Solid state switched power channels.

[mclemens1969]

I am wanting to turn on and off some simple resistive loads with a microcontroller, and monitor the currents per load while enabled.

I am considering a pair of these for each switchable "power channel":

Switching:
http://www.mouser.com/ds/2/205/CPC1907B-268909.pdf

Monitoring:
http://www.allegromicro.com/en/Products/Part_Numbers/0711/0711.pdf

The switch is optoisolated, so that's a nice bit of protection as the resistive loads are connected or disconnected externally via connectors. The switch spec says 0.06 ohms on-resistance. At a max draw of 6 amps that would be 6*6*0.06 = 2.16 watts... on an 8 pin DIP.

I was just wondering before I order a few of these if I am "missing the boat" somewhere, if there is something important I should consider.

The maximum continuous load I would need to support is 5 amps per channel.

Thanks for any insights,
Mike in Alaska

[qwaarjet]

Hall effect sensors are quite noise especially at lower currents, if  all you looking for is course current indication it is fine but if you want high precision go with shunt resistor  or current transformer.

Check the derating curve if your dumping 1.5W per channel your enclosure may exceed the 40C operating range at that current, active cooling may be required.

[BrianHG]

Everything looks good, just 1 thing to remember, even though your are are using opto-isolators, the output side where you will be connecting and disconnecting the load, when the relay is off/open, if the voltage between to 2 pins exceeds the maximum 60v, you will still potentially fry the opto-isolator's output.  Since I don't know the rest of the circumstances of your circuit, you may want to add 2 back-2-back 50v zener diodes to prevent the voltage gap from exceeding the 60 volt maximum.

As for the 5 amps per channel, as stated above, you may need some cooling if you have a small case with something like 10 5 amp channels mounted side by side.

This may add up to a lot of power as well, make sure your PCB had an adequate thickness of copper, not just trace thickness, but the actual copper clad thickness on the PCB.

[mclemens1969]

I mocked this up and tried various currents from 0-2 amps while monitoring the Allegro output voltage. The sensor output range was 1.65 at 0 amps to 2.15V at my allowed full load of 5 amps.

So the sensor is 1.65V - 2.15V.   (0.5V range)
My Xmega ADC input range is 0V - 2.06V    (2.06V range)

Can I use a single op amp circuit to amplify the sensor output and shift it to that ADC range? Any hints?

Thanks
Mike

[BrianHG]

My Xmega ADC input range is 0V - 2.06V    (2.06V range)

Can you not configure the Xmega ADC input to run 0.0-3.3v, GND-VCC.  In one of my circuits, I'm using the Allegro's output, with a simple capacitor filter, feeding my PIC's ADC's input directly.  The Allegro already has an internal op-amp buffered output.

Remember, the allegro's output should go from 3.025v to 0.275v, & 1.65v with no load, powering the IC with a 3.3v supply. (This is + max current to - max current & no current.)

If you only need to measure current in one direction, you can run the device in reverse meaning it's output will go from 1.65v with no load down to 0.275v (25 amps) at full negative current load.  This you can feed you current Xmega ADC directly.  This means you swap the IP+ and IP- inputs.

Your final choice would be a simple 2/3rd resistor divider at the output of the Allegro to drive your ADC input.

For your application, the Allegro ACS722LLCTR-10AB-T is a far superior choice.  You'll get much better definition with a voltage of:
0.330v for 0 amps, 1.65v for 5 amps.


 

[mclemens1969]

Awesome Help !

[BrianHG]

For your application, the Allegro ACS722LLCTR-10AB-T is a far superior choice.  You'll get much better definition with a voltage of:
0.330v for 0 amps, 1.65v for 5 amps.
This one is a 0-10amp device, one direction only, so, the output is at 0.33v with no load.

[mclemens1969]

Huge Thanks, that part looks perfect, it is a much better match to my ADC without using any adapting circuits.

[BrianHG]

Now just to make sure, we are talking about driving a DC power source, not AC...

[mclemens1969]

Yep just DC resistive loads.

[mclemens1969]

The Allegro and IXYS solution has a benefit of being able to hook up the power supply completely backwards and not hurt anything. Neither device really cares about the polarity of what is flowing through. Seeing that the loads are purely resistive, it would even work !! (Though, the Allegro would read 0 amps flowing, but nothing would break.)

BUT......  I need to sense the raw input voltage as well. (So I can do wattage calculations.)   If I tap off Vinput with a voltage divider, if the power supply were hooked up backwards, this could bring a negative voltage right in to my microcontroller. What is a safe way to get this Vinput sense line into the uC?

Thanks!
Mike

[BrianHG]

Without knowing your hookup, I can only say that the PIC's IOs have diode protection from VCC-GND, others may not.  Your choice is to use a small signal dual schottky diode at the input pin of your MCU, one diode to VCC, one to GNG while feeding your supply voltage through a 10k resistor (This may be part of your voltage divider network).  This will protect the input from going below -0.3v and 0.3v above your VCC.  If you look at your MCU data sheet, this slight over-range should be safe.

Now, like the new Allegro 0-10 amp current sensor, you are safe to operate in the negative voltage supply, as in nothing will blow, but, your readout will be 0 with a negative source power.

Otherwise, you would need to run your power through a bridge rectifier which will introduce that 1.5v voltage drop.

[mclemens1969]

Thanks, I will switch to that.

I have not chosen the ideal divider resistors yet, but they should typically,  easily, be of a size to protect those diodes right? This is for sensing DC and speed is not important.

[tkuhmone]

Interesting thread to follow, this might help me also on some future projects (after got enought basic knowledge on Arduino programming).

Does anyone have idea, how sensitive to external RF fields those Allegro sensors are ? I have thought to use them current sensing in environment, where is strong nearby RF-fields (ham radio transmitters) in the frequency area up to 432MHz (and power levels up to 50...100W feed at antennas).

[BrianHG]

Does anyone have idea, how sensitive to external RF fields those Allegro sensors are ? I have thought to use them current sensing in environment, where is strong nearby RF-fields (ham radio transmitters) in the frequency area up to 432MHz (and power levels up to 50...100W feed at antennas).

Off topic, just start your own thread with the question.

[BrianHG]

For diode protection, use this guy, it has 2 schottky diodes in one package, wires in series exactly the way you need it:
http://www.digikey.com/product-detail/en/on-semiconductor/BAT54SLT1G/BAT54SLT1GOSCT-ND/917811
Though it can handle 300ma, set up your series power-supply resistor voltage divider so it is not possible to exceed 50ma with the maximum possible voltage.
This means, with 50v, don't use a primary resistor of less than 1k.  Though, wasting 50ma just to measure your main 50v supply voltage is extreme, I would use 10k, then something like a 1k to GND.

[mclemens1969]

I am thinking that part link is incorrect?

[BrianHG]

Ooops, sorry, correct link...:
http://www.digikey.com/product-detail/en/on-semiconductor/BAT54SLT1G/BAT54SLT1GOSCT-ND/917811