Charge Controller in Street light getting damaged

[Mihir2308]

Dear All,

I have installed a PWM Solar Charge Controller for a off-grid street light system. Where specs are as follows: LiFePO4 25.6V Battery 60 Ah, Streetlight = 30W (rated to operate at Vin: 16-32 V DC), Charge Controller I have made is PWM charge Controller which is designed as per 24V System i.e. Panel's Max Voc= 44V DC. Panel used = 200W Voc= 44V, Vmp= 36V, Imp= 5.5 Amps.

My PWM charge controller has Transistor and Zener Diode based Linear Voltage Regulator to generate 15V stable voltage from Panel's Connection (0-44V DC) for using it in Gate Control of the Charging Mosfet (N channel Mosfet Vgs≤ 20V). Mosfet Vds= 100V, Id= 110A.

NPN Transistor Used for Voltage Regulation= BD139 i.e. 80V (Vcbo & Vceo) 1.5A (Ic) NPN Transistor. Zener Diode = 15V / 1W.

The Voltage Regulator (of Transistor) has 6.8k Ohms resistor in series after output of +15V before it goes to switching circuit for Mosfet. Thus the power rating of transistor is always under limit. The Base Resistor of NPN Transistor is 20k Ohms where Zener is attached in reverse bias to ground (Panel -ve).


Now this system is working well in other locations from almost last 1.5 to 2 years but in one system the transistor which regulates 15V is getting burnt repeatedly. Only strange situation is, that suspected system is close to a High Voltage Transmission Line which is passing in near vicinity of the garden where the system is installed.

Also what I have noticed it that my Charge Controller don't have MOV (Varistor) of 45-60V DC which many manufacturer use at the Panel +ve & -ve connections in charge controller to withstand / bypass the surge from Panel Side. I am unaware of this type of surge from Panel side in off-grid system.

If anyone has some data regarding same or any International Standards to test the charge controller with respect to surge from Panel Side then please share the same.

For now I am not considering the impact of the HV line passing nearby but is there any other possibility of the transistor getting damaged ?.


I know it is an electronic cum Solar based question, I am not sure if this should be in thid forum or not but I would appreciate your valuable input, Thank you in advance.

 

[MisterSandals]

For now I am not considering the impact of the HV line passing nearby but is there any other possibility of the transistor getting damaged ?.

Have you tried shielding this charge controller to rule this out?

 

[Mihir2308]

Have you tried shielding this charge controller to rule this out?

Thank you very much for your suggestion.

But Shielding as in ?

 

[MisterSandals]

But Shielding as in ?

Electromagnetic shielding - Wikipedia

en.wikipedia.org

 

[AntronX]

the transistor which regulates 15V is getting burnt repeatedly.

Does it have a heatsink? With 6.8k series resistor it will dissipate 200mW at 44V input.

 

[Hedges]

Possibly the transmission line induces 60 Hz in the loop formed by PV panel. That would be a continuous signal, higher when higher power carried. Check for it with a scope.

Transient loads might cause higher frequency and voltage coupling. That is where protection devices might help. Clamping PV+ to PV-, also to ground.

In the event you do measure 60 Hz riding on PV panels (voltage, or current which goes into capacitor and doesn't develop much voltage), consider making a loop in the wires that cancels what PV panel picks up. Effectively a 2-turn twisted pair, one "twist" is the loop formed by panel, the other is the loop formed by wires.

 

[Mihir2308]

Electromagnetic shielding - Wikipedia

en.wikipedia.org

My housing is of Aluminium Diecast (Glass in front) so i guess probably it behaves as shield for the charge controller. I might be wrong too. I will consider and check about same.

 

[Mihir2308]

Does it have a heatsink? With 6.8k series resistor it will dissipate 200mW at 44V input.

Hello Mr. Antron,

My Transistor is feeded by 44V (worst case) and Output= 15V, 6.8k is after emitter (which is at 15Volts), thus

Power Dissipated in Transistor (Neglecting base current flow) = (Current Through Collector - Emitter) * (Vin - Vout)

(CE Current= 15V/6.8k = 2.2 mA)

power in transistor=
= (29Volts) * (2.2mA) ≈ 0.06 Watts

please correct me if I am wrong.

 

[Mihir2308]

Possibly the transmission line induces 60 Hz in the loop formed by PV panel. That would be a continuous signal, higher when higher power carried. Check for it with a scope.

Transient loads might cause higher frequency and voltage coupling. That is where protection devices might help. Clamping PV+ to PV-, also to ground.

In the event you do measure 60 Hz riding on PV panels (voltage, or current which goes into capacitor and doesn't develop much voltage), consider making a loop in the wires that cancels what PV panel picks up. Effectively a 2-turn twisted pair, one "twist" is the loop formed by panel, the other is the loop formed by wires.

Dear Mr. Hedges,

Thank you very much for your suggestion.

I probably have missed this, will try to take the scope to the site and check the same.

 

[Hedges]

Just focusing on the transmission line being only difference. Of course, could be marginal circuit and only this instance happened to be over the limit. You could stress test another copy on the bench, margin it with higher voltage and current. Check temperatures.

Reading the comments, I now realize it is the bipolar regulator circuit, not the power transistor. Should be easy to RC filter what it sees from power supply. Unless there is something coupled back from MOSFET (I don't have a good idea of your circuit in my head.)

I've used Zener regulation and MOSFET switching, not both in same circuit and haven't built what you're doing.

 

[Mihir2308]

Just focusing on the transmission line being only difference. Of course, could be marginal circuit and only this instance happened to be over the limit. You could stress test another copy on the bench, margin it with higher voltage and current. Check temperatures.

Reading the comments, I now realize it is the bipolar regulator circuit, not the power transistor. Should be easy to RC filter what it sees from power supply. Unless there is something coupled back from MOSFET (I don't have a good idea of your circuit in my head.)

I've used Zener regulation and MOSFET switching, not both in same circuit and haven't built what you're doing.

True Sir, actually I have tested all possibilities in my lab, i intentionally increased temperature to 100+⁰ C by external heater which actually would never go in real life. Plus I also tested my system with Voltages upto 60V at Panel's connection. Still that component didn't failed like it does on site. Tomorrow morning I will share the schematic of the circuit in discussion here.

Thank you very much for your time Sir.

 

[AntronX]

power in transistor= (29Volts) * (2.2mA) ≈ 0.06 Watts. please correct me if I am wrong.

No, you are correct. Can you post a schematic of this voltage regulator and how it connects to PV input? I wonder if you may have RF oscillation.

 

[Mihir2308]

No, you are correct. Can you post a schematic of this voltage regulator and how it connects to PV input? I wonder if you may have RF oscillation.

Thank you very much for your patience & helping me out.

I missed attaching schematic today, tomorrow after reaching office I will share the same here.