agree with 'you get what you pay for' however the device I want to add battery backup is a VERY cheap device and I din't want to spend more on a battery backup than the device cost. The batteries I considered using are only rated at 600mAh (there were others with the same name rated 1000mAh and one rated at 2000mAh) and they are marketed as replacement for solar lights and so would only ever be trickle charged although only during daylight hours.
Thanks for the simple circuit to isolate the battery using a MOSFET which I will consider.

 

I have many cheap solar garden lights. The Ni-MH batteries they come with have such bad quality that they work for only a few months and I replace them with Energizers.

 

You can likely safely trickle charge the NiMH cells if you keep the current low, no more than .03C where C is the battery AH capacity.
If there's only a 0.2V difference between the USB voltage and the battery voltage, then the resistor value to the battery would be about 10Ω.

That would give about 100mA of charging current if the batteries were fully discharged to 1V per cell.

 

Sorry to come late to the party.

I have a power failure light in my living room that uses two Ni-Cd cells that are trickle charged in a circuit similar to your circuit in post #1. Been there for 16 years and the light still comes on after power goes out. After 17 years of service they have lost a lot of their capacity but they still work.

You might consider whether Ni-Cd cells would make you life simpler.