320 V DC 15 amps current limiter circuit

[Esq]

I am looking for a tried and tested circuit for my DC Motor protection. I would like to limit the load current to 10 to15 Amps, variable, for my DC Motor test bench and my need is to have it trip if an overload/short perhaps lasts to long. My search on the internet was so far not succesful.
Cobus

 

[BobK]

Current sensing resistor, comparator and relay.

Bob

 

[Esq]

Current sensing resistor, comparator and relay.

Bob

Hi Bob,
Thank you for your response. I would like to use the shunt resistor but the 0.033 ohm shunt I have in the circuit at present will give me 0.33 Volts at 10 amps and I am using PWM 20Khz. So the pulse width is 25 micro seconds and I cannot afford to filter it because the IGBT`s need to be interrupted within the 25 micro seconds. The rated short circuit time is 10 micro seconds max depending on fault level impedance down stream. I see noise also on the scope when looking at this pulse, so I will need to filter then and make sure the delay is minimum. Any ideas?
Rgds
Cobus

 

[Bluejets]

The rated short circuit time is 10 micro seconds max

So how do you get the motor started..?

 

[Harald Kapp]

I cannot afford to filter it

I will need to filter

Now there's a contradiction...

You could sample the analog voltage across the shunt by an ADC at 200 kHz sample rate or more. Then run the sampled data through a digital filter, using a microcontroller or an FPGA). Then analyze the change (dv/dt) of the filtered signal. A steady current should give a (nearly) static signal. A short circuit should be indicated by a step in the filtered signal. Use the step signal as an indicator to turn off the IBGTs.

You could also sample the voltage across the sense resistor by a track and hold circuit such that the voltage is tracked while the IGBTs are on, but held while the IGBTs are off. Thus you will reduce the pwm modulation from the sensed signal to a large extend and can use a faster filter to eliminate the remaining noise.

make sure the delay is minimum

Any processing wil add a delay. You don't want to act on the change in current due to the pwm controlled drive signal. Therefore you cannot react in a time span shorter than the period of the pwm signal (at least as far as I know).

 

[Esq]

So how do you get the motor started..?

The motor is a Switched Reluctance Motor and start current is very low.

 

[WHONOES]

How are the IGBT's switched?

 

[Esq]

I use a Micro and a TLP250H gate driver.

 

[WHONOES]

So, can you inhibit the fet via you micro or gate driver. Does the gate driver have an inhibit pin?

 

[Esq]

Dear Whoknows, thank you for your response. I have just completed a circuit that will inhibit all the gatedrives when an overload occurs within 4 to 5 micro seconds and I have a 2nd design that could half that time. However I need to develop a current limiting control as well and I can do so with the PWM control to the IGBT`s but I am looking for an analog current limiting device that could perhaps limit the current at a setpoint level so it does no need to trip at all and that will have to be analog for speed. Rgds

 

[WHONOES]

It is easy enough to monitor the current in your system, it just depends on how you are going to integrate it. You would presumably want to limit the PWM pulse width to a minimum to achieve this or, do you want to bypass the digital control altogether?

 

[Esq]

I need to control the motor speed with PWM in the SRM, but that is "slow" relatively, but when the motor is in operation in a working environment, where I have no control over the fault level impedance, current limiting is preferred to tripping. So analog as well as digital control is required. The IGBT`s require very fast shutting down as you know and that is the only way I know to manage the unknown fault level (I2T) for the starter to endure.

 

[WHONOES]

So you need to turn of the drive to the IGBT's.

 

[Harald Kapp]

So you need to turn of the drive to the IGBT's.

That is imho not the issue. The op wants to limit the current in the first place, turning it off only if the overcurrent situation persists for too long.

I am looking for an analog current limiting device that could perhaps limit the current at a setpoint level so it does no need to trip at all and that will have to be analog for speed.

Use a current limited power source. However, a linear current limiting device will dissipate a lot of power. You don't state what your operating voltage is. In the worst case (short circuit) the current limiting circuit will dissipate the same power the motor normally does as the full voltage will drop across the current limiting circuit and full limit current will flow, too. You're asking for a big heatsink.
Once current limiting has set in, you can star a timer (e.g. using one in your microcontroller) to turn off power after a certain time.

Here's a discussion of simple and more elaborate analog current limiting circuits.

 

[73's de Edd]

Sir Esq . . . . . . .( . . .be it Esquivel or Esquire ? )

Take a good deep read on the aspects of the wealth of info found within some of these procedures.

https://www.google.com/search?q=dc+motor+overcurrent+protection+circuit&newwindow=1&tbm=isch&source=iu&ictx=1&fir=OoDGeSHDTIGQuM%253A%252CrKSC09g3Eqjl6M%252C_&vet=1&usg=AI4_-kTu70xpaWPA1SHkNeldnJD_drjxBg&sa=X&ved=2ahUKEwj-iqzA08TiAhUOi6wKHTSwC6YQ9QEwA3oECAAQCg#imgrc=OoDGeSHDTIGQuM:

73's de Edd . . . . .

Typically . . . . . success always seems to occur in private, and failure, . . . . . in FULL public view.

 

[Esq]

So you need to turn of the drive to the IGBT's.

My Voltage is max 320 V DC and amps at present somewhere between 10 and 15 Amps. My current trip circuit is very similar to the one circuit on "current limiting article" because I also use an op amp in conjunction with a SR Latching (Nor Gate dual IC). Then I cut or close the Gatedrivers common from the Micro when overcurrent happens in about 4 to 5 Micro seconds using a 30 A current sensing unit on the Comparator. I like the advanced circuit current limiter and will look deeper into that idea with 320 V in mind.
Thanks for the participation.

 

[Esq]

Sir Esq . . . . . . .( . . .be it Esquivel or Esquire ? )

Take a good deep read on the aspects of the wealth of info found within some of these procedures.

https://www.google.com/search?q=dc+motor+overcurrent+protection+circuit&newwindow=1&tbm=isch&source=iu&ictx=1&fir=OoDGeSHDTIGQuM%253A%252CrKSC09g3Eqjl6M%252C_&vet=1&usg=AI4_-kTu70xpaWPA1SHkNeldnJD_drjxBg&sa=X&ved=2ahUKEwj-iqzA08TiAhUOi6wKHTSwC6YQ9QEwA3oECAAQCg#imgrc=OoDGeSHDTIGQuM:

73's de Edd . . . . .

Typically . . . . . success always seems to occur in private, and failure, . . . . . in FULL public view.

Thanks for the google links, found a few ideas to follow up. Appreciate it lots!!!!

 

[Esq]

Thanks for the google links, found a few ideas to follow up. Appreciate it lots!!!!

Esq is for Esquare!!!