Test & Measurement


Overcoming the challenges of multi kilowatt source/sink testing

19 November 2014 Test & Measurement Power Electronics / Power Management

Satellites, hybrid electric vehicles (HEV), uninterruptible power supplies (UPS), alternate energy power sources and a number of other modern power systems supply power when it is opportune. They rely on bidirectional and regenerative energy systems and devices for storage in order to provide a continuous supply of power on demand.

Examples of these systems and devices include rechargeable batteries, supercapacitors, motor-generator systems, bidirectional DC/DC converters, battery management systems (BMS) and regenerative braking systems. Many of these systems and devices operate at multi kilowatt power levels.

When developing and manufacturing these systems, it is necessary to source and sink at kilowatt and greater power levels, which presents a formidable challenge for test engineers. The most common approach is to use separate instruments for sourcing and sinking, because of their availability. But this approach has shortcomings that can be overcome only by having the sourcing and sinking functions fully integrated into a single instrument or system.

Two-quadrant versus four-quadrant operation

Test engineers sometimes misunderstand what is needed for sourcing and sinking power for testing bidirectional and regenerative energy systems and devices. Bidirectional is often interpreted as bipolar, suggesting a bipolar power source is required for test purposes. These are actually two different things.

Figure 1. The four quadrants of an I-V plot.
Figure 1. The four quadrants of an I-V plot.

A unipolar, bidirectional power source is one that operates in quadrants I and II of a four-quadrant I-V (current-voltage) plot, as shown in Figure 1. It can source and sink current, but only with positive voltage. It operates as both a DC source and electronic load; it is a two-quadrant DC source.

In comparison, a bipolar power supply can transition through zero volts and operate with either positive or negative voltage. It can source power in quadrants I and III and sink power in quadrants II and IV.

As these bidirectional and regenerative energy devices and systems source and sink current and have unipolar voltage, a unipolar two-quadrant DC source is almost always adequate - and often preferable - for testing them. It is important that the two-quadrant DC source has fully regulated continuous operation when used as an electronic load in quadrant II. Simple down-programmer current sinking is not sufficient for adequately testing these bidirectional energy systems and devices.

Separate DC source and electronic load with deadband

The selection of suitable two-quadrant DC sources at multi kilowatt power levels is extremely limited. Engineers often resort to using separate DC sources and electronic loads to provide the combined power sourcing and sinking needed for testing their bidirectional regenerative energy systems and devices.

Independently, the DC source and electronic load are well suited for continuously sourcing and sinking power while providing good DC accuracy, stability and fast dynamic response, regardless of the DUT. This kind of performance is needed, as these DUTs are active and dynamic, alternately sinking and sourcing power, depending on their status and operating conditions.

One arrangement for combining a DC source and electronic load for source-sink operation is depicted in Figure 2, often used as a battery simulator system (BSS).

 Figure 2. Common DC source and electronic load arrangement for a battery simulator system (BSS).
Figure 2. Common DC source and electronic load arrangement for a battery simulator system (BSS).

A BSS is predominantly a voltage device; both the DC source and the electronic load normally operate in constant voltage (CV) mode. Their voltage is offset to provide a deadband so that they do not overlap in operation, with the electronic load’s voltage being set slightly higher than that of the DC source. Most often the DUT being tested with a BSS is a battery management system (BMS) but this setup is applicable for a variety of DUTs that need to be tested with a unipolar, two-quadrant DC source.

When the DUT is drawing or sinking power, the voltage is maintained by the DC source. When the DUT is generating or sourcing power, the voltage rises, the DC source cuts off and the electronic load becomes active operating in CV mode, clamping the voltage at a slightly higher level. A blocking diode at the output of the DC source is often required to isolate and prevent any reverse current from flowing back into the DC source when the DUT is actively sourcing power.

With this configuration, the source current is then directly read back from the DC source while the sink current is directly read back from the electronic load. There are a number of performance compromises with this approach:

• The DC source needs to be used with local voltage sensing as the blocking diode will destabilise it if remote sensing around the diode.

• The deadband between sourcing and sinking is high impedance.

• Voltage-level programming commands need to be sent to both the DC source and electronic load so that they track each other as the BSS voltage level is changed.

• A much higher level of complexity is generally needed for coordinating the activities of the DC source and electronic load during test.

• The electronic load has to transition between cut-off and active CV mode, compromising its dynamic performance.

• The voltage drop of the blocking diode is variable based on current levels and temperature, leading to the need to use a substantial deadband voltage of a few hundred millivolts between the DC source and electronic load voltage levels.

In particular, these last two items limit the two-quadrant flexibility, accuracy and general performance of this configuration to static operation. To compensate for the deadband voltage under static operation, the BSS voltage can be programmatically brought up or down as needed, to within a reasonably close voltage level. However, the deadband voltage step is inherent for dynamic transitions, further compounded by the electronic load’s CV mode transient crossover, as depicted in Figure 3.

Figure 3. Source-sink transition for separate DC source and electronic load with deadband.
Figure 3. Source-sink transition for separate DC source and electronic load with deadband.

Separate DC source and electronic load with overlapping source-sink operation

Fully overlapping operation can be used to avoid many of the problems associated with non-overlapping operation as previously described. A DC source and electronic load configured for fully overlapping operation are depicted in Figure 4.

Figure 4: DC source and electronic load with overlapped operation for a battery simulator system (BSS).
Figure 4: DC source and electronic load with overlapped operation for a battery simulator system (BSS).

Now the electronic load operates in CC mode instead of CV mode. The electronic load’s current setting is fixed at a value in excess of the maximum value expected to be sourced from the DUT. This way the electronic load always remains in CC mode, drawing a fixed level of current and power. The electronic load no longer has to contend with any mode crossover issues.

The DC source always remains in CV mode and is always sourcing current. Because of this, the blocking diode is no longer required. As a result, this BSS configuration is always in CV mode throughout the entire range of sourcing and sinking, free of the electronic load mode crossover and the deadband voltage transients affecting the BSS configuration that has non-overlapping operation. There are a few drawbacks as a result:

• The DC source needs to be considerably larger so that it can furnish both the maximum current and power required by the DUT, plus the full amount that the electronic load continually draws. As one example, the DC source needs to be more than twice as large for 100% current sinking.

• The electronic load is constantly dissipating full power, which is considerable for a large system.

• Measurements require reading back both the DC source and electronic load currents and taking the difference, often of two large values to get one small value. Measurement accuracy suffers as a result.

Integrated source-sink solutions

The disadvantages of configuring a power sourcing and sinking solution with separate DC source and electronic load are mitigated when the sourcing and sinking functions are integrated into a single instrument.

When integrated, these functions operate under closed-loop control to provide seamless, transient-free crossover between sourcing and sinking current and power. There is no need to constantly dissipate large amounts of power to accomplish this.

DC accuracy and dynamic performance are now optimised instead of compromised. Measurement performance is greatly improved by having a single measurement system for all currents. The main challenge has been a lack of instruments available that adequately address the test needs of today’s bidirectional and regenerative energy systems and devices, leaving engineers no choice but to use separate DC sources and electronic loads.

Keysight power supplies featuring integrated sourcing and sinking

Figure 5. Keysight APS family.
Figure 5. Keysight APS family.

The APS N6900A/N7900A DC power supplies, pictured in Figure 5, are tailored for the test needs of today’s bidirectional and regenerative energy systems and devices. They comprise energy efficient, 1U, 1 kW models and 2U, 2 kW models with 10% current and power sinking capability, which can be increased to up to 100% sinking with the optional N7909A power dissipator unit.

The wide selection of output voltages offered by these units addresses the diversity of today’s DUTs and applications. Their voltage- and current-priority operation give greater sourcing and sinking test flexibility, regardless of the DUT’s nature.

These features add up to a two-quadrant measurement system for accurate voltage, current, power, charge and energy measurements. Advanced sourcing and measurement capabilities in the N7900 can be used for creating dynamic output events, making transient measurements, continuously logging voltage, current and power, and more.

Advanced trigger signal routing with configurable logic is useful for creating application-specific controls, triggers and protection features to assist in solving particularly challenging test problems. The system’s modular architecture is scalable up to 10 kW for testing higher-power DUTs.

APS integrated sourcing and sinking for seamless operation

When using separate DC sources and electronic loads, built-in power sinking is not provided. The APS units, however, feature 10% power sinking built in. This can be increased up to 100% integrated sinking capability with the addition of one or two 1 kW N7909A power dissipator units, depending on the APS unit’s power rating. All that is required is to connect a power cable and a control cable between the two units to provide truly integrated operation, as shown in Figure 6.

Figure 6. Integrating an APS and an N7909A.
Figure 6. Integrating an APS and an N7909A.

As shown in Figure 7, even with 10X greater voltage resolution, an APS and N7909A provides seamless, stable voltage performance while transitioning between sourcing and sinking, when tested under conditions comparable to the separate DC source and electronic load BSS setup shown in Figures 2 and 3.

Figure 7. APS and N7909A seamless source – sink response.
Figure 7. APS and N7909A seamless source – sink response.

APS advanced measurement performance

Making accurate voltage, current, power, charge and energy measurements is a necessary part of testing bidirectional and regenerative energy systems and devices. These measurements are complicated and problematic when using a separate DC source and electronic load as a basis for a source-sink test solution.

As a minimum, separate currents read back from the DC source and electronic load have to be managed and assimilated. In all likelihood the DC source and electronic load will not have adequate capabilities for making accumulated charge and energy measurements, requiring the addition of external logging measurement capability with its associated complexity and issues.

The APS’s integrated two-quadrant measurement system fully complements its integrated sourcing and sinking capability, making accurate measurements a simple matter. As one example, Figures 8 and 9 depict the current, voltage and energy sourced and then subsequently recovered on a supercapacitor when using an APS and N7909A power dissipator unit for charging and discharging the supercapacitor while capturing its voltage, current and energy.

Figure 8. Super capacitor charging and discharging voltage and current.
Figure 8. Super capacitor charging and discharging voltage and current.

Figure 9. Super capacitor charging and discharging energy.
Figure 9. Super capacitor charging and discharging energy.



Credit(s)



Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

E-mobility: Navigate safety, interoperability and conformance
Concilium Technologies News
In this whitepaper, these challenges are discussed in more detail and the question is asked: How can EV and EVSE manufacturers navigate a complex regulatory landscape and deliver a quality product, without compromising time-to-market projections?

Read more...
Digitisers upgraded with pulse generator option
Vepac Electronics Test & Measurement
Spectrum Instrumentation has added the Digital Pulse Generator option to its ultrafast digitisers (with up to 10 GS/s speed) and arbitrary waveform generators.

Read more...
Network Master Pro to provide support of OpenZR+
Tamashi Technology Investments Test & Measurement
Anritsu Corporation has introduced the 400G (QSFP-DD) multi-rate module MU104014B that supports the new interface standard.

Read more...
Upgrade brings extra layer of detection to Fluke’s acoustic imagers
Comtest Test & Measurement
The firmware 5.0 update helps to boost efficiency and allows maintenance technicians to scan large areas quickly, and visually pinpoint technical issues before they become critical.

Read more...
Companies collaborate on EnviroMeter
Avnet Silica Test & Measurement
STMicroelectronics and Mobile Physics have joined forces to create EnviroMeter for accurate air-quality monitoring on smartphones. Time-of-flight optical sensing enables an accurate personal air quality monitor and smoke detector.

Read more...
Transmitting power to remote places
Altron Arrow Editor's Choice Power Electronics / Power Management
The new single-pair power over Ethernet (SPoE) allows for power and data to be transmitted over longer distances of up to 1000 metres.

Read more...
PCB test points
Vepac Electronics Test & Measurement
Maintaining these access points in the final production versions will prove invaluable during the life of the equipment for service, adjustment, and debug, or repair activities.

Read more...
RFID reader
Test & Measurement
The EXA81 from Brady turns any smartphone or tablet into a personal radar that can pick up radio signals from all RFID-labelled items.

Read more...
Proximity sensor with VCSEL
Avnet Abacus Test & Measurement
Vishay’s newest small package proximity sensor, the VCNL36828P, combines low idle current with an I2C interface and smart dual slave addressing.

Read more...
CNH data output devices for AI applications
Altron Arrow Test & Measurement
STMicroelectronics’ CH family of time-of-flight sensor devices feature compact and normalised histogram (CNH) data output for artificial intelligence applications requiring raw data from a high-performance multizone ToF sensor.

Read more...