Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

Amplifier for NI 9263 current

I bought a NI cRIO chassis including a NI 9263 module. The 9263 module was meant to control a pneumatic valve accepting a 0-10V control signal.

At the time of purchase, I was convinced that all 0-10V control modules and valves should work together and power requirements were some industrie standard... Unfortunately, looking at both datasheets it turns out not to be the case. The NI 9263 can output max. 1mA/ch, while the valves need at least 1.6mA to function.

Buying a higher-rated NI module is very expensive just to get the required current, so it was suggested I build an amplifier to place between DAQ and valve.

I'm no electrical engineer, so my knowledge is just basic. I was hoping someone more knowledgeable could tell me wether my idea will work, or will fail.

 

My idea was to use an operational amplifier with enough output power to control the valves. Since I want to amplify all 4 output channels of the 9263, I was looking at packages which include 4 op-amps in one chip. I would connect the opamps in a voltage follower configuration, so there is no amplification and output current comes from the opamp supply rather than the NI 9263.

My idea was to use a Tracopower TMT 15215 power supply to convert AC voltage to +/- 15V DC, as a power supply (15 Watt) for the op-amp. I would use a National Semiconductor LM837 as the op-amp chip.

 

As far as I am capable of checking this system, I found that it should be able to source/sink almost 40mA per op-amp. Using +/- 15V supply leaves sufficient headroom between my 0-10V and rail to work over the entire range I want. Controlling valves is something that's not super-fast anyway, so I think it's dynamic characteristics (rise-time) are also more than sufficient. (I even believe it to be over-dimensioned, but with little knowledge, I prefer to be on the safe side).

 

Can anyone confirm this will work? Or let me know what problems I might expect, or what I could possibly do better?

0 Kudos
Message 1 of 8
(5,798 Views)

I think that should work.  The LM387 seems like a good choice for that driver.  I recommend that you put a resistor from each amplifier input to ground (about 100 000 ohms) so that the output will be off if the amplifier power is turned on and the cRIO is off.  I also recommend that you connect a diode across each valve control input with the diode cathode connected to the positive power supply.  This protects the amplifier from the energy stored in the valve coil when the control signal turns off. 

 

Lynn

0 Kudos
Message 2 of 8
(5,793 Views)

Thanks for your reply!

 

I've been trying to understand your suggestions, but keep failing...

 

1. Why would I place large resistors between the cRIO output and ground? I assume your proposing a solution to the offset voltage, but can't understand how this is supposed to work.
2. I can follow the logic in protecting the opamp from coil currents when I turn off the control signal (or lower it sufficiently fast). But connecting a diode between supply (+15V) and coil in the manner you propose will only allow current to flow if the coil goes over +15V? And even then, it still has a free path to flow to the op-amp. Wouldn't it be better to place a diode between op-amp and coil (cathode side towards coil), and place a large resistor to ground behind that? That way, current can flow from op-amp to coil, but not backwards. And coil currents after shut down will flow to ground through the large resistor.

 

I could be talking complete non-sense. My knowledge of non-ideal op-amp behavior is almost non-existent. 

0 Kudos
Message 3 of 8
(5,788 Views)

1. The resistors are part of the amplifier circuit.  So if the amplifier is connected to the valves and has power applied, but the cRIO is disconnected or powered off, then the valve will not turn on due to stray charges picked up by the input.  This is a parallel ground path at the input of the amplifier.  Its impedance is relatively lower than the input impedance of the amplifier but high enough that it is a negligible load for the cRIO.

 

2. The diode is connected in parallel with the coil, not in series.  The polarity is such that no current flows through the coil when the amplifier is powering the coil.  It allows a path for the energy stored in the magnetic field of the coil to be disipated without creating high voltages which might damage the amplifier.

 

Lynn

0 Kudos
Message 4 of 8
(5,785 Views)

To be 100% sure, the circuit below is what you mean with the diodes?

 

emf.jpg

0 Kudos
Message 5 of 8
(5,781 Views)

That should work.  What I had in mind was only one diode. The upper diode would be omitted and the anode end of the lower diode woudl connect to ground rather than V-, so that it is in parallel with the coil.

 

Lynn

Message 6 of 8
(5,777 Views)

Does anyone have a recommendation for a product that I can just buy to do this?

 

I'd rather not spend the time to build an amp, and buy one so i can focus on other tasks.

 

Lookig for up to 5 amps, 32V and capable of DC-10kHz. It will operate in the range of 1A 18V most of the time, around 2kHz.

 

Thanks so much in advance!

0 Kudos
Message 7 of 8
(5,415 Views)

 


@dierks wrote:

Does anyone have a recommendation for a product that I can just buy to do this?

 

I'd rather not spend the time to build an amp, and buy one so i can focus on other tasks.

 

Lookig for up to 5 amps, 32V and capable of DC-10kHz. It will operate in the range of 1A 18V most of the time, around 2kHz.

 

Thanks so much in advance!


One suggestion here

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


0 Kudos
Message 8 of 8
(5,405 Views)