How to build a Floating Ground Buffer Box?
Posted by vincentcheung at
What you need is a differential amplifier with high common mode rejection ratio. A simple unity gain differential can be built by four identical resistors, one unity gain stable operational amplifier and a +/- 15V split power supply with proper filtering.
The construction is simple so I describe it here in words:
+15V of power supply connects to operational amplifier positive power pin;
0V (power supply ground) connects to DAQ module ground;
-15V connects to op-amp negative power pin;
1st resistor connects between motor controller output and op-amp "+" input;
2nd resistor connects between op-amp "+" input and DAQ moduile ground;
3rd resistor connects between motor controller ground and op-amp "-" input;
4th resistor connects between op-amp "-" input and op-amp output:
op-amp output connects to DAQ module input.
You may use op-amp like MC1458 and four 10k ohm 1% resistors to build the buffer for small signal frequency up to 10 kHz.
Replies
-
Thank you for the reply. I have bought the components and I am now going to build the circuit. I will let you know what the results are. Regards
-
Problem description:
I would like to measure the feedback (speed and torque, -5V to 5V) signals from a motor controller using a DAQ module. The problem is that the DAQ module (computer) has a floating ground relative to the motor controller (earth ground). Because the floating DAQ ground and motor controller grounds are not on the same level, there is an offset in the signal. Since this offset is too big for the measuring range of the DAQ, the DAQ gets flooded, which results in the DAQ stalling and reading an unusable signal.
I would like to build a ground buffer box which compensates for the floating ground. I remember using one of these at varsity in a project in which I wanted to achieve this exact same thing. It worked very well, however I cannot remember how the buffer box worked or how it was built. I suspect it has something to do with an operational amplifier.
Is there someone that can perhaps point me in the right direction (maybe an internet resource) as to where I can find the information to build one of these ground buffer boxes?
Thank you very much
Anton Esterhuyse
***email address removed*** -
Hi, To keep things well matched, precision, and noise lower, you might want to take a look at an INA132 instrumentation amplifier. You can easily find this device at DigiKey or other vendors and download the manufacturers datasheet. The COOL thing in that the four (4) resistors are integral to the amplifier and they are laser trimmed for precise matching and for cancellation of diffrential thermal drift issues! External resistors will perform -say, 100 times worse if it doesn't matter, and if you like a larger total footprint circuit.
Good luck, kwck
-
Hi Anton!
I am astonished that you meet problems between the earth (or the electronic ground?) of your speed/torque sensors. Let's try to identify what's going wrong. By the same way, I'll tell you how I get rid of this kind of problems.
For proper operation, and thus avoid earth loops problems, the electronic circuitry of the sensors must be fully isolated from the housing, providing thus a pure floating output signal. The “zero” of the electronic circuit must in any way be connected to the case!
If this is correct, and if problems persist, investigate the power supplies of the sensors. Leakages may occur between the primary and secondary windings of the (poor quality or damaged) transformers; in most cases, you can't identify them with an ohmmeter, as they become only significant when line voltage is applied. A good idea, for sensitive circuitry, is to use isolated transformers (with a shield between the two windings, NOT a shield around the transfo, although this latter prevents the appearance of unwanted magnetic fields, and thus 50/60 Hz pick-up). This shield must be connected to and only to the case, with case connected to earth. If ALL the sensors are equipped in such a way, you avoid problems due to parasitic coupling through the 230V line.
Wiring between sensors and measuring equipment (DAC, e.g.) must be realized with a two wires (signal) + shield cable, (all) the shield(s) being connected solely at one unique “electronic ground point" at the DAC. Proceeding this way, completed with isolated transformers, creates a star-like electronic ground network, without any loop.
To terminate the job, insert instrumentation amplifiers, like the INA128 (better choice than the INA126). The first scheme in fig.3 (Burr-Brown or TI datasheet) should meet your requirements. This latter provides easy return paths for bias currents, but don't forget to adjust the gain at a value which avoid overloading the output!
And for perfect operation, the INA105 datasheet (fig.17) gives you the principle of a guard drive generator for your shielded cables, which can be tailored to your particular needs. Fig 13 offers a less sophisticated circuit to perform this fonction, if the 47K resistors in the aforementionned fig.3 are each replaced by 2 x 22K resistors in serie, the middle point being connected to the inputs of the summing circuit, and the output driving the shield.
I hope this few ideas will help you, as well as other readers.
Regards,
B.B.