Monthly Online News For The Successful Thermal Manager

 Volume #3: October 2006: www.jarothermal.com

 

Inside this Month’s Issue:
(click on a topic below)

 

 

   RETURN TO JAROTHERMAL HOMEPAGE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The Technology Forum

As increasingly complex, electronics applications continue to shrink in size, thermal management continues to grow in importance. To avoid malfunctions or failures which inevitably arise in “heat challenged areas”, the importance of accurate temperature measurement can not be overstated by today’s “thermally-conscious” engineers. In effect, this is forcing a creative, rethinking of the entire pre-design process.

 

According to Gayle Bullock, who works with National Semiconductor Corporation’s Analog and Networking division, National Semi  is now proactively addressing thermal management with a new, forward-thinking technology called TruTherm™”.

 

TruTherm™ technology solves the problem of inaccurate remote temperature readings caused by variations in diodes located within deep sub-micron microprocessors, micro controllers, application-specific integrated circuits (ASICs) and field-programmable gate arrays (FPGAs). Inaccurate remote temperature readings can lead to higher acoustic noise and reduced system performance. National's TruTherm™ products improve the accuracy of temperature readings, allowing designers to achieve higher performance and efficiency in their applications, while lowering cooling-fan speed, reducing acoustic noise and extending system life. TruTherm™ allows managers to successfully monitor complex sub-micron cores, in order to optimize heat dissipation. One such example of this new TruTherm™ technology (in action) is the LM94 Hardware Monitor.”

 

To set fan speed, the LM94 has two pulse-width modulation (PWM) outputs that are each controlled by up to six temperature zones. The fan-control algorithm can be based on a lookup table, a proportional/integral (PI) control loop that supports a temperature control (Tcontrol) function or a combination of both.

If it can be measured, it can be controlled

According to JaroThermal’s marketing director: “Our speed controlled DC fans optimize technologies such as National Semiconductor’s TruTherm™. JaroThermal fans use pulse-modulated speeds, which are controlled by an external SPWM signal. In line with the push for better measurement and control, our fans can run at lower speeds. In fact, we currently utilize1.3 ~ 5 points of speed control on the production line. This makes for a more energy efficient (lower power consuming) device. In addition, it increases the life span of the components. By utilizing high-frequencies, our fans optimize coil switching. At the same time, they maintain a soft (silent) operation.”

Fan speed can be measured with tachometer inputs which includes limit and status registers for all measured values. It also works well with the monitoring of dynamic power supply voltage for the processor (VccP), dual processor thermal throttling monitoring (PROCHOT) functions and general purpose input/output (I/O) pins.

PWM, as it relates to JARO’s DC Fans

Features

Ø      Single-phase full-wave driver

Ø      SPWM variable speed control

Ø      Soft switching (silent)

Ø      Low speed setting

Ø      Frequency Generation output

Ø      Division frequency setting

Ø      Built-in current limit (60% modulated )

Ø      Build-in Auto-restart & Lock protection (C saving)

Ø      Low power consumption and high driving efficiency (Vol under 0.4V)

Ø      High driving current capability

Ø      Notebook PC useable (5V only)

Tachometer Background

On July 13, 2006, a New York Inventor developed an accurate Fan Tachometer Reading Generation Method. According to the U.S. Patent & Trademark Office: "The duty cycle of a pulse width modulated (PWM) signal provided by a signal generator may control the speed of the fan. The fan may generate tachometer pulses that may be used by a tachometer reading-unit to monitor the revolutions-per-minute (RPM) of the fan. Very low frequency test pulses may be generated and provided through a sampling signal multiplexed with the PWM signal to sample the fan generated tachometer pulses even when the PWM signal is low. "

 

  

 

 

 

 

 

 

 

 

 

 

 


 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hybrid Vehicles and The Future

 

THE AUTOMOTIVE INDUSTRY

 

According to a recent automotive industry report[1], the Hybrid Electric Vehicle industry is preparing for a growth spurt.

According to the report, the North American hybrid business is growing by leaps and bounds. As North American hybrid programs move towards the manufacturing stage (perhaps “assembly” is a more accurate term), overseas suppliers are gearing up for the predicted contracts. Leaders like Toyota Motor Corp. and Honda Motor, Co., already enjoy a developed supply network. By default, their keiretsu suppliers are also in a good position to benefit from the growing hybrid business.

On the other hand, no one can reliably predict how fast the hybrid business will actually grow, due to the volatility of today's fluctuating gas prices. Either way, the long term trend for increased hybrid business seems undeniable.

According to Vice President of Sales and Business Development at TRW Automotive Inc., Peter Lake, “manufacturing requirements for hybrid vehicles can be quite different than that of conventional vehicles.”

"Fuel efficiency, packaging flexibility and weight reductions are key factors," Lake said. "And with smaller volumes, vehicle manufacturers want to reduce development costs for individual systems and components.”

“The supplier requirements for hybrid electric vehicles differ from those for conventional vehicles," he said earlier in the year, in a statement accompanying the introduction of a collection of TRW hybrid technologies. “In a micro hybrid, the internal combustion engine gets an assist from features such as a stop-start system, an efficient generator and regenerative braking,” he added.

THE MILITARY

The idea of a growth spurt inside of the United States military is more illusive than ever. According to a new military report[2] on the subject, it has been very difficult to adopt commercial technology for military use. The effort is made more difficult by the fact that military vehicles operate in unique military conditions, with larger electric and use-load requirements than standard commercial vehicles. With size and weight challenges in most military applications; small, light-weight, high powered components can be hard to come by.

The other concern is cooling. Most cooling systems do not lend themselves towards the development of hybrid technology. The very hot environment underneath the armor appears to be one of the culprits.

Despite the fact that these challenges have always existed, some companies are bucking the trend by investing in the future of hybrid technologies.

One company, Oshkosh Truck Inc., created a brand new, hybrid electric diesel power train, as a replacement for an existing conventional diesel engine. Dubbed as the HEMTT A3, this “ProPulse” hybrid system propels vehicles with 140-hp ac induction motors, with one mounted on each of the four axles[3].

The company’s technical director is quick to distinguish between the HEMTT A3 and standard commercial hybrid vehicles. "Unlike consumer hybrids, we don't have a mechanical drive train at all," notes Dan Binder, Oshkosh Truck's technical director. “The HEMTT A3 has no transmission, no drive shaft, no transfer case and no torque converter. Instead, each motor directly drives a set of wheels, via the differential on each axle.” [4]

SYNOPSIS

The following paragraphs appeared in a September (05) article in Military & Aerospace Electronics[5]. They are still relevant to the current thermal issues at hand.

“Will chassis designers be able to handle the new thermal-management challenges in military and aerospace systems? The "more functionality in less space" market force is not new, yet there are many new factors that affect the way the packaging enclosure company provides solutions. Higher currents, hotter processors, and more switched-fabric technologies are making their way into military and aerospace systems. How will this affect the packaging?

Higher system performance-often coupled with faster and hotter processors, as well as denser packaging-can be two major nemeses for the chassis designer. Dissipating 200 watts per slot like in AdvancedTCA chassis for communications systems is certainly a challenge, but it is not just the watts; the size is a factor, too.

Military and aerospace applications can bring added challenges not often found in other industries. Components for handling EMI and shock and vibration can impede airflow. Physical space constraints, the effects of altitude, and limitations of the environment can make the thermal management even more challenging. Dynamic packaging companies, however, are finding ways to provide cooling solutions for each new challenge.

Space constraints are common in many industries. However, in many military and aerospace applications the physical space is a true limitation. Saving space is not just a convenience or a bonus; it is an absolute necessity. The packaging become denser and denser, and cooling the chassis demands innovative design.” [6]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Thermal Simulation & Testing: Chassis Design

In order to optimize thermal management solutions, you need the right tools.

This is confirmed by the most experienced chassis design teams of military COTS (Commercial Off The Shelf systems).

One of these (cooling solution) tools is thermal simulation and testing.

Software simulation programs like the Icepak® suite from Fluent Inc. in Austin, Texas are design tools for modeling electronics cooling. According to Fluent:

“Advanced thermal design is made easy with powerful thermal modeling software to transform your electronics designs. The ease-of-use, speed, and accuracy of these tools allow you to save time and money by reducing the need for physical prototyping and cutting time-to-market.“

According to an excellent analysis of this topic by Mr. Ram Rajan (with Elma):

 “The chassis design engineer can then determine the best way to eliminate the hot areas by optimizing the location of fans, baffles, air intake, and exhaust during the design phase. The software provides the ability to perform several "what if " scenarios to arrive at the ideal design. It allows designers to build conceptual 2-D/3-D models of an enclosure and run preliminary simulations. With animation capabilities, it can illustrate the airflow through an enclosure and scan the heat distribution and airflow on a slot-by-slot basis.

Thermal simulation can also help locate hot spots in a chassis design, where sensors then can make accurate measurements. Then thermal analysis can determine how fans should be grouped together.

Working early with the board vendors can help optimize a solution for the exact configuration of the chassis. The chassis vendor can analyze how to best cool the vendor's boards within the confines of the particular chassis and components. If the customer's card is also modeled, the chassis designer can verify the cooling and make sure the hotter components of the card are placed in cooler parts of the chassis (or modify the cooling arrangement of the chassis).

Simulation also can be helpful in balancing airflow and static pressure. Because of static pressure, slower fans can at times provide better airflow and cooling than faster fans in a push-pull setup

Other newly addressed issues, according to the experts at Fluent:

Increasing trace and via density on printed circuit boards makes it necessary to model the effects of traces in greater detail. Prior attempts to model traces merely approximated them using a lumped value of conductivity for the entire board. This approach worked so far as the trace density was low, and the roles of thermal vias were minimized. With the 10-20 layer boards of today, this is no longer a suitable approach, especially, if local hot spots play an important role in determining the junction temperatures for the package. With the tighter margins that designers have to work with, accuracy in temperature prediction has become increasingly important.

 

Icepak 4.3 introduces a new capability whereby the effect of the traces is captured on a local level using orthotropic conductivity values that change depending on the location on the board. The conductivity values computed depend on the local copper and the orientation of the traces. The pictures below show the temperature contours on a board with a specified heat flux on the top and fixed temperature on the bottom. Figure 1 shows the traces on the printed circuit board, Figure 2 is the result of a temperature solution on a coarse mesh, while Figure 3 shows the results of a temperature solution at a finer resolution.

 

 

Figure 1. Trace details in a printed circuit board

Figure 2. Temperature contours – low resolution

 

Figure 3.  Temperature contours – high resolution

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EMC Across The Board

 

 

http://www.cieonline.co.uk/

 

Potential electromagnetic compatibility problems in circuits should be considered early in the design to avoid difficulties later. Steve Rogerson looks at some of the options.

Designers of electronic circuits naturally concentrate on functionality in the early stages. The most important aspect is that the circuit will do what it is designed to do. Then they will look at board layout, making sure that there is a way to connect these circuits up in the correct manner. In the latter stages of design, aspects such as heat dissipation will be considered and then, almost as an afterthought, will the problems of electromagnetic compatibility (EMC) be looked at.

This could be a mistake, especially for highly populated boards as often ensuring EMC may mean more than just sticking the circuit in a metal box. Sometimes different parts of the circuit will need different levels of protection and so some areas will require individual cans that have to be fitted into the board. If space hasn't been allowed for this, it could be quite tricky in the final design processes to accommodate, and may involve a redesign of some sections of the board.

"With EMC you can end up retrofitting the design," says Paul Duxbury, consultant engineer with Flomerics. "Thermal, mechanical and EMC issues need to drive the design more, and so should be considered earlier."

The key areas to look for are potentially noisy circuits that will emit radiation and sensitive circuits that are susceptible to radiation. Keeping the two apart is often a good start, but what also needs to be considered is the environment in which the circuit is to be used; a board that will be in an industrial area will probably need more protection, for example.

Filters at the input and output stages of the circuit can help; even a simple surface-mount PCB filter will be useful if the threat level is low. At its simplest, any capacitor or inductor can be a filter and these can be designed in to the circuit.

"Mains-borne interference needs to be attenuated early so as little of this as possible enters the product," says Andrew Skinner, advanced development manager at Lambda. "For interference that is being generated within the product, this should be dealt with as close as possible to the source so that harmful effects are reduced."

Tips for avoiding noise
Noise sources can be magnetic or electrical. A magnetic field will generate a voltage if it can flow through a loop, and any track that goes round the board and comes back on itself creates such a loop. Circuit design therefore needs to be done in a way that keeps such loops as small as possible. For example, putting the signal and return elements on different layers can reduce the size of loops.

Using the ground plane as signal layer may be desirable but other signals may also be going through the ground plane creating holes around which current can flow, damaging sensitive circuitry. Moving a hole to the edge of the board stops it creating a loop, but for some circuits it is better to keep the ground plane intact otherwise its effectiveness is reduced.

For complex filtering, a component may be needed at the input to screen everything but the signal that is wanted. However, such perfection is never achieved in practice and a balance of signal versus noise needs to be found. Again, this filter will be a collection of capacitors and inductors. These can be custom built to match the circuit but many engineers can get away with a standard design that is closest to their requirements. Cost is also a factor here.

"Any EMC component to be added is also an added cost," says Alex Tennant, sales manager at Spectrum Control. "So people should develop the equipment to keep this down to a minimum. Most good designers analyze the potential threat and then decide what filters would work well in that environment."

Here, a lot depends on the extent of qualification testing that is needed. For example, the military and aerospace qualification procedure is long and expensive, and not something that a company would want to go through more than once for a particular product. In this case, therefore, all the filtering should be sorted out before the qualification process starts.

For less critical commercial applications where the final cost of the product is a major factor, the number of extra EMC components needs to be kept down. So a product may be put through the procedure with low EMC control in the hope that it will still pass. If it fails, then swapping out a standard D-sub connector for one with built-in filtering can be done quickly. This can be proved to work in advance by coupling the filtering element to the connector.

If some circuits are going to require their own shielding cans, this must be considered in association with cooling the equipment.

"It may be difficult later in the design to allow provision for a shielding can," says Duxbury.

Some cans themselves can resonate if not designed properly. A can resonating at the same frequency as what it is protecting can increase rather than reduce the problem. Thus the use of an absorbing material inside the can should be considered.

Likewise, just putting a bank of PCBs in an enclosure is unlikely to be enough. The actual presence of the boards will change the characteristics of the box.

"It is useful for designers to understand this so they know how to place the boards," suggests Duxbury. "If you have a board that is sensitive, you want it in a part of the box that is quiet in terms of emissions."

Problems solved
On the enclosures themselves, there have been no major developments for a decade or so.

"We solved all the problems in the early to mid-1990s," says David Bowring, product manager at Rittal. "And all the changes from then have been formalized within IEC specifications, so there is a standard way of doing things with products from different manufacturers fitting in the same space."

Many in the market want a system-wide approach rather than just off-the-shelf components. For example, look at a telecoms system. A gas discharge tube in the primary position is very robust in that it will protect against a lightning strike but it does not react fast enough to protect sensitive components. A semiconductor thyristor in the secondary position is fast enough but not sturdy enough to guard against a lightning strike for more than a few milliseconds. Thus, a system approach would use the thyristor as a trigger for the gas discharge tube. To coordinate this, a third element is needed, such as an overcurrent device.

"It is a system approach," explains Werner Gretzke, sales manager at Raychem. "You need all three elements to make it work."

But looking at the emissions for a complete system can be harder because there are multiple sources, and these might affect each other. There are software tools that can help, but the final product will still need testing for compliance. Good design can reduce the risk of failure but not eliminate it.

"A first-time pass is the ideal," says Duxbury, "but it should be OK if you can cut it down to one or two changes."

Tennant adds: "A lot of top level design work uses simulation software, but it is difficult to generate a simulation of the whole system. It is very much a guide."

KEY POINTS
- EMC should be considered early in the design process!

- Different parts of the circuit will need different levels of protection - some will require individual cans

- Look out for noisy areas that will emit radiation, and sensitive areas that are susceptible to radiation. Keeping the two apart is probably a good start

- Filters at the input and output stages can help

- Reduce magnetic noise by making any looped paths as small as possible. Put signal and return elements on different layers

- A can resonating at the same frequency as what it is protecting can increase rather than reduce the problem

- Good design can reduce the risk of failure but cannot eliminate it.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


[1] Bradford Wernle. Automotive News. Detroit: Jun 19, 2006. Vol. 80, Iss. 6208; p. 28 (1 page)

[2] Technology Limitations Stall Military Hybrids, David Axe. National Defense. Arlington: Sep 2006.Vol.91, Iss. 634;  pg. 31, 3 pgs, ENGINE CHECK:

[3] Small, Cool, Powerful, Joseph Ogando. Design News. Boston: Aug 14, 2006.Vol.61, Iss. 11;  pg. 56

 

[4] Small, Cool, Powerful, Joseph Ogando. Design News. Boston: Aug 14, 2006.Vol.61, Iss. 11;  pg. 56

 

[5] Solving thermal-management challenges in military and aerospace applications, Ram Rajan. Military & Aerospace Electronics. Westborough: Sep 2005.Vol.16, Iss. 9;  pg. 14, 1 pgs

 

[6] Solving thermal-management challenges in military and aerospace applications, Ram Rajan. Military & Aerospace Electronics. Westborough: Sep 2005.Vol.16, Iss. 9;  pg. 14, 1 pgs