Simple Battery Discharger

[Quank]

Hi everybody!
I'm trying to build a battery discharger. The main goal is to discharge a battery LiPo of 40AH in aprox 4 o 5 hours (at aprox a 15A pace!).

The battery gives 3.7 V aprox when is charged and decreases its voltage as will discharging.
The nature of the battery implies not to discharge beyond 2.8 V (to preserve charging properties).
The circuit I've attached in this thread connects the battery to 2 power resistances, of 0.5 Ohm each, in parallel. In order to disconnect the battery when voltage is too low, I use a transistor and 2 relays. The transistor is switched when the comparator detects that the voltage is too low. The comparator uses a diode to set the voltage reference and a voltage divisor for comparing the voltage.
I built the circuit and it works perfectly when no resistances are connected. When I connect the resistances the relays start to vibrate a lot when the voltage is upper than 2.8 V. How I can prevent relays vibration?
Thank you in advance.

Quank

 

[Reloadron]

What is U1B actually? Place a meter across the battery and note what is going on with the battery voltage. The comparator has no hysteresis. Where did that circuit design come from? You load the battery and the voltage drops, when that happens the load is removed, when removed the voltage increases. That is my guess.

Ron

 

[Quank]

Yes, I'm sorry, U1B is STMicroelectronics LM293N Comparator in a 8-DIP encapsulation. You can get its specifications in the link:

**broken link removed**

Yes, the comparator has a little hysteresis (see specification). Is does not enough?
The circuit came from my head (maybe this is the problem).

How can I avoid this transition in voltage? Is the capacitor not correct?
Thanks

 

[Reloadron]

You are starting with a battery voltage of about 3.7 volts. That same voltage powers the circuit as well as the relay coils. What is the correct relay coil voltage?

Actually LiPo batteries do not need discharged for any reason I know of so why discharge them? They do not require discharge prior to charge, at least I don't think so.

Ron

 

[Quank]

Ron,
The relays are Omron G5LE-14, with 3VDC rated voltaje and 10 A. You can see specifications in:

https://www.digikey.es/scripts/DkSearch/dksus.dll?x=0&y=0&lang=en&keywords=Z3326-ND

without the resistances of 0.5 Ohm the circuit works pretty well. In fact I'm not connecting the circuit to the battery to test it, I use a laboratory power supply.

The reason for building a discharger is for practical purpouses. I'm designing a charger and an autonomous system that uses these batteries, and with 40 AH capacity, I need a lot of time to discharge them!! I need a discharger for testing the charger more often.

Thank you

Mark

 

[ronv]

I think ron has it right. The battery is discharging ok until it hits the trip point at 2.8 volts. The relays open and the battery voltage "bounces back" a little to say 2.9 or maybe even 3 volts, so the relays re-apply the load. Several ways to fix it. The easiest is probably to add some hysteresis with say 100k from pin 5 to pin 7 of the comparitor. You may have to experiment with this value until you get good results. You should also add a clamp diode across the relay coil - anode to CTRL cathode to +3.

 

[Quank]

Thank you ronv. I will try to put this histeresys and the protection diode. Witch properties must have the clamp diode?
Is there any problem to put two relays with two resistances in parallel? Is possible to melt one relay? I was thinking to put only one clamp diode for both relays and not one diode for each relay..
What do you think?

 

[ronv]

The clamp diode just needs a current rating close to the relay current so a 1N4002 would work fine.
The 2 relays should work ok. One diode would be ok.

 

[Quank]

Thanks ronv!!