Variable 0 to 28v analog ouput from +/-10v on a NI-USB-6003

[SparkyOne]

Does anyone here have some thoughts on how I can convert the +/-10V Analog Outputs from a NI-USB-6003 module to a low-current (under 100mA) variable voltage input from 0v to 28v?

I have a legacy Test Set with some old 20k potentiometers wired-up as shown in the attached diagram.   Looking to modernize the whole setup a bit, but I'm struggling with these?

I'm open to dumping the 6003 module....just don't know what would be a good replacement?

http://sine.ni.com/nips/cds/view/p/lang/en/nid/212385

Thanks,

[johnsold]

You will need some kind of amplifier.  One simple approach might be to use the LM317 adjustable voltage regulator. Connect the regulator input to a power supply of 32 to 36 V. Connect the AO to the Adjust terminal of the regulator. Connect the power supply ground to the USB-6003 analog ground. Note that the LM317 requires a minimum output current of 10 mA, which is more than the 6003 AO lines can provide so the typical resistor from LM317 output to adjust cannot be used.

Lynn

[SparkyOne]

I was actually thinking about using a LT1012 Amplifier.....but I'm not sure I'm reading the datasheet correct.    Looks like it is rated to handle a 40V differential on the rails (absolute maximum), but am I violating a standard electrical practice by tying pin 4 to Ground, and pin 7 to +28V?

http://cds.linear.com/docs/en/datasheet/1012afbs.pdf

They have an example -kind-of- what I was thinking about doing, but only drive the chip to +15V.  See attached image (also on page 14 of the URL link).

[johnsold]

Well, Maybe.

In looking over the complete thread again, I may not be understanding your application. In your orginal post you referred to "...a low-current (under 100mA) variable voltage input..." The 100 mA part is what I do not understand. Generally voltage inputs are specified by their input impedance rather than by a current unless the internal circuitry is current activated. If you need a method to produce 0..+28 V at up to 100 mA, that is very different from a requirement to replace the 20000 ohm pots.

I will not go into a lot of detail until I know more about what you are actually trying to do.

However, the circuit using the LT1012 will not do either.  You can certainly run it from +28 V and 0 V but the output will go from approximately 2 V to 26 V. Look at the Output Voltage Swing specifications on page 3 and the graph on page 10.  The output current is limited to about 13 mA at room temperature and about 8 mA at maximum temperature.  Also on a graph on page 10.

Lynn

[SparkyOne]

I'm still doing research on the legacy test set, and I have incomplete information at the moment....so yea, the 100mA is probably bogus (I'll have to recheck that reading in the morning, as I'm starting to question it myself).

What i do know is what the 20k Pot is feeding.   

It's going into a LT6220 Op Amp which feeds into a LTC1863 Analog to Digital converter.   They have a voltage divider on the front-end of the LT6220 which is knocking-down the input from from the 20k pot from 0v to approx 4v.    I'm not sure I completely understand how the Op Amp is configured...looks like a Non-inverting...but is missing some components (or at least I think it is).

Attached a quick sketch, see what you think....?

[Buechsenschuetz]

There are several approaches to control a 0V...28V voltage with a -10V...+10V analog output. 

The main problem is that you will need something that runs on a supply voltage above 28VDC and also will accept input voltages down to -10V. 

Anyhow, there is always some theory behind such problems. One approach to consider operational amplfifiers as what they were originally designed for - to perform mathematical operations such as adding/subtracting or multiplying/dividing values reprented as voltages. 

To get  0V...28V from a "source" with -10V...+10V, you can use a first stage adding +10V to the control voltage, so you will get 0V....20V. Multiplying this voltage by 1.4 will yield 0V....28V. All this can be done using basic circuitry with operational amplifiers, only problem is the voltage range since you will need operational amplifiers with a supply voltage range of approx. -15VDC to +30VDC (to allow sufficient headroom for inputs and outputs). 100mA output current is beyond the range of most op amps. There are power op amps available, but - as discussed before - check whether you really need the 100mA. 

[SparkyOne · Accepted solution]

Thanks for you help guys, apparently I just needed to discuss through the problem......

I made a mistake on my initial evaluation, and did not read my meter properly.   The legacy circuit is reading 100uA, I need to pay closer attention. 

So that said, I found a nice guide online for Scaling and Biasing analog circuitry:  http://www.symres.com/files/scalebias.pdf

I have a prototype circuit built with the components specified above, and it appears to be working properly.   I had to do two things:

1)   I had to re-scale the +/-10V input to 0-10V

2)  I had to bump-up my Op Amp rail voltage to 31V

I still need to tweak the resistor values a bit, but I'm getting almost full-range deflection.   I'm getting 0.46V with a -10V input, and 27.85V with a +10V input, and there are no "dead spots", as with the old Potentiometers.    Here's a quick circuit simulation that I whipped-out: