THE MAGAZINE FOR COMPUTER APPLICATIONS
Circuit Cellar Online offers articles illustrating creative solutions
and unique applications through complete projects, practical
tutorials, and useful design techniques.

LOOKING AT THE SPECS

by Gerard Fonte

Start ı Are You Flexible? ı Whoıs the Boss? ı So, What are Your Options? ı What are the Hazards? ı To be or Not to be Specified? ı Unhappy Customers ı Sources and PDF

WHAT ARE THE HAZARDS?

It is important to understand the difference between hard and soft specifications. Basically this is a measure of importance. Some specifications are more important than others. It is important that your voltmeter read the proper AC voltage at 60 Hz. Itıs not so important that it read properly at 200,000 Hz. Of course, your meter probably has a specified high-frequency limit. Have you ever tested that? Was it an important point in your decision to buy it? Probably not.

The most important (or most difficult) specifications deal with personal safety. It is absolutely critical that these specifications be conservative. What is the AC leakage current? What fuse rating is proper? Do you know how to properly specify a fuse? Are the power supply components rated properly? If not, an exploding filter capacitor or fire could result. If a part fails, will it cause the product to become a shock hazard? Will a person be at risk? Always search for possible dangerous situations. Most likely they wonıt occur, but you should be prepared.

For example, I was asked to design a custom doorbell to be installed by the user. It worked on the 16-VAC doorbell transformer, but I knew that someone somewhere was going to connect it to 110 VAC. So, I took a little extra time and designed the system to tolerate 110 VAC. It only required changing two quarter-watt resistors to half-watt parts.

Did you notice the change in perspective? I went from creating product specifications to identifying part specifications. This is another point. Specifications of a product depend on the specifications of the parts. Although obvious, it is surprising how often it is overlooked. I frequently see designers fail to specify a proper part, and then they wonder why the final product fails to meet expectations.

One of the most common and disturbing examples of this is in simple linear power suppliesıa transformer, bridge rectifier, and capacitor. Many times the capacitor and rectifier are underrated. Suppose the transformer is 24 VAC. A 35-V filter capacitor should be okay, right? This is the standard 50% voltage overrating for power supply circuits. Wrong.

The unloaded voltage will almost certainly be greater than 35 V. Remember that the capacitor (without load) will charge up to the peak voltage. This is 1.414 times the rated (RMS) voltage, or 33.9 V. Also, the transformer is rated under load. Without load, the transformer output is 3 to 5 V greater than its nominal rating. Even after subtracting the rectifier diode voltage drops, the no-load voltage will still be greater than 35 V. And, donıt think that there wonıt be occasions when the load will be removed. Donıt you sometimes remove the load to check the power supply? How would you react if the power supply filter capacitor exploded when you did this? A 50-V capacitor is the minimum voltage acceptable. A 63-V rating would be better.

The PIV of the bridge rectifier should also be a minimum of 50 V, however, 100 V is better and doesnıt cost much more. Just donıt forget inrush current. Suppose the power supply provides 0.5 A. You choose a 1-A transformer. At powerup, the capacitor acts like a dead short, and the full power of the transformer goes through the diodes. This also occurs if there is a short circuit to the power supply. So, if you chose 1-A rectifiers for the 0.5-A supply, they could easily fail. Admittedly, diodes are rugged, but for a few pennies, safety and reliability can be enhanced.

I am always conservative when I design a power supply. Cutting corners here always seems to create problems later.

PREVIOUS • NEXT