Icon

Accurate AC Current Sourcing (With The ACS041)

1. Analysis of Design Points in Further Depth.
2. Technical Overview And Description.
1.1. High Impedance Leads.
1.2. Power Supply Layout And Earthing Points.
1.3. Electrostatic Coupling.
1.4. Electromagnetic (Inductive) Coupling.

This article concerns an APCS product the "ACS041 ac Current Source" that has been discontinued. This article has been retained for information purposes only.

The ACS041 Current Source has been purposely designed and built for injecting a stable, low distortion mains frequency current signal into ac transducers, meters and recording equipment. Why have we developed such a product?

There is an increasing trend in most industries these days for higher quality and accuracy with the requirement to provide traceability for the measurement accuracy. Manufacturing companies and those organisations involved in supply and/or maintenance are required by ISO9000 standards in their Quality Assurance program to ensure that their products and services meet stated accuracy limits and that their testing instruments are calibrated and traceable to National physical standards. This provides confidence both to the customer and the supplier. This also ensures that the correct information is obtained which can be acted upon to optimise efficiency and ensures that the product is functioning correctly which can be important for safety. This is especially true for those involved in the AC power frequency fields of instrumentation, ac metering, power system protection and control, and those involved in plant or building services maintenance and relay testing.

(Back to top)

Relay testing includes such items as earth leakage relays, ac over-current relays and shearpin relays. These all need to be tested periodically to confirm their correct function. These relays are used to protect people and expensive machinery in various industries ranging from mines to mills to machine shops. AC transducers, panel meters, multimeters and recorders are designed to provide information to confirm correct and efficient operation of the plant. The performance of all these items should also be regularly checked to ensure they do so. A.P.C.S. had a need for a reliable, stable AC Source for testing our own AC transducers and trip modules. We perceived a need in the marketplace by other people for such a device. The requirements for an ac current source can no longer be met simply using a mains variac, step down transformers and suitable power resistors to develop a current. Using these devices creates the problems of:-

  1. instability and fluctuations which increases the uncertainty of measurement;
  2. costly calibration time is required to obtain a reasonable measurement or make a suitable calibration;
  3. measurement errors caused by distorted sine waves.

(Back to top)

These days we require quick, efficient and effective calibration techniques. E.g. in calibrating a 5 Amp current transducer, we need to be able to quickly source a 5A signal from a calibrator, which can be relied on to be accurate and stable; use this to calibrate the zero and span potentiometers in the transducer and then check multiple points in the range to confirm the linearity of the transducer. With high labour costs, this needs to be done in a quick efficient way. In testing an earth leakage relay, we need to connect to a current source and simply and smoothly increase the current level and note the level of tripping. Many companies have trouble justifying price tags of $20K to $100K for AC calibrators (which may have accuracy figures to 0.01%) when their requirements could be met by a stable source with reduced accuracy specifications. With these requirements in mind A.P.C.S. has developed their AC CURRENT SOURCE - model ACS041 to supply currents up to 10A with very good stability. This unit solves the above problems by being inexpensive and at the same time a tool which is simple to understand, operate and quick to use. Our initial target price was under $2K, but realistically costs increased as quality and stability improved to the level we required, and after two years development, we go to the market with a price around $3K. At this price, it should be affordable by most organisations or technicians requiring such a product. Why the emphasis on stability?- Many tests require us to generate a current and then rely on the source to maintain this value without drifting or jumping around, e.g.:

  • If we are calibrating a transducer, we do not want to waste time by continually adjusting the calibration knob if the source is drifting, or
  • If the source is beating about the required value we do not want to have to average the measurement and output values in order to get confidence in our calibration, or
  • If we are performing long term drift tests on a particular product, we require the source to be stable so our results are not distorted by the source long term drift.

(Back to top)

The design is not based on the usual concept of adjusting the amplitude by means of a row of dials or keypad entry (which relies on precision resistors or processor control). These methods rely on the stability of the design to maintain the output to be the same as the dialled indication. Errors can be caused by varying the frequency or load or by the degradation of attenuation resistors and the user is not aware of a problem until the source is recalibrated. The ACS041 has been designed so that the amplitude is adjusted by a precision multi-turn potentiometer until the required output is obtained as indicated by a precision meter actually measuring the output (refer block diagram). This gives the user confidence that the output is the required value. There are also switches available to quickly change the amplitude to user preset values to enable rapid testing at required levels.

(Back to top)

A problem with keypad or keyboard entry for amplitude adjustment is that, in order to incrementally change a value, the total value has to be re-keyed in. For example, to check the accuracy of an analogue meter, the input is incremented until the pointer aligns with a major division scale mark and the value of the input is noted and used to calculate the error. Trying to "creep up" on a division by continually retyping numerical values is very annoying and time consuming. This is one reason why the ACS041 was designed with a simple to use precision multi-turn potentiometer for amplitude adjustment. The front panel controls are clear and ensure that it is easy to use. The range switch selects 1A, 5A or the 10A position. The instrument under test can quickly checked for linearity with a switch which provides 100%, 50% and 10% of the range. These settings can be easily adjusted to other user required percentages. There is also a continuously variable setting for easily adjusting 0-100% of range. Frequencies of 50Hz, 60Hz or 400Hz may be selected or an external frequency from 40 to 500Hz may be injected. The output is displayed on a 14mm led digital panel meter which has a 10mA resolution. The unit is mains powered, however, it uses an independent internal oscillator to generate low distortion AC signals. The ACS041 operates into 10VA burden (1.0V drive). This is generally more than sufficient for the user's requirements since most instrument CT inputs impose about 1VA burden on the source. Thus, allowing for wiring losses, we can test 5 to 10 units simultaneously.

(Back to top)

The basic accuracy of the ACS041 is 0.2% which make it suitable to calibrate class 0.5 instruments. If higher accuracy is required , e.g. in a standards laboratory, the user only has to place his "known and trusted" ammeter (or precision shunt) in series with the output and dial the amplitude to the value required as measured by this reference. It should be noted that most digital multimeters do not have very good accuracy specifications for ac current. Even 6 and 1/2 digit bench multimeters of highly reputed brand names can have ac uncertainties in measuring 5A at 50Hz of 0.25%, or even 1.0% for 5 digit units. The design of a mains frequency current source has many problems. One of the most complex is that of eliminating or minimising what is referred to as "mains beating". This occurs when the frequency of the oscillator is the same as the frequency of the power supply. Part of the power frequency signal can be superimposed on the signal frequency. The coupling can be caused by such means as capacitive coupling or induction. Since the phasing of the two signals is random, the effect is a slow amplitude modulation of the signal. The problem can be addressed by such methods as:-

  • Careful layout of the PCBs especially in sensitive areas of the design such as high input impedance leads to operational amplifiers;
  • Selected layout of the components on the PCB around the power supply section and in all ground tracks utilising one point earthing;
  • Screening of the magnetics of the transformers;
  • Correct placement of wiring, especially conductors carrying heavy currents;
  • Attention to wiring, using twisted pair and shielded wires where required.

(Back to top)

1. Analysis of Design Points in Further Depth

1.1. High Impedance Leads

The high impedance input leads need to be kept as short as possible in order to minimise noise pickup, current leakage paths and capacitive loading. Shielding of sensitive operational amplifier inputs can be accomplished by using guard tracks around the input pins and applying a low impedance potential (bootstrapped to the signal level) to this shield.

(Back to top)

1.2. Power Supply Layout And Earthing Points

The layout of tracks can be critical to circuit performance. In particular, the layout of the system reference voltage tracks needs careful consideration. It must be remembered that every wire or track on the PCB has a small but finite series resistance and inductance and shunt capacitance. Therefore, we can generate IR drops along small tracks which can act as an additional error signal if the signal voltages share reference tracks with power supplies or switching digital signals. A critical area is the rectifier area of the power supply where large non-linear currents are generated by diodes charging large filter capacitors. The regulators need bypassing with tantalum and ceramic capacitors right on the IC pins and tracking layout around them is critical. Single point earthing should always be used where practical from a suitable reference point.
The layout and shielding of transformers and PCB wiring is very critical to minimise the problem of coupling unwanted error signals into the signal path. Coupling can be galvanic (resistive) as mentioned above, or electrostatic (capacitive) or electromagnetic (inductive).

(Back to top)

1.3. Electrostatic Coupling

Electrostatic coupling occurs due to the capacitive coupling between signal wires and say power cables. When two conductors are near each other, a capacitance forms between these conductors which increases with area (A) of the conductors and decreases with separation distance (d) between the conductors. A charge Q develops between the conductors, where Q = CV. (C is the capacitance between the conductors and V is the voltage between the conductors). If V changes, this causes a change in the charge on the conductors which implies a current flow through the conductors. The magnitude of the current flow depends on the rate of change of the voltage: I = C dV/dT Induced currents I1, I2 cause errors or beating effects

This coupling can be minimised by:

  • increase separation between wires
  • reduce area of overlap (different cable routes, cross at right angles)
  • use screen - connected to earth at one point to provide a path for the capacitively induced current. Use a highly conductive material for the screen (Cu or Al).

(Back to top)

1.4. Electromagnetic (Inductive) Coupling

Electromagnetic (inductive) coupling is one of the most frequent causes of problems with noise and interference.

"Mechanism"

Any conductor carrying current will produce a magnetic field around it. If there is an adjacent conductor, then a portion of this magnetic field will link with that conductor: flux line a' does not link with B but a' does and is linked or coupled with conductor B.

By Faraday's law, if the magnetic field produced by conductor A changes, due to the current in A changing, then a voltage will be induced along the length of B which is proportional to the rate of change of flux linking B. If A is a power cable and B a communication cable, then this induced voltage is a noise voltage.

This effect can be reduced by:
  1. Increasing separation
    • Increase distance d or run different cable through routes;

  2. Twist the two signal cables;
    • This reduces the effective circuit loop area (need balanced differential signals).
    • The induced polarity reverses with every twist and tends to cancel.
  3. Screen the cables.

    A screen earthed at both ends to provide a path for circulating current i' whose field will oppose the original magnetic field. This is assisted by the disturbing magnetic field producing eddy currents in the shield, which also opposes the original field. The screen should preferably be of magnetic material e.g. steel.

(Back to top)

One of the other difficult problems is to design the source so that it operates correctly into a wide range of capacitive and inductive loads. This is achieved by starting with an amplifier with a good stability margin and applying suitable feedback loops. It is important to actually measure the output current with a well designed low leakage isolating current transformer for feedback control (refer block diagram). These problems have been successfully addressed in the A.P.C.S. Precision Current Source, model ACS041. This affordable instrument provides a solution to those requiring a efficient, simple to use calibrator of high quality and accuracy in their testing and calibration of ac instrumentation. It provides the user confidence that their testing results are traceable to National physical standards as required by Quality Assurance Programs.

(Back to top)

2. Technical Overview And Description.

The ACS041 consists of 4 main functional areas.

  1. The power supply is of the linear regulated type with a toroidal power transformer specifically selected to optimise demand and efficiency. A 5 amp fuse is located on the rear panel for power circuit protection.
  2. The sine generator has internally trimable RC oscillators as it's clock which drives an A-D converter to generate a pseudo-sine wave. This signal is then filtered by 3 cascaded 2 pole low pass filters achieving better than 0.2% distortion. Internal full-scale amplitude adjustment (10 amp range) for the 50, 60 and 400 Hz sections are provided after the filters. These signals go to the frequency selector switch. From there the signal goes through attenuators (5 amp and 1 amp) to be selected by the range toggle switch. From there the signal passes through a further set of attenuators which are selected by the calibration rotary switch and on to the power amplifier.
  3. The power amplifier is configured to drive the transformer with a constant current as defined by the signal from the sine wave generator. Stabilisation of the output against temperature and load variations is provided by a precision current transformer in conjunction with active feedback circuitry. Comparators monitor the amplitude of the power amplifier output voltage and trip the over load indicator if such a condition occurs.
  4. The indicator consists of a precision shunt in series with the output. The signal from the shunt is amplified and converted to DC via a precision rectifier and 2 pole filter. The digital indicator in the front panel displays this value.