Help with Current Sensing Circuit Needed

[originator]

If someone could help me get this in the right perspective:

I want to take a digital output ( logic 1 pulse at an adjustable threshold) from the current consumed by an AC servo motor. I have a 24v regulator that is driving the entire motor section, my theory is to insert a current sensing circuit between that reg 24v and the motor/controller.

Using the multimeter inserted in the path in 10amp mode, the meter reads .011 when idle. When the motor is in motion (no load) there is no change. If I grab the timing pulley and put a lot of friction on it, the meter raises to around .1o or higher.

What is needed is a way to experiment with the best level of current drain to trip an output to tell the motor circuit to stop/reverse or whatever. A pot would be ideal for tuning the level in the real world application to determine the best threshold.

Would it be correct to assume that there are at least two methods to do this:

A circuit inline to the load with an output, that output goes into an 4 or 8 bit a/d. The parallel outputs of the a/d can be crudely arranged so that if a certain set of outputs are present ( ie only 0011) then a pulse is sent to the logic board for further decisions on what to do.

or

A circuit inline with the load with an output that goes to a comparator. If the output exceeds a voltage, the circuit turns on( or shunts, either way)

I am sure there is an easier way of thinking of this though.

Thanks

Todd Chapman

 

[waross]

Insert a shunt resistor in series with the motor. As the current to the motor rises, the voltage across the shunt resistor will rise proportionally. The lower the resistance of the shunt the lower the losses in it. Find or build a circuit that will turn on repeatably at a low voltage level. Use a potentiometer across the shunt resistor to match the turn on point of your circuit with the appropriate current level through the shunt.

 

[nbucska]

Simplest solution is to use a comparator: Let's suppose we
allow at max. current( About 1 A ) .1 V this means a
.1 ohm 1/4 W sense resistor.

You can use any comparator which can work around the GND.
Use hysteresis and an input lowpass filter to slow it down.

Generate the power supply from the 24 V .
It may be a $5.- project -- without ( or with home-made ) PC board

Plesae read FAQ240-1032
WEB: <

http://geocities.com/nbucska/>

 

[itsmoked]

I am confused... An AC servo motor fed with a regulated 24V supply? A regulated 24V AC supply???

 

[originator]

well, the 24 reg supply feeds the power to the controller, to which the motor is connected. the motor is separate from the controller, but the motor clearly says ac servo motor, therefor the controller must be creating the ac? thats my guess anyway. in any case, the dc current can be measured.

in placing a resistor in series with the supply, and measuring with a volt meter, the voltage drops at the shunt resistor when more load is put on the motor.

to use a comparator in this case, would not a reference of a lower voltage than the 24 motor supply be used? ie 22 volts reference voltage to the comparator, and a tap from the shunt resistor side to the motor...

if the shunt resistor is sitting at normally 24v, and the reference is 22, then when the shunt voltage drops with a heavy load on the motor, the amp turns on.

 

[waross]

A shunt resistor is connected in series with the load. It shunts most of the current around your measuring device.
You will probably get better results by putting your shunt in the motor circuit rather than the supply to the motor controller.

in placing a resistor in series with the supply, and measuring with a volt meter, the voltage drops at the shunt resistor when more load is put on the motor.

You have either made a mistake in your connections or in your measuring points. With a properly sized and connected shunt, the voltage across the shunt will be proportional to the current through it. The current may not be exactly proportional to the load but it is closely related to it.
yours

 

[itsmoked]

originator; I would check the actual output as DC is WAY more common and much easier for your desired monitoring solution.

Next check whether the controller is a half bridge controller or a full bridge controller.

If it is half bridge you get another point for simple as then you can put a very low value in the ground lead of the motor and reference the comparator circuits to ground making things very much easier.

Otherwise use a shunt current monitoring chip and put the shunt pretty much wherever you want.

http://pdfserv.maxim-ic.com/en/ds/MAX4373-MAX4375.pdf

Here's a page of them:

http://para.maxim-ic.com/cache/en/results/5102.html

Here's a pretty inclusive APP note:

http://www.maxim-ic.com/appnotes.cfm/appnote_number/746

These could also be used in the quasi-odd place you are trying to read the current now.


Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[originator]

Thanks for the info guys. I was looking at Maxim chips earlier, I think they are a good option. I have been breadboarding comparators for a few hours to a feel for how to use them as a cheap easy option as well.

I have tried a .1 resitor as a shunt in series to the motor, with the meter in series as well. With no load the meter reads .145 DC amps. With a large load it reads almost 1 amp. I don't know if I miscommunicated earlier, but when I pull the meter back out of the circuit, and just meter from the shunt and GND normally, it reads around 25 volts, and when I put a load on the motor it drops a few volts. I am working on a way to take the change in current and send it to the comparator, and when the voltage hits a threshold there is an output. I am not clear on how to convert the current increase into a voltage increase as noted before by waross:

"As the current to the motor rises, the voltage across the shunt resistor will rise proportionally"

As stated, when I pull the meter back out of series with the motor/controller, I do not find any voltage reflecting an increase proportional to the current. I apologize for my limited experience with this type of measuement.

If there were a way to get that voltage increase mentioned by waross down in the range of under 5 volts, that would be easily compared to a reference on the 5 volt comparator (LM339) and get a simple output when it exceeds the adjustable ref.

I suppose crudelely I could take a photocoupler off the supply to the motor, put 5 volts on the transistor side and get an output under 5 volts that would drop on higher loads, and use that output as the ref in the - input, and use the + as a reference instead. Kinda hacked up but may work. Essentially measure the supply voltage drop instead of current indrease since I can't undrstand a way to convert current increase to voltage increase as of yet.

 

[originator]

AHA! ok i get it now. I was measuring from GND to one side of the shunt. Now, when I measure across the shunt for volts I get the small voltage increase as you said.

Thanks!


Maybe I can take that and make something happen.

 

[itsmoked]

Voltage across a resistor is given by Ohm's law:
Voltage = Current * Resistance.

More current = MORE VOLTAGE.

You should look at the app note I provided.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[originator]

Thanks Keith, I'll get the Maxim 4173 shipped tomorrow and try it out. I was attempting to use a differential op amp to utilize the sense voltages but had no luck, I'll wait on the Maxim stuff and save the headaches of reinventing the wheel.

 

[originator]

I talked to Maxim today sorted out the part I needed, but nobody they recommended stocked it. So I kept fooling with a dif op amp circuit and with a 3.3 ohm 1w sense resistor got it to work beautifully with a comparator. There is some slight oscilation at the turn on point on the op amp but I'll stretch the pulse and not worry about trying to get it oscilation free.

Thanks again all for the help.

Todd

 

[itsmoked]

Look up "hysteresis comparator".

This will stop the possibly damaging oscillation problem you have.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Madcow]

Alot of people make the high side current monitors.
If you cant get the maxim , I have used the TI INA198.

http://focus.ti.com/docs/prod/folders/print/ina198.html

only 80 cents in volume.

But Linear and Analog Devices make them too.

Good Luck