overload protection circuit in inverter

[red_alert]

Glad to hear that!

Yes, one opamp is enough. Still, you might want a prealarm before shutdown (when battery is 10.5 or something) - in that case you need a second opamp, too. I'll draw & post both schematics a little bit later.

Edit: A simple voltage divider (two 10K resistors and a capacitor) it's ok as you don't need a great precision for that delay stage.

- - - Updated - - -

That's a sample circuit:



If you need precision thresholds, you might need a zenner diode (D3) to regulate the HIGH voltage of the hysteresis resistors. Same thing for the reference voltage (5.4V).

Btw, you can choose whatever Vref you want for the "5.4V" reference (if you already have a particular zenner diode or any small power voltage reference) then I would calculate those resistors accordingly.

 

[--BawA--]

Red_alert , i am using 5.1V (pin 16 of sg3525) as a reference.
How to calculate the hysteresis resistors and value of that zener diode ?

If you need precision thresholds, you might need a zenner diode (D3) to regulate the HIGH voltage of the hysteresis resistors.

please throw some light on it.

 

[red_alert]

So you want to set two thresholds: 10.2V and 10.8V. To use the reference you have (5.1V) you should put a voltage divider on your battery voltage, to translate the battery voltage around 5V at the opamp input.

Vref (5.1V) being at the middle of opamp HIGH (12V) and LOW (0V) output range, the hysteresis thresholds would be almost symetrically around this value (by example, you could get 4.9V and 5.3V) so you have to translate the battery voltage a little bit lower than half (10.8V -> 5.3V, 10.2V -> 4.9V or something like this). Anyway, I'll do the calculations for you!

please throw some light on it

The hysteresis thresholds depends on HIGH and LOW voltage levels of opamp. If the opamp has unregulated voltage, those output voltages will be variable too (well, except the LOW state - almost GND level).

So, for accurate thresholds, I put the D3 zenner diode which limit the output voltage (as seen by the hysteresis network) to a fixed value (7.5V in the example above).

Anyway.. like i said, only the upper threshold (10.8 -> 5.3) will be affected. Moreover, if I assume the opamp voltage is the same as the battery voltage, once the battery is recovering from discharge status it will reach 10.8V first, anyway.. so I could calculate the upper threshold for this opamp HIGH state value.

Nevermind, maybe it sounds too complicated. If you really want accurate upper threshold (10.8V) you better use that zenner diode.
If not, you could have a slightly variable threshold.

Btw.. if you want to manually (and independent) adjust each of those thresholds, I might add some diodes, to have two separate hysteresis networks for upper and lower battery limit.

 

[--BawA--]

thank you so much red_alert for helping me..

Btw.. if you want to manually (and independent) adjust each of those thresholds, I might add some diodes, to have two separate hysteresis networks for upper and lower battery limit.

how?


one more thing to ask , i intend to design a battery charging system for 12V, 150AH battery , using a UC3842 chip, now my question is " is it possible to change the output current using a pot? because this charger is intended to charge 12V battery of different sizes (for exp 12V , 65AH etc). ?

my second question " is it possible to incorporate the float charging and overcharge protection in UC3842 based design?

Any supporing schematics will be highly appreciated...

 

[red_alert]

About the independent threshold adjustments:



D4 and D5 helps selecting different resistors (R2 / R22) for setting the upper/lower thresholds. When opamp output is HIGH, you have R1 and R22 as hysteresis resistors (substracting the voltage across diode, of course) and when the opamp output is LOW you've got R1 and R2 as voltage divider.

About UC3842:

It has a current sense input (actually, it's reading the voltage drop across an external current sense resistor). It also have an internal 1V reference (IN- of the internal opamp comparator) so all you have to do is to adjust your external current sensor circuit to have 1V at the pin 3 input for the desired output current.

It's not usual to adjust the current sense resistor so you better use an opamp to amplify the voltage drop across that resistor (this way, you could choose a even smaller resistor, for better power efficiency) and put an ajustable voltage divider at the opamp output to feed the pin 3 (CURRENT_SENSE) of UC3842 with that 1V.

For float charging (and overcharging) you have to read the battery voltage and dynamically change the output current sensing voltage to keep the battery voltage at a steady value (14V or something).

That takes a little bit of logic circuits and it could be way easier if you could use a uC (or the cheapest arduino board). This way you could easily implement all those previous logic circuit (overcurrent protection, battery discharge) and the current one (current limiter, battery charging).

- - - Updated - - -

That's an useful app note for UC3842: https://www.ti.com/lit/an/slua143/slua143.pdf

 

[--BawA--]

Red_alert , could you please tell me the mechanism ,how current is adjusted in uc3842? i mean when i connect the 12v/7AH battery , the charging current should be 0.7A , and when i connect the 12V/ 150AH battery , the charging current should be 15Ampere,
i just wanted to know how this can be achieved , because constant current technique is used in both the cases ,so the output voltage should adjust automatically to keep the current constant ,.

now please correct me if i am wrong anywhere
i will design the system of maximum power 250 watts.
now i'll sense the voltage across the curent sense resistor and will feed it into non-inverting amplifier .
now i'll put a pot at the output of the amplifier and feed the output of the pot to the pin 3.

kindly explain internal mechanism of how the current is limited to desired level?

 

[red_alert]

When voltage drop across the current sense resistor raises over 1V, the PWM output pulse is inhibited (activating the RESET signal of the internal PWM latch - see image bellow).

Then, after next oscillator pulse, the output pulse is starting over (the internal oscillator set the PWM latch) an so on.

So the current is kept at a constant value by modifying the output pulse width - thus the classical PWM technique.

 

[--BawA--]

ok red_alert, i got you ,
thanks for your support , i'll design the hardware and will update you in couple of days.

 

[red_alert]

Good luck with your design! Btw, in case you still have doubts about comparator threshold hysteresis, I've made a more explicit diagram:



(It's all about simple voltage dividers)

 

[--BawA--]

red_alert, in post#25 ,
in that schematics , i don't have 7.5v zener , however i have used 6.2v zener diode , and the low battery indication led is continuosly glowing, i have checked the voltage across the zener and it is 6.28v, i think the value of resistor should be modified ,
could you please tell me the modified resistor values and also how to calculate them?


Thanks

 

[red_alert]

Oh, in that schematics I've used random resistor values (was just a topology sample, sorry).

Have you used D4/D5 diodes, too (to select a different resistor for the high/low threshold)? Can you post the values of the resistors you already choose?

 

[--BawA--]

i am using the same resistor value as you specified , i thought , you have calculated them.
by the way , what is procedure of calculation if i use different value of zener?

also please calculate the values of resistor for the zener diode of 7.5V, as i got a 7.5v zener now.
reference is 5.1V from the pin16 of sg3525.

and yes i used both diodes d4 and d5

 

[red_alert]

Okay.. just tell me what are the battery disconnect/reconnect thresholds. (i.e. 10.8V - 11.5V)

 

[--BawA--]

its 10.2V and 10.8V

 

[red_alert]

R1 = 10K, R22 = 47K, R2 = 1M.

Thus you got 5.1V and 5.4V thresholds at opamp IN+ input. R3 and R4 are identical so the battery voltage is halfed at IN- input.

10.2V / 2 = 5.1V
10.8V / 2 = 5.4V

By the way, you have to take into account that battery reading depends on battery status. If is charging, the battery will have a much higher voltage than if it's discharging.

So a battery with a steady (open) voltage of 10.5V could read 10V on a quite light load and 11V on a rather small charging current.

That's it, your thresholds might be inadecvate. (The battery will disconnect at 10.2V but after disconnection (no load) the voltage will rise to 10.8V in few minutes and it will reconnect again - though the battery wasn't charged in the mean time).

 

[--BawA--]

(The battery will disconnect at 10.2V but after disconnection (no load) the voltage will rise to 10.8V in few minutes and it will reconnect again - though the battery wasn't charged in the mean time).

so what do you suggest , should i increase the upper threshold?

also one more thing to ask : i want to design a circuit for buzzer using opamp , in case of overload or low battery , the buzzer should beep with a frequency of 1 second.
when the input to the circuit is logic high , the buzzer should start beeping and when it's logic low , the buzzer should stop ,i have seen some oscillator circuits using opamps , but i don't know how to implement a trigger system in it ,

 

[red_alert]

Just put a diode from your detection circuit to one input of the buzzer opamp, to inhibit its oscillations.

- - - Updated - - -

Regarding the battery thresholds.. I actually live off-grid (just solar & wind energy, no other grid connection) so I learnt a lot about lead acid batteries (mostly by direct experience).

Depending on how strong your battery string is (mine is over 20 kW), the battery voltage readings react slower or quicker to battery charging status.

If you only have a small battery, you have to carefully choose those thresholds. By example, upon a quick load change (a powerful appliance is turning on) the battery could report a false undervoltage. Thus you better use a delay circuit before disconnecting the battery (wait a minute or something).

I'm continuously monitoring the battery/load/charger current and voltage and I use more complex alghoritms before a disconnecting/reconnecting decision.

 

[--BawA--]

thanks red_alert , i'll design delay circuit , n will update you by tomorrrow..

 

[--BawA--]

red_alert , i made the delay circuit and it's working fine, now before disconnection , the sircuit waits for 15 seconds.

but i am facing problem in buzzer circuit , can you draw any schematics of buzzer circuit using opamp .
when the input to the circuit is logic high , the buzzer should start beeping with a frequency of 1hz. and when the input to the circuit is logic low , the buzzer should stop .

thanks

 

[red_alert]

That's a typical circuit for an opamp oscillator:



It's actually a hysteresis comparator, with a variable IN- signal (the charging/discharging capacitor). To modify the oscillation frequency you must change the R1 and/or C1 (to increase/decrease the C1 charging current).

By putting two of this in cascade (one with an oscillation frequency of 1Hz and the other with a frequency of 1-3kHz), you got a 1-sec buzzer. Just put a diode between first opamp output and a input of the second one.

Similary, put a diode between your detection circuit and a input of the first opamp to inhibit its oscillations. Just choose the IN+ or IN- as to get a steady output state of the opamp that will further inhibit the second opamp.

PS: I could have been posting a complete schematic of this but it's better for you to understand the basics of this circuit to (re)design it as you wish. Good luck!

- - - Updated - - -

I've designed a complete schematic (just in case):



You have to choose R1/C1 and R2/C2 to get the output frequencies you want.