I would like to setup output modul with over-current ( short-circuit ) protection.
So here is my plan:
Output power transistors as are: ULN2803 / UDN2981 or TIP122 for high current applications.
So example for UDN2981, on common Vce would be assembled Hall Effect-Based Linear Current Sensor IC ACS712.
Sensor will sens current flowing into tansistor array and send measured value to Arduino via analog multiplexer (multiplexer is just to save some native analog pins).
UDN2981 has 500mA max. current limit for one channel, so I would use 500mA as maximum current limit for whole IC.
In case that arduino analog read value will be higher than 500mA ( reading will be in voltage of course ) software will immediately switch off corresponding digital outputs ( close transitor ) and stop currend going thru.
So I have just one question, will be this software counter-measure fast enough to protect the transistor?
Surely it mainly depends on ADC sampling frequency and program. Thats my worried, with multiplexer I cannot use software interrupts.
However in my point of view:
for over-current and short-circuit protection is generally used PTC, but in according to it´s datasheet "time to trip" is much longer in compare to Arduino ADC sampling frequency
my proposed solution would be essential because of feedback ( means information about output error state )
Any your feedback is welcome.
Thank you in advance for your opinion. Best regards Miro
It is not clear to me, what your problem is. How are you sensing this motor current ? Are you concerned that your current sensing device is going to give more than 5V to your arduino analog input and blow up your arduino ?
Hi, hall sensor is sensing current, reading is not a problem.
If value readed by arduino analog pin exceed preset value, means transistor measured by sensor is overloaded
(over preset current).
Question is if read time and response time will be fast enough to protect the transistor output in case of short circuit.
Respons means transistor pin goes low to close the transistor, and stop high current going thru.
There are lot of "short circuits" that you might possibly get. Although really you are more likely to get an overload from stalling a motor than from a "short circuit".
If you are concerned about the time, that is, how quickly you can detect an overload, there are different kinds of failures. Some overcurrents will destroy a component effectively instantly. Others take a while to heat up and are not destroyed until seconds or tens of seconds have gone by, enough time for you to react and shut them off.
yep, overcurrent...time depens how much is power transistor overloaded.
thats why I have written short-circuit protection
Of course it might be dependent on the power supply for the transistor.... because if transistor is rated at 500mA
it would be big difference if power supply is rated at 1000mA or 10A.
As I wrote in my first post, as short-circuit protection is generally used PTC, but i have seen few datasheet and there was indicated that ¨"time to trip" is almost 1 second. So in this case software protection would be much faster....
Transistor datasheets usually have a graph of maximum power against pulse-length,
which should give a value for how many us or ms the device can stand for any particular
overload condition.
Less commonly you may get a similar graph for max current pulses. High current pulses
can melt/vaporize the bonding wires even if the device total power isn't a problem, so
ideally you'd like to know the maximum current and maximum power that the circuit can put
through the transistor, then work out the maximum safe delay to act in protecting the
device.
In general high current and voltage circuitry gives very little margin for error in a
hard short condition (sometimes 10's of kW are available to dump into a device, and
microseconds matter). For instance shoot-through in a bridge is like this and you'd
normally use a comparator to catch a hard-short condition as fast as possible.
When the "short" is through a motor you've got a less extreme condition as motors are
inductors and the current rises more slowly - so knowing the winding inductance is
useful for calculating how fast the current can rise to dangerous levels.
Having a bridge/switching configuration that prevents shoot-through means you will have
an easier time with protection circuitry (but remember a wiring-short is as fast as
shoot-through).
yep,maximum current and pulse duty cycle is clearly defined in transistor´s datasheet.
So if I understand well, software protection based on feedback from analog current sensor will be good in case of overloading, or short circuit in motor for example ( means still will be a lot of impedance in the circuit ).
But for wiring short circuit will be not fast enough to protect the transistor from damage.
It is simply because the power supply will be most probably very strong and in case of direct wire short circuit current spike will be huge....so transistor will not survive longer then few micro seconds..
In attachment is short circuit protection from wiki, based on resistor sensing current and direct feedback from control transistor to the power transistor.
So this circuit in attachment would protect the transistor in case of direct short circuit without any impendace?
Means voltage directly to ground
Sorry for my english, and thank you for your feedback. Miro
If you current limit the circuit then you only need to worry about the power
parameters - how long at maximum power before it needs to switch off,
which is usually on a nice convenient graph of pulse ratings on the datasheet
