I've tried the LM317 as a current limiter, see diagram attached. I tried with a variac setting with approx. 165 VDC and a load getting exactly 100 mA without current limiting.
I used a 20 Ohm resistor and expected the current to be limited to 62.5 mA. However in real life testing the current is not limited at all, still giving 100 mA.
What could possibly go wrong with just a single 3 pin device and one resitor? I checked that the output was taken from the Adj. pin.
I just don't get why it's not working.
Regards, Gerrit
I used a 20 Ohm resistor and expected the current to be limited to 62.5 mA. However in real life testing the current is not limited at all, still giving 100 mA.
What could possibly go wrong with just a single 3 pin device and one resitor? I checked that the output was taken from the Adj. pin.
I just don't get why it's not working.
Regards, Gerrit
Jon, this current regulator function is shown in the datasheet as well. See page 13 section 9.3.3.
Lineup, the input voltage is 165 Volt (to get the 100 mA load for testing). The LM317 is not in any way connected to ground (only through the load) so it doesn't see the "high" voltage. Of course with the variac I started with low voltages (20 - 30 Volt), but the voltage doesn't seem to be the issue.
Regards, Gerrit
Lineup, the input voltage is 165 Volt (to get the 100 mA load for testing). The LM317 is not in any way connected to ground (only through the load) so it doesn't see the "high" voltage. Of course with the variac I started with low voltages (20 - 30 Volt), but the voltage doesn't seem to be the issue.
Regards, Gerrit
Of course I can use a new one tomorrow, but I’m not convinced this is the issue. But time will tell.
Regards, Gerrit
Regards, Gerrit
Good morning , can I ask you design all the scheme ...power supply connection included. As lineup told you... I think high voltage is the problem...regards Ros
Hi Mooly,
The resistor is 1650 Ohm (but of course I can test with other values). If the input voltage is 165 Volt, the output is slightly lower (2 Volt due to R1), so the voltage across the LM317T is still very low.
Regards, Gerrit
The resistor is 1650 Ohm (but of course I can test with other values). If the input voltage is 165 Volt, the output is slightly lower (2 Volt due to R1), so the voltage across the LM317T is still very low.
Regards, Gerrit
Interesting. I'm surprised a simulation for overvolts works as well as it seems to do.
A 1650 load works over a limited input voltage range as you can see here. You need at least 100 volt input to allow the load current to meet its target and then when the IC has 40 volts or more across it then it all breaks down (literally I suspect for the chip).
So we have Vin, Vout and Load Current displayed here for a 0 to 165 volt input. The circuit works from about 100 volt input to 140 volt input beyond which is damages the chip.
A 1650 load works over a limited input voltage range as you can see here. You need at least 100 volt input to allow the load current to meet its target and then when the IC has 40 volts or more across it then it all breaks down (literally I suspect for the chip).
So we have Vin, Vout and Load Current displayed here for a 0 to 165 volt input. The circuit works from about 100 volt input to 140 volt input beyond which is damages the chip.
The Zener diode in parallel with the integrated circuit is missing.
If you use the exact schematic posted, then the Zener diode must be placed between the input and adjust.
This will not necessarily solve the problem but will protect the integrated circuit.
Is the power supply in rectified and filtered DC?
If you use the exact schematic posted, then the Zener diode must be placed between the input and adjust.
This will not necessarily solve the problem but will protect the integrated circuit.
Is the power supply in rectified and filtered DC?
Mooly, I will try again tomorrow (also with another LM317 as I have several new stock from a main supplier).
Sesebe, I don’t see any zener on the datasheet. How would you put the zener in the circuit? Where do you put the zener’s anode / kathode?
The input voltage is filtered and regulated with a gyrator (very clean output, less than 0.5 mV ripples & noise with 100 mA load).
Regards, Gerrit
Sesebe, I don’t see any zener on the datasheet. How would you put the zener in the circuit? Where do you put the zener’s anode / kathode?
The input voltage is filtered and regulated with a gyrator (very clean output, less than 0.5 mV ripples & noise with 100 mA load).
Regards, Gerrit
A crude voltage regulator in front of the 317 prevents chip overvoltage. Power dissipation rises with source voltage.
Referring to post #1.
62.5mA into 1650ohm is 103V.
That leaves 62V across the LM317 + current setting resistor.
The LM317 is unlikely to survive.
When it burns through, current is limited only by the resistor.
165V / 1650ohm = 100mA.
The solution is either to use LM317HV (just survive 60V).
Or more robust solution is to cascode with a depletion Mosfet.
https://vwws.wordpress.com/2011/06/21/dn2540-lm317-cascode-ccs/
In that case preferably with LM1117 instead because of the lower drop-out voltage.
Patrick
62.5mA into 1650ohm is 103V.
That leaves 62V across the LM317 + current setting resistor.
The LM317 is unlikely to survive.
When it burns through, current is limited only by the resistor.
165V / 1650ohm = 100mA.
The solution is either to use LM317HV (just survive 60V).
Or more robust solution is to cascode with a depletion Mosfet.
https://vwws.wordpress.com/2011/06/21/dn2540-lm317-cascode-ccs/
In that case preferably with LM1117 instead because of the lower drop-out voltage.
Patrick
Last edited:
You need to make sure that the LM317 never 'sees' more than 40 V from input to output. This includes during start-up, power-down, and current limiting operation. You also need to ensure that it doesn't overheat during any of these scenarios.
Adding a zener diode across the LM317 might save the LM317, but then you lose the current limiting function as soon as the zener conducts. And depending on your circuit, you may fry the zener, which will then fry the LM317.
You might be able to use a depletion MOS device as a current limiter. There are some available that can handle hundreds of volt.
Tom
Adding a zener diode across the LM317 might save the LM317, but then you lose the current limiting function as soon as the zener conducts. And depending on your circuit, you may fry the zener, which will then fry the LM317.
You might be able to use a depletion MOS device as a current limiter. There are some available that can handle hundreds of volt.
Tom
The cascode MOSFET is usually rated at 400V, and it ensures that the LM317/LM1117 does not see more than 3V at 62.5mA.
The 317/1117 still regulate the current even with no load.
The MOSFET will also take most of the dissipation.
All details described in Walt Jung's article :
https://audioxpress.com/assets/upload/files/Sources_101_P2.pdf
Patrick
The 317/1117 still regulate the current even with no load.
The MOSFET will also take most of the dissipation.
All details described in Walt Jung's article :
https://audioxpress.com/assets/upload/files/Sources_101_P2.pdf
Patrick
Last edited:
@tomchr The zener diode when is broken because of overpower will became a short and will still protect the IC.
The zener diode will not solve the problem of this schematic but will protect the IC. To solve the problem of schematic a solution with a preregulator will help but is not the best and anyway, no matter what solution is used, it must remain in the circuit.
A better schematic is this:
It will only limit the maximum voltage at the terminals of the LM317 integrated circuit but will not limit the voltage range in which the current limiting will work, as in the version with zener diodes to GND. The zener diodes must therefore be chosen to limit the voltage to a maximum of 35V to protect the integrated circuit.
The zener diode will not solve the problem of this schematic but will protect the IC. To solve the problem of schematic a solution with a preregulator will help but is not the best and anyway, no matter what solution is used, it must remain in the circuit.
A better schematic is this:
It will only limit the maximum voltage at the terminals of the LM317 integrated circuit but will not limit the voltage range in which the current limiting will work, as in the version with zener diodes to GND. The zener diodes must therefore be chosen to limit the voltage to a maximum of 35V to protect the integrated circuit.