High-Amp Current Sensor anyone?

Philldapill

The concpet is simple enough. I wanted to build a high-amperage current sensor for a number of uses. Some of these uses include:

1. Monitoring the current from a large PWM power supply for DC loads.
2. Monitoring and adjusting current output from my solar array
3. Making an AC phase angle/power factor meter for numerous household loads
4. Building a smart battery charger to regulate and program charging
5. about 20 other things...

I want your input. What sort of things would you use one of these sensors for? I've written an assembly program for a single channel ADC that samples at about 100ksps. This interfaces with a hall-effect sensor from Allegro. The Hall-Effect sensor is capable of up to 150A, but the one I'm using now is programmed for only 50A(which should be good for my applications). The whole thing will be mounted on a single PCB with ADC, sensor, and secure lugs for the cables to solder to. The sensor is bi-directional, meaning you can allow AC through it and still measure it. The output is a 5V scale, with 2.5V being 0A, 5V being +50A, 0V being -50A. Again, the driver is already written. The interface will be 1 +Vcc, GND, Data, CS, and Clock pin for a total of 5 pins, but only 3 I/O pins.

If you don't know what a Hall-Effect sensor is, simply put, it measure the magnetic field produced by current. The higher the current, the higher the magnetic field. Many methods of measuring current use a shunt, which is just a low resistance resistor. The current through the shunt creates a voltage across the shunt, and this voltage is measured. The only problem - shunts get kind of hot. The Hall-Effect sensor I'm using has about 100 micro-ohms of resistance. With 50A through it, it should only generate about 1/4W of power. Thats the amount a small resistor can dissipate. The sensor isn't even warm.

I've got these things working pretty well now. There is a SLIGHT bit of magnetic hysterisis in the sensor, so when current stops flowing, the sensor may say that 0.2A is still flowing. That's 0.2 out of 50A. This can be corrected in software.

So, anyone want one? What would you use them for? The price would probably be about $15 assembled.

FrameShift.

Very interesting idea. I was looking into this as well recently. Could we see some pictures of the system assembled?
Also, is the hall-effect sensor inline with the power cable? or just around/next to it?

-Tom

Timothy D. Swieter

I think the sensor has to be inline with the cable based on what I have read about Allegro Microsystem sensors in the last. I started a thread in the sand box a couple weeks ago regarding current sensors. I think for my upcoming project I am going to use a current clamp transformer because I am measuring current on an AC line and I won't have access to make the cable in-line.

Phill I am interested to see what you have made. It sounds like a great project.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com

Philldapill

Tim, your thread is actually what inspired me to get back to making this thing. I have tinkered with the sensors before, but what I had was rather crude. One of the first things I plan on doing, is making a stand alone power meter. Like I described in the first post, it would measure phase angle, current, voltage, Watts, VA's, impedance(maybe), and all the AC stuff a single Propeller can handle.

I don't have pictures yet, but I'll get some when the first project is finished. I want this power meter to fit in a small 3"x 4" box with a 2x16 LCD screen, a couple of buttons to toggle between display items, and two sturdy bolts to connect the current to, along with a couple of probes to connect to the voltage lines.

I wish I could use a clamp-around current transformer type of sensor, but this is intended to be a versitle AC/DC gadget, so a Hall device seems best. BTW, Timothy, that was a good thread. Keep it going.

Timothy D. Swieter

Thanks Phil. I think I will pop over to the other thread and post an update.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Timothy D. Swieter, E.I.
www.brilldea.com - Prop Blade, LED Painter, RGB LEDs, 3.0" LCD Composite video display, eProto for SunSPOT
www.tdswieter.com

$WMc%

Philldapill

I think a non contact current sensors would be more useful to the members on this forum that need to monitor household current or Industrial current.

A CT is noncontact but for AC only

A simple analog HALL-SENSOR Will do this, AC or DC. Buy measure of the GAUSS on the wire. NONCONTACT!

But don't take My word for it. Check out the Fluke 771 Milliamp Process clamp on meter.

Yes this is 24volts DC at 4.000mA to 20.000mA, an industry standard. The Fluke 771 has as much accuracy as a milliamp meter placed in series with the loop!


___Just something to think about_____________$WMc%_________

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there············································ BoogerWoods, FL. USA

$WMc%

Pilldapill

Have You had any success with this project?


____________$WMc%_______

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there············································ BoogerWoods, FL. USA

Philldapill

WMc, I sure have! I have the current sensor all finished - PCB and driver. I'll get some pictures up Thursday when I'm done with school for the week and at my parents place(my REAL lab).

Here's the basis of how the sensor works. The current flows through a Hall Effect Sensor, and the sensor outputs a voltage from 0-5V with 2.5V being 0A, 5V being 50A, and 0V being -50A. My PASM driver runs an ADS7816 ADC chip - 12-bit, single channel differential input. I just checked the sample rate using my DMM, and it looks like it's sampling right at 250ksps. This is actually TOO fast since the ADC is rated for only 200ksps, so my readings may not be 100% accurate... I might use a faster ADC, or just add some extra instructions into my code to slow the sample rate to 200ksps. These extra instructions may include some code for doing a windowed average, further stabilizing the output readings.

So, with 200ksps, I should be able to get 3,333 samples per cycle for 60Hz AC. I think this is more than enough resolution for determining RMS values of current.

What do you think? Faster ADC?

$WMc%

Philldapill

Excellent !!!

I think >3000 samples per/cyc. is plenty of resolution for an accurate RMS value. I also think the ADC is fast enough.

Adding more code to slow the rate (250k to 200k) could really be beneficial, Like adding that windowed average. A MIN/MAX reading would also be useful.

2 more COGs instant 3 Phase "I" meter, Widely used in Industry/Commercial use.

I,d like to see Your Elc. layout and the Hall Effect sensor Your using.

_____Great Project____________$WMc%_______

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there············································ BoogerWoods, FL. USA

FrameShift.

Cant wait to see some pictures of the board.