Dear All,

I am looking for advice on improving the reliability of my circuit.

I am using a 12.6 V, 10 A Li-Poly battery pack (3S) with an integrated protection circuit, together with the LTC4162 Li-ion battery charger configured according to the datasheet.

Occasionally, the battery suddenly stops operating and outputs a very low voltage. After some investigation, this appears to be related to the protection circuit. When measuring the individual cells directly, each cell reads above 4 V (totalling approximately 12 V across the pack). However, the output from the protection circuit can show an abnormal value (for example, around 5.6 V).

I suspect that voltage transients may be damaging or gradually degrading the protection circuit, leading to reduced reliability over time and eventual failure.

On my PCB, I generally have:

• A 12 V pump driven by a MOSFET load switch, with a flyback diode for transient suppression
• Long cable runs between the LTC4162 charger and the solar panels

These are two possible sources of transients. My concern is whether long solar leads could allow voltage spikes to propagate into the system, potentially exceeding 12.6 V for a short duration (milliseconds or less), even if charging is not actively occurring. Could such short transients bypass the charger’s regulation and stress the battery protection circuit?

I considered adding a TVS diode on the 12 V rail. However, I realised that the clamping voltage of a suitable TVS would likely be above ~12.8 V, which may already exceed the maximum protection threshold of the battery pack.

Could anyone suggest:

• A more robust protection approach for this type of system?
• Recommended methods to damp or clamp voltage spikes without overstressing the battery?
• Whether additional filtering (e.g., LC filtering, RC snubbers, dedicated surge protection, ideal diode stage, etc.) would be advisable?

These units are deployed in the field, so we cannot easily monitor transient behaviour directly. At present, we are assuming transients may be the root cause of the battery failures.

Any guidance or recommended best practice would be greatly appreciated to improve my 12V power rail or keep it clean.

 

When fighting those nasty transients it is very useful to understand what the circuit is doing, and just HOW it is doing it.
With the exceptions of lightning and physical damage, transients are CAUSED, they seldom "Just Appear".
One other thing is that the excess voltages can often be prevented from causing damage by having the system elements rated for a higher voltage. CERTAINLY that adds to the cost, but reliability and the ability to surviveare quite valuable in many instances.
Transient suppression and prevention is a fairly mature field of knowledge, and quite a bit of information is available.

 

When we carry out our lab testing, the issue does not appear. It is only in the field that, out of 50 units, one or two may be affected, and it is difficult to replicate the problem.

In terms of cost, this is not an issue, as it is cheaper than the time spent rectifying the fault. I just need some assistance in designing a circuit between the battery and the rest of the system.

 

If the TVS suppression diode is not near the TVS generating device it is possible that excessive impedance in it's connection could be allowing excess transient energy to travel back to the battery protection system. That "excessive impedance could be as simple as a single connection not adequately torqued, or possibly a single crimped connection not being adequate. That would explain why it never happens in lab checks, and only rarely in the field.

 

Correct @MisterBill2

Yes, I have placed TVS diodes near the devices, and in some cases I have also used Schottky diodes (SSB43LHE3), for example when driving higher-power devices such as the IXDI630MYI.

The issue seems to be the limited headroom of the battery protection circuit, which has a maximum of 12.8V (as shown in the attachment). My concern is that the margin is quite small, so even a transient reaching 13.5V at the battery protection circuit could potentially cause damage.

Is there a recommended circuit or filter to prevent these short transients from reaching the protection circuit?

 

If there is not much energy in the transients, possibly a simple ZENER diode clamp scheme could be used to shunt the voltage, but with a suitable FUSE between the diode and the battery protection system. In case of a zener diode failure.

 

OK, so the overall circuit will be as follows. I will prepare a proper circuit diagram shortly:
Power Vin → Battery Charger → Battery → Fuse → Zener → Application Circuit.
Should I also place capacitors, such as a 100 nF ceramic and a 100–105 µF electrolytic capacitor, on the battery side before the fuse?

 

The main purpose of the fuse that I suggested is to protect the battery in case the capacitor fails short circuited. That means that the fuse must be between the battery and the capacitors. I should have explained that before.

 

OK it will now be

Power Vin → Battery Charger → Battery → Zener -> Cap -> Fuse → Application Circuit.

 

The main purpose of the fuse that I suggested is to protect the battery in case the capacitor fails short circuited. That means that the fuse must be between the battery and the capacitors. I should have explained that before.

Is this a normal issue to expect when using a Li-Po battery with an inductive load?

How fast would a Zener diode react in the case of a transient or voltage spike? Is there a recommended Zener diode that we should use? I am thinking it should be around 13 V, as the battery’s maximum voltage is 12.6 V. Would that be correct?

Is it appropriate to place a TVS diode together with a Zener diode?

This load driver is controlling a 12 V pump/motor, so I am wondering whether the diode would be very effective in this case.

 

Voltage spikes produced by the rapid cutoff of current thru an inductance, or by coupling from an outside electrical event such as a lightning strike.
The circuit in post #10 leaves out a lot of detail, so it is not very useful. It shows neither the battery nor the motor nor where the long run of wires from the charger circuit to the battery. That makes any evaluation rather challenging.

 

@MisterBill2


Apologies if my previous post was overwhelming. The schematic contains several parts of my system, so I will break the explanation down into smaller steps and sections.

1. The previous post shows how the existing setup can power a similar pump to the following: HSEAMALL Brushless Submersible Water Pump DC 12V Amphibious Pump 800L/H 5M for Pond Aquarium Fish Tank Solar Fountain Pool Water Circulation System : Amazon.co.uk: Pet Supplies

The pump connects to Port J1 from the previous post.

2. The VCC_12V0 is provided directly from the battery


3. This is my charger. Please ignore connectors J1 and J2, and note that the charger is configured to work with 12.6 V Li-Poly batteries, not 4.2 V. The charger is a small module placed on the board as component U8.





The aim is to give additional protection to the battery BMS circuit.