0 - 36 mA current Transmitter with 0 - 10 V input signal

Hello Kamal,

0 - 36mA is a non-standard range and it will be difficult to find a dedicated part. You will find many options for 0 - 20mA, 4 - 20mA which are standard current loops.

Let us wait for a response from TI.. perhaps they may have a part which will work for you by tweaking reference, gain components etc.

Regards,

Jayant Deshpande

Hey Jayant, 

As you mentioned I was able to find many 4 - 20 mA and 0 - 20 mA in the current loops.

I was waiting for TI to comeback with the answer.

However, thanks a lot for your input !

Kamal Joshi

Kamal;

I would suggest using a Howland voltage-controlled current source; with the proper components, you can build it to deliver any current that you like.

Hi Kamal,

Neil's suggestion of a Howland current pump topology is a good one. Some information I don't see in your post is what the load impedance is, and what bandwidth you are trying to achieve. Both of these pieces of information are necessary to select a proper device for the design.

Hello John,

I am trying to design this circuit to mainly provide a current signal between 0 - 36 mA to a hydraulic valve that has the coil resistance of 190 - 200 ohms.  Also, in terms of the howland current circuit, I was unable to find some good design documents on the web.

Please share if possible.

Thanks,

Kamal Joshi

Kamal,

SBOA046 "IMPLEMENTATION AND APPLICATIONS OF CURRENT SOURCES AND CURRENT RECEIVERS" has some information on Howland current pumps employing difference amplifiers. Also, if the current source is not required to be bi-directional, an opamp paired with a series pass device may meet your needs as well. See figure 17 in the above mentioned application note. In your case, rather than employing the 100uA current source on the input you would simply apply your 0-10V control voltage as shown below:

The value of R3 determines the current pulled through your load. At 10V input, the load current will be: 10V/274= 36.5mA. The transfer characteristic of the circuit is shown here:

Remember that if you chose to run the opamp single supply, you will need to select an opamp capable of handling an input that includes the negative rail (rail-to-rail input) and meets your power supply requirements. The LM7321 might be a good fit. Unfortunately, TI doesn't sell the Ideal opamps that Tina lets us simulate with...yet...

Hopefully this helps!

I should add that the inductance of the solenoid coil in your valve can cause flyback voltages that may damage the MOSFET in the above circuit. It would be smart to include some protection diodes in your final design.

Hey John,

I do not need bi-directional current, but if I was to add protection would I just add a diode before the R3 resistor?

Also, what is the simulation software that you use for simulating these circuits ?

Thanks for your input.

Kamal Joshi

Kamal,

At a minimum I would place a diode in series with the FET, between the FET and the coil such that current can only flow one direction through the sink. You may also want to put a flyback diode in parallel with the coil on your valve.

Those circuits were simulated with Tina-Ti v9 which is available for free on our website. It is a nice simulation program that has quite a bit of functionality and no node limit. It can be downloaded here: http://www.ti.com/tool/tina-ti

I've attached my simulation file if you would like to duplicate the results. I changed the opamp to an LM7321 (rather than an ideal model) and the FET was switched to a simple 2N2222 BJT because your output current is not very high. The two protection diodes I mentioned are also included and you can remove them to see the effects on your system. The inductor named "valve" represents your coil, I guessed at a 100uH inductance which might be pessimistic.

Hey John,

I see that you have power supply of 20 V, but for my design the maximum voltage is 12 V DC.

Are there components that work at lover voltage level to provide the same result?

Thanks,

Kamal Joshi

Kamal,

In the design I posted, your input control voltage is the voltage that will be forced across the current sense resistor R3. At 10V in, you will have 7.2V across your coil, and 10V across R3, as well as a drop across the transistor, this is why a 20V supply was used. There are many ways to get around this, one possible way is to add gain to the feedback loop, so the voltage across the resistor is amplified, allowing you to use a smaller resistor. However, the addition of a second opamp adds complexity and the chance of instability. The Howland current pump previously mentioned in this thread will probably be a more simple solution:

Two areas of concern are going to be power dissipation in the opamp and protecting its output against high voltages developed by the load inductance. The circuit should be easy to to prototype on a breadboard to confirm if it meets your needs.

Hello Kamal,

you can use for example the XTR111. Here an exerpt from the data sheet:

The XTR111 is a voltage-controlled current source

capable of delivering currents from 0mA to 36mA.

I

OUT = 10 • (VVIN/RSET)

You can find more information here:

http://www.ti.com/product/xtr111

Hope this helps

Andreas

Thanks Andreas.

I am not sure if the response of XTR111 is linear for the whole range of 0 - 36 mA for input signal 0 - 10 V.

Please Advise.

Kamal Joshi

Hi Kamal,

to be honest, I am not the expert for this product. What I understand the resitor basically sets the tranfer function of the amplifier.

There is a typical plot in the data sheet page 8 figure 15.

Andreas

Hello Kamal,

Andreas is correct, the XTR111 is specified for a derated output current range of 0-36mA.  The effect on the linearity of the IC in this extended output range can be seen in Figure 15 of the datasheet which I've copied below.

As you can see in the figure, the nonlinearity begins to take a hit after the output current exceeds 25mA with the largest effect on the nonlinearity after the output current exceeds 35mA.  Therefore with a 2 point calibration at 0.1mA and 36mA you should find that the nonlinearity is improved from the figure below.   

That said, John and Neil's suggestion above regarding the use of a Howland Current Pump may still be your best option depending on your exact application's needs.

Regards,
Collin Wells
Precision Linear

Hello,

There was an error in the previous post.  The device I meant to refer to is the XTR111, not the XTR110.

Regards,
Collin