PWM coupled with current limiting circuit design

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
Hello there
the following is a current limiting circuit template with low voltage drop

but, how can I couple it with a PWM adjustable circuit (both Frequency and duty cycle), so I can control the pulse and current in one go.
the device must be capable of working with 12V, up to 10 Amps, Frequency range of 5 kHz to 500 kHz, duty cycle from 50% up to 90%
thanks for your kind support and reply
 

dendad

Joined Feb 20, 2016
4,447
Assuming your circuit works, another FET across Q2 could be PWM controlled, or a FET in series with the load.
Do you need the current adjustable? A PWMed lamp driving an LDR in place of R5 is a way to control the current.
There are lots of ways it could be done.
 

crutschow

Joined Mar 14, 2008
34,201
how can I couple it with a PWM adjustable circuit (both Frequency and duty cycle), so I can control the pulse and current in one go.
You could replace Q2 and it's associated circuitry with another op amp driving the opto.
One op amp input is from IC1, and the other is from the average PWM signal, which serves as the reference voltage for the current limit.

If you used a P-MOSFET for Q1 (source to +12V), you could drive the gate directly from the 2nd op amp, and eliminate the need for the optocoupler.

What do you want to change the PWM frequency?
Only the duty-cycle, not the frequency, has an effect on the PWM average voltage.
the device must be capable of working with 12V, up to 10 Amps
Note that at 10A, the 0.1Ω shunt will be dropping 1V and thus dissipating 10W.
Suggest you use a smaller shunt resistor to reduce the voltage drop and power dissipated.
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
You could replace Q2 and it's ...
the idea of using another op amp is very interesting but actually I am looking for something like this, since I have to optimize the frequency and duty cycle for my load to see which is more suitable.
pulse-generator.gif
but I don't know how to couple it with current limiter, and also thanks for shunt resistor advice, yes you right. but a MOSFET driver is prefered I guess, so no thanks, I think I can work with optocoupler that way.
any idea how to couple it with the above circuit????
 

crutschow

Joined Mar 14, 2008
34,201
the idea of using another op amp is very interesting but actually I am looking for something like this,
The 555 PWM filtered output would go to the new op amp to generate the DC current level reference.
Do you understand that the PWM signal must be filtered to give a DC reference voltage?
Or were you planning of applying the PWM chopped signal directly to the load?
 

Thread Starter

farzad latifeh

Joined Oct 3, 2017
82
Do you understand that the PWM signal must be filtered to give a DC reference voltage?
I'll be really appreciate if you explain more or draw it on the paper so I can see what exactly your information is
BTW, I am not really good at electronics
Or were you planning of applying the PWM chopped signal directly to the load?
if I knew how to do that, I would never asked for it, clearly :)
in addition, the purpose of using PWM is to send pulse to the load and not use it as a reference for current limiter, I guess there is misunderstanding
 
Last edited:

crutschow

Joined Mar 14, 2008
34,201
I guess there is misunderstanding
Yes.
You originally stated:
how can I couple it with a PWM adjustable circuit (both Frequency and duty cycle), so I can control the pulse and current in one go.
You can't readily control both the pulse width and the current in "one go".
So you would need one signal to control the current, and another signal to control the pulse width (duty-cycle).
Is that okay for your purposes?
 

MrAl

Joined Jun 17, 2014
11,342
Hello there
the following is a current limiting circuit template with low voltage drop

but, how can I couple it with a PWM adjustable circuit (both Frequency and duty cycle), so I can control the pulse and current in one go.
the device must be capable of working with 12V, up to 10 Amps, Frequency range of 5 kHz to 500 kHz, duty cycle from 50% up to 90%
thanks for your kind support and reply
Break the connection between Q2 and the opto coupler input anode.
Insert a VCO between the Q2 drain and opto coupler input anode.
The VCO should be a voltage controlled duty cycle oscillator.
Depending on the VCO frequency C2 may have to be increased in value.
Dont expect super fast overcurrent response. For that you would use a transistor current limit circuit rather than an op amp current limit circuit.

Really though this kind of circuit is usually made from a buck converter. That way you get a smooth DC output as well as voltage regulation and current limit.
 

ronsimpson

Joined Oct 7, 2019
2,954
This is how I would do it...
When I make a differential amplifier, C2 has a mate on the (+) input. With out it I think noise will drive through the opamp and go directly to the output. I think the time constant on both inputs need to match.
1575242145834.png
What is the purpose, to limit the average current or the peak current?
 

MrAl

Joined Jun 17, 2014
11,342
This is how I would do it...
View attachment 193501
Hi,

I dont think that is such a good idea.
For example, if the circuit were to go into current limit that would be controlled by IC1, and pulsing the signal where you show would just mean IC1 would try to compensate less while still maintaining the output level. If it was in voltage regulation mode though, then your pulse width would control the voltage and since there is no feedback there would be no voltage regulation.
I would think that any addition should also handle voltage regulation. That's usually something very desirable.
 

ronsimpson

Joined Oct 7, 2019
2,954
The peak volts will be 12V, but does that matter?
That is what I am trying to find out. Because with out a inductor the circuit will force 12V on the cell and likely the current is 20A or much more. (peak) I do not see anything that limits the current. I do not understand the cell and can't find answers. Is 20A at 50% what you want? A real PWM will have a current of 10A. (maybe it ramps from 9 to 11 and back to 9)

I dont think that is such a good idea.
This picture is not exactly what I was looking for. I have built many audio diff amps (line amps) where you want a small signal that is riding on noise. Much like the LT1636 circuit. What I built needs the RC time on the (-) input to exactly match the RC time on the (+) input. Maybe I was assuming that we want the voltage across the 0.1 ohm resistor and do not want the 0 to12V signal on the resistor.
1575259536436.png
 

ronsimpson

Joined Oct 7, 2019
2,954
I built the LT1636 circuit and ….. well I don't understand how the LT1636 works. (LT Spice)
If I remove the 0.1uF cap it almost does what I want.

The LT1636-jig works like I think it should.
 

jeffl_2

Joined Sep 17, 2013
74
Look at the opto (VOM1271) turn-on and turn-off times, if that needs to switch at the PWM rate you'd be lucky to get that up to 13 kHz on its own, and the delay isn't symmetric either. That's kind of the "Achilles heel" of this circuit for what you're trying to do, but that isn't what it was designed to do either. Generally high isolation and fast switching tend to be mutually exclusive but there are much faster devices available.
 

acdc_lab

Joined Mar 18, 2019
1
Hello there
the following is a current limiting circuit template with low voltage drop

but, how can I couple it with a PWM adjustable circuit (both Frequency and duty cycle), so I can control the pulse and current in one go.
the device must be capable of working with 12V, up to 10 Amps, Frequency range of 5 kHz to 500 kHz, duty cycle from 50% up to 90%
thanks for your kind support and reply
Hi, I don't think the circuit will work. Q1 is missing 10-15V bias.
 

MrAl

Joined Jun 17, 2014
11,342
Hi, I don't think the circuit will work. Q1 is missing 10-15V bias.
Hello,

I think maybe Q1 will turn on with around 8v but you brought my attention to this relatively new device. It's a photovoltaic isolated mosfet driver.

The problem with this driver is that the output short circuit current is specified in the tens of microamps which is very very low for a mosfet driver. More typical would be at least 500ma, yet this driver is spec'ed at roughly 50ua which is 10000 times lower max drive current!

Now also looking at the turn on and turn off times, we also see they are spec'ed in the tens of microseconds, which is also far too slow for a mosfet driver.

With the capacitor shown (0.1uf) all this means the turn on time could be as long as 16ms which is very very very slow. The turn off time would also be as slow, which is not very good.

All this points to the device as not being suitable at all for a normal switching circuit like PWM or a buck circuit but rather perhaps a Solid State Relay which usually does not have to be as fast because it is meant for an application that turns on once in a while and turns off once in a while, never too short of a time in between.

So yes this circuit even as is, is very questionable and probably wont work except as an on/off switch.

Solution: redesign the mosfet driver circuit.
 
Top