Tech Tips January 2007

By Control Engineering Staff March 22, 2007

January 30, 2007

TECH TIP OF THE WEEK:

Learn from Jack Spratt.

The first stanza of a traditional English folk poem goes:

Jack Sprat could eat no fat
His wife could eat no lean
And so betwixt the two of them
They licked the platter clean

Today, we know that Mrs. Spratt’s arteries must have resembled the sewer pipe at my house when the tree root got into it last fall. When the plumber shoved a borescope in there, he found a sign saying: ‘NO FLOW!’

Mr. Spratt’s lifestyle, however, was much healthier. Wise production managers have learned to emulate it as closely as possible. The current management-consultant jargon term is ‘lean manufacturing.’

Control Engineering and automation equipment manufacturer ABB have published a white paper that explains lean manufacturing concepts and how control engineers can use the concepts to improve the automated-manufacturing systems they develop.

According to this white paper, entitled TEN STEPS to Lean Electrical Controls , by John E. Boyer, President of J. E. Boyer Co., ‘the major objective of lean manufacturing is to ‘banish waste and create wealth in your corporation.”

Fine words no doubt, but waste is not banished by a silver bullet and wealth is not created by applying a simple formula. Only hard work and persistent application of proven practices over a period of time yields positive results. This pragmatic approach has allowed many companies to benefit from lean manufacturing.

Boyer’s paper shares 10 specific lean-manufacturing strategies a manufacturer who uses electrical controls can deploy to help reduce waste in the total supply chain. If applied properly and completely, the result can be lower overall costs, faster throughput, higher quality, and better service for suppliers, customers, and customers’ customers.

Waste is any activity that does not add value to the company’s product or service. These non-value-adding activities do not add form, fit, or function to the product, and are not things customers would be willing to pay for if given the choice. Value-adding activities, on the other hand, are things that do add form, fit, or function, and are things that customers are willing to pay for.

If the time associated with all value-adding operations (fabrication, machining, assembly) is added up and compared to the time for non-value adding operations (moving, stocking, transacting, testing, waiting, etc.), the ratio of non-value to value would likely be very large – 500:1 or more. An electrical component that has 10 minutes of total work done on it could be ‘in the flow’ for 10 days! One goal of lean manufacturing is to reduce this ratio by including fewer non-value-adding activities.

To help understand the difference between value and non-value, consider in which category the following belong:

  • Starting/stopping a motor.

  • Marking the customer’s part number on a box.

  • Surge protecting a circuit.

  • Moving a component to an inventory location.

  • Stripping insulation from a wire to be terminated.

  • Wiring the electrical components in a panel.

  • Testing the machine before it ships.

  • Detecting the proximity of a machine component.

Some do not add value, from the final customer’s point of view, to the operation. For example, if a wire could be terminated without being stripped, it could save a lot of time and still have the machine operate appropriately. Better yet, if there was a prewired cable system available to snap into place, there would be no need to cut, mark, and bundle individual wires. Finally, what about redesigning to reduce the wiring required and still perform the electrical functions needed?

High performance electrical controls suppliers use lean manufacturing to reduce unnecessary motion, overproduction, defects, inventory, transportation, waiting, inspecting/testing, overprocessing, and other non-value adding activities in the supply chain to lower total cost and increase throughput.

The 8-page white paper TEN STEPS to Lean Electrical Controls , by John E. Boyer is available free of charge by visiting the web page www.lean-electrical-control.com/5 and downloading it in PDF form.

For more information, visit the Control Engineering website at www.controleng.com and type ‘lean manufacturing’ into the search box on any page.

SOURCE: TEN STEPS to Lean Electrical Controls, by John E. Boyer
www.rhymes.org.uk/jack_sprat.htm
www.simonsays.com
www.abb.com

January 23, 2007

TECH TIP OF THE WEEK:

Hedge your vehicle-guidance bets.

If walking around the ProMat 2007 expo floor taught me anything, it’s that the big trend in warehouse automation is automated guided vehicles (AGVs). These little puppies (while ‘little’ may not be apropos many of these AGVs, ‘puppy’ is getting to be more appropriate as the devices’ intelligence grows) turned up in nearly every booth I visited.

There are two problems AGV designers must solve—beyond the obvious one of giving them enough muscle to carry the loads required. First, they have to be able to navigate from point A, where they accept their load, to point B, where they drop it off. Second, they have to make the trip without running into or over anything significant, such as employees, walls, stacks of product, or other AGVs.

In general, AGV designers seem to be dealing with these problems separately. That is, they design a navigation system that simply navigates. Avoiding the supervisor’s toes is a separate problem usually handled by a separate collision-avoidance system, such as the ultrasonic proximity sensors mounted around all sides of the AGVs demonstrated at the Egemin Automation booth. I could write a book on either system, but want to briefly explain how combining multiple navigation technologies makes AGV navigation systems more robust.

Three AGV navigation technologies are in general use:

Wire guided systems are the oldest, having been in service since the 1940s. Typically, the installer attaches some physical guide, such as a magnetic tape, to the floor. The AGV carries a sensor that finds the guide structure and follows it around.

Laser guided systems have an infrared laser mounted in a turret that rotates around its vertical axis on top of the AGV, and targets strategically located in the space the AGV roams around in. As the laser beam reflects from each target, a photodetector picks up the echo returning to the turret. The time at which the system sees a target correlates with the line-of-sight direction to that target. A computer triangulates lines of sight to multiple targets, computing the AGV’s position and orientation in the horizontal plane.

Inertially guided systems use the fact that the distance moved is the second integral of acceleration. System designers use accelerometers mounted anywhere in the AGV to monitor linear and rotational accelerations in the horizontal plane. An onboard computer does the double integrals numerically to determine the AGV’s current position and orientation. This system is particularly useful in outdoor applications like the one Allan Quimby described at the Transbotics booth, where laser and wire guided systems are harder to maintain.

The problem, of course, is that any of these systems can get lost. If a wire-guided system loses contact with its guide, it has no way to find it again. Laser guidance works only as long as the laser can see at least some of its targets. Inertial systems work only until inevitable measurement uncertainties and numerical-integration errors accumulate beyond tolerable levels.

Some AGV manufacturers are starting to deploy more than one guidance method on each AGV. This strategy allows the AGV to crosscheck its position, catching, and in most cases correcting, guidance errors before they become problematical.

For example, a wire-guided AGV that also has a laser guidance system can find its way back to the nearest point on its guide track should it get lost. A laser guided AGV has no problem zipping along an isle between cases stacked higher than its turret.

As time goes on, expect to see additional AGV guidance devices, all with some failure mechanisms. Deploying more than one guidance system will keep automated warehouse facilities running smoothly.

Read also ‘ Sensors: Getting into Position .’

For other information about automated warehouse technology, visit the Control Engineering website at www.controleng.com and type ‘automated guided vehicles’ into the search box on any page.

C.G. Masi, Control Engineering Senior Editor, charlie.masi@reedbusiness.com

www.jerviswebb.com
www.egeminusa.com
www.transbotics.com

January 16, 2007

TECH TIP OF THE WEEK:

Keep your feet on the ground.

Even differential measurements have a ground reference. Source: Control Engineering
Floating differential measurement circuits include a virtual high-impedance path to ground. Source: Control Engineering

Every control system starts with a measurement of whatever it is that needs to be controlled. And, whether you’re measuring temperature, pressure, position, speed or anything else, most automated control systems use a voltage level analog of the process parameter the system’s trying to keep under control. That means you’re usually making voltage measurements.

Voltage measurements are always made with respect to a reference point, which is commonly called ‘ground.’ Even differential measurements, which ostensibly measure the voltage difference between two points, actually use an analog-computer circuit that computes the difference between voltage readings at those two points, each measured with respect to ground as Figure 1 shows. Even when doing floating (non-referenced) differential measurements, there is always a virtual high-impedance path to ground as shown in Figure 2.

Whenever you make any voltage measurement, therefore, it is critical that you keep in mind where your ground reference is.

Suppose, for example, that the system is closed-loop control of temperature using a fast-acting thermocouple in the circuit of Figure 3. Everything looks okay, but the temperature never seems to stabilize. It’s as if the setpoint keeps jumping around!

What’s happening is the signal leads are acting as a big antenna to pick up ambient EMI, which appears as millivolt-level noise across the virtual high-impedance path to ground. The reference junction rectifies that high-frequency noise, making it appear hotter. Whenever the EMI level changes, the reference-junction temperature appears to change, upsetting the control loop.

All the methods for eliminating EMI, as well as many other measurement problems, start with establishing a solid ground point and maintaining it throughout the circuit. To avoid such problems in the first place, always know where your ground is.

Poorly shielded measurement signal paths are open to EMI. Source: Control Engineering

Planning a ground system is as much an art as a science, and the best approach depends on the application. For example, I once worked as a test engineer with a company that mainly developed dc high voltage equipment. The bandwidth for their control signals was invariably in the extreme low frequency (ELF) range from dc to a few thousand Hertz. One day, the head of the electronics department brought out a prototype oscillator for a Cockcroft-Walton (CW) voltage-multiplier power supply that they just could not get to work. They were so frustrated that they were willing to ask anyone—even we test engineers—for help.

The grounding system was exactly right for ELF frequencies: fat, heavy green wires connecting each component to a common ground point in a star pattern. The problem was that the CW operating frequency was a few megahertz (in the low RF range). Being an old RF guy, I realized that those wires all acted as RF chokes. While they might have looked like short circuits at ELF, they blocked the oscillator’s operating frequency very effectively. There was no RF ground, anywhere!

What they needed for that frequency range was a Faraday cage with every ground tied to the metal chassis by as short a lead as possible. Once they rewired the chassis as if it were a ham radio transmitter, the prototype oscillator worked perfectly.

In general, the basic rules of thumb are:

  • Use a strategy appropriate to the signal bandwidth;

  • View signal paths as transmission lines;

  • Use twisted pairs for all transmission lines;

  • Enclose each transmission line in a shield;

  • Ground each shield at one end only;

  • Ground single-ended measurement equipment at the detector/signal-conditioner end, not at the sensor end.

These rules of thumb work for every frequency range up to about 100 MHz. For higher bandwidths, you have to start planning for standing waves in the transmission lines, and that’s beyond the scope of this article.

For more information about, visit the Control Engineering website at www.controleng.com and type’grounding and shielding’ into the search box on any page.

Have a really good idea to share with fellow engineers? Email your control-system tip to controleng@reedbusiness.com with ‘Tips’ in the subject line and question, and full contact information in the body of the email. If we use it, we’ll send you an ‘Engineer and proud of it!’ pocket protector.

January 9, 2007

TECH TIP OF THE WEEK:

Use variable-speed drives to save energy.

The only reason to specify a variable speed drive in a control application is because you expect your control system to vary the motor’s speed, right?

Wrong!

According to Akseli Savolainen at control-compent supplier ABB, variable-speed drives are a green technology as well. That is, users can save energy costs and reduce greenhouse-gas emissions by using variable-speed drives to match motor output to process demands.

‘So much energy is wasted by inefficient constant speed and mechanical control mechanisms,’ according to Savolainen, ‘that every industrialized nation around the world could make several power stations redundant simply by using AC drives and high-efficiency motors.’

Controlling motor speed makes matching process demand and machine output possible. Control of a pumping system provides a good example. Rather than controlling flow in the normal way via a throttle valve, a variable speed drive can be more efficient. Suppose the flow rate is to be half maximum, reducing the speed by reducing the motor speed reduces energy consumption by a factor of eight. The throttle valve, which reduces flow by introducing an impedance, actually increases the motor’s load, increasing energy use.

Pump operation provides an example of another way to save energy with variable-speed drives. The traditional way to maintain liquid level in a tank is by turning a pump on when the level reaches a low set point and off when it reaches the high set point. With a variable speed drive, one can control pump output to very nearly match flow requirements. Since startup current can be as much as 7 to 8 times full-load value, slowly varying the motor’s speed as requirements change can save considerable energy compared to duty-cycle control.

For more information about how to save energy using variable speed drives, contact Mark Kenyon at ABB. mark.kenyon@us.abb.com or visit the company’s website at us.abb.com .

For more information about running control systems efficiently, visit the Control Engineering website at www.controleng.com and type ‘energy savings’ into the search box on any page.

Have a really good idea to share with fellow engineers? Email your control-system tip to controleng@reedbusiness.com with ‘Tips’ in the subject line and question, and full contact information in the body of the email. If we use it, we’ll send you an ‘Engineer and proud of it!’ pocket protector.

January 2, 2007

TECH TIP OF THE WEEK:

Electrical noise influences.

The reliability of a control system can be reduced when it is subjected to unusually high amounts of electrical noise, but proper power and grounding can minimize the effects.

Various components of a system can be affected to different degrees. The following observations have been made during system troubleshooting:

  • Properly installed, standard, non-communicating (called traditional or classic) I/O products that use 4-20 mA analog signals only are typically not affected by electrical noise, except high frequency noise.

  • High frequency electromagnetic noise generated by ac drives, insulated gate bipolar transistor (IGBT) power switching devices, and other sources can affect all control products without bias to a particular vendor.

  • Digital data signals used in bus communication systems are susceptible to electrical noise.

These effects can be minimized by:

  • Using isolated ac power sources

  • Grounding at single points

  • Minimizing undue influence on signal wiring from stray magnetic fields

  • Selecting appropriate cables and pathways, including adequate cable separation.

Source: ‘Control Engineering: Control System Power and Grounding Better Practice,’ by David Brown, David Harrold, and Roger Hope; Elsevier Inc. 2004.