Circuit protection basics - Part 2: Case study - understanding and using predictive analysis

New research-based technology protects against multiple power grid disturbances through microprocessor driven technology and predictive technology

David McGirt, V.P. of Engineering, Innovolt -- EDN, February 7, 2012

The circuit protection solutions that most electronic equipment designers and facility managers are currently implementing are missing a key component of true, mass-market protection: predictive analysis and flexible implementation.

In part 1 of this series on circuit protection, I covered the basics, the issues and the component and design solutions that are available to resolve them. Now let's look at a more advanced and reliable circuit protection technique that can truly meet the needs of today's complex electronic equipment.

Whereas previously discussed technology solutions, such as surge protectors and UPS, leave a wide gap in levels of protection, new research-based technology easily protects against multiple power grid disturbances through microprocessor driven technology and predictive technology.

As I mentioned, UPS offers some technology advancements, protecting against disturbances on the higher end by isolating electronics from the grid and powering them by battery. However, UPSs are extremely expensive and too large for integration, resulting in a costly and inflexible solution. Due to the high expense associated with UPS, there also is a high demand for inexpensive surge protectors, which effectively shield electronics and circuits from less than one percent of power disturbances. The combination of UPS and surge protection, while reliable for a select few, leaves a wide gap in the market place for electronic equipment protection.

Intelligent electronics protection such as Innovolt's entered the market to fill this gap. Its technology guards against damage from 99.5% of power disturbances including an infinite variation of voltage sags, brown outs, over voltages, power outages and damaging current inrushes. The predictive modeling, which anticipates and responds to disturbances, is a key and unique component.

In part 2 of this series, I will look specifically at the core innovation, predictive technology, how it works and implementation.

Understanding predictive analysis
The system is based on long-term analysis of grid data, as defined by I-grid (www.i-grid.com), a web-based power quality and reliability monitoring and notification system developed by Dr. Deepak Divan, who also is founder of Innovolt. I-grid uses I-sense monitors to capture and transmit power data. With this collaboration and a selection of partnerships with independent research centers, Innovolt has amassed millions of points of data, and thus can effectively identify disturbances on the power grid, opening up the gateway of predictive analysis.

To better understand predictive analysis, let's review the key points. To increase the protection of circuits from power anomalies, we need to predict what the power grid is doing or what it is about to do. In order to do this, we need to analyze incoming power. However, the need for predictive analysis is two fold. In addition to analyzing incoming power, there is some thought that must go into prediction. For example, not all power disturbances affect the grid the same.

You probably have a freezer in your garage that was built in the 60s or 70s, right? Inside your house, you may have updated to a digital fridge. And, it probably hasn't been serviced in several years. Meanwhile, the fridge in your house has to be serviced every two to three years.

Based on knowledge of the power grid, we can state that protecting an old 60s fridge and protecting a digital fridge are very different. With that said, the analysis has to do with being aware of what a circuit protection device is protecting and in turn, the custom guidelines that need to be applied to the protection technology. Once the guidelines are set based on knowledge and type, the technology uses a microprocessor to continually monitor the power line condition.

How it works
Let's take a look inside an example of this technology. The protection technology resides between the public power grid and the equipment that is being protected. It uses a series of patented algorithms and protocols to first recognize potential power issues, then quickly remediate these issues before any damaging effect occurs.

The technology platform is broken down into a selection of distinct and protected components that work individually and together to improve electronics lifespan and reliability (Figure 1).


The first component is measurement and signature creation. Based on knowledge of the grid both through internal research and partners, signatures are set for each device to measure and act accordingly. The technology rapidly monitors and measures the details of the incoming power to create signatures that can be profiled in real time against known disturbances. At the core, the signatures are defined by Innovolt; however, the technology is flexible and can be defined by a customer or end user.

The signatures, being a certain voltage or irregularity over a certain threshold, are detected and documented. Once the signature is detected, the next component kicks in: predictive processing. A central microprocessor system compares the real-time power signatures to profiles that are known to cause damage to electronics. With millions of potential combinations of profiles and signatures, each with a representative damage signature, the process then determines the most effective steps for remediation and activates a core protection circuitry.

Another key component of the predictive analysis stage is diagnostics. An overview report details the protection device location, the equipment manufacturer, product details and occurrence type. In addition to an overview report, as seen below in Figure 2, a specific report for an individual device is available showcasing a more detailed breakdown of events.


Comprehensive, real-time diagnostics about disturbances enable operators to custom define the equipment profiles to tailor the protection to their electronic application. It also details the area's power quality. This data is useful for service technicians, allowing for the tracking of power related events or trends that often lead to equipment failure or downtime.

The last step is the core protection circuitry, which provides a buffer between the damaging power event, its effects, and the protected electronics. The circuits act in concert with the predictive processing to provide a significant high value workload, with a small power and size footprint. Fundamentally, the core protection circuitry is all about diverting power or turning off power. It's the active intelligent decision process that is taking place.

To put this in a real world perspective; imagine being in an office workroom that includes two multifunction printers, two servers, a computer and phone system. Somewhere on the power line, a voltage sag, the "silent killer" of disturbances, which happen 30 - 100 times a year on average, has occurred.

Going back to part 1 of this series, we know the consequences of the recovery of a voltage sag - a huge current inrush as the voltage returns to normal, negatively impacting circuits of electronic equipment by inducing an inrush that far exceeds the component rating. However, the electronics are using predictive analysis technology, which is continually monitoring the power line with its microprocessor. An example of this technology is included below in Figure 3.


The technology immediately detects the voltage sag and through predictive analysis, works quick enough to isolate the connected equipment from the inrush. The circuitry now makes its intelligent decision. It can divert the power or turn off the power. In this case, it cues to derail the inrush and its core protection circuitry goes into buffer mode, absorbing the inrush. This allows the printers, servers, computer and phone system to continue to operate. Its high volume workload can handle the "silent killer" as much as needed, whereas the electronics in the office workroom over time would weaken and burn out.

Implementation
In addition to offering a holistic approach to circuit protection, deployment isn't limited with this platform. The technology is designed as a daughter-board, sub-assembly or chip-set to be integrated into OEM electronics. It can either be placed in front of electronic power supplies to shield the equipment from all power anomalies, or it can be integrated into the products themselves.

Conclusion
We have come a long way in solving the problem of circuit failure due to grid issues and power disturbances. Using predictive technologies to know that a grid problem is coming before it occurs and taking steps to protect equipment in advance of the issue is cost-effective and a meaningful and proven long-term safeguard for electronics.

Over time, the reputation of highly protected electronic equipment will be associated with longevity and reliability, making consumers realize that products aren't necessarily poorly or cheaply constructed. Microprocessor-based and predictive analysis technology is changing the landscape of electronics protection and management and the entire industry, beyond the individual products and manufacturers, will benefit from the improved reputational qualities.

About the author:
David McGirt joined Innovolt in spring of 2011 as the senior vice president of engineering. He has nearly 20 years of experience in technology and operational management within telecommunications and data center environments around the world. McGirt is focused on leading Innovolt's technology development and expanding the technology to fit new markets, as well as develop its cloud capabilities. Previous to Innovolt, McGirt was the Founder and Chief Technology Officer of Xiocom Wireless, a global broadband company, as well as the Chief Technology Officer of ITC^Deltacom & Quality Technology Services. He earned his Bachelor of Science from Auburn University.