AC/DC (power conversion not the band) – buy it or build it?

At some point in the product chain, almost all electronics require conversion from an AC source to a DC one. Whether your product ships with an external adapter, battery charger, or if you design your own offline (AC/DC) converter, that originating source power most likely comes from AC mains.

In many cases, it makes the most sense for engineers to buy off the shelf AC/DC converter solutions. There are many good options from manufacturers like Artesyn, Lambda, Bel Power and SL Power. These solutions can range from PCB mount to rail mount to fully enclosed external supplies. For those engineers who want to design their own chip down solutions, manufacturers like Infineon, ST Micro, Maxim, Microchip, Intersil, Rohm, OnSemi and many others provide conversion IC’s as well as reference material to help guide you in your design. You also need to consider the power needed from your supply. Designing a 5W converter isn’t nearly as complex as higher power designs. So how do you decide if you want to tackle designing your own or if you want to buy an off the shelf solution? Let’s take a look at some pros and cons to each.

The Case for Build 

The most compelling reason to try and design your own offline converter nine times out of ten is that you can do it for a lower cost than buying a module. You can also plan for procurement challenges by designing in as many second source capable components as you can. In a previous role I designed equipment in the telecom space. In that market every penny counts. Additionally, we designed for industrial environments and had plenty of in house expertise in not only circuit design, but design for compliance (regulatory, CE, UL, etc) testing… ah yes, compliance testing. Compliance is a huge part of the design process when dealing with hazardous voltages (not only AC/DC, but high voltage DC/DC). More on that to come. You also may have unique requirements that necessitate special features that you can’t get off the shelf or maybe a strict form factor that requires custom design. Bottom line, there are very valid reasons to design your own, however in my experience at least strongly considering a buy mentality is worth investigating. If you are comfortable designing your own custom magnetics, specifying the proper safety rated capacitors, meeting minimum creepage and clearance requirements and insuring that hazardous voltages aren’t exposed to where the user could touch them then maybe designing your own is the way to go. Like anything else it’s hard until you’ve done it. If any of the things I’ve mentioned in this section are news to you, it might be better to focus on the next paragraph…

Case for Buy 

In my opinion, the compliance and certification requirements would be reason 1A with safety being 1B. I should probably make safety number 1, but I’ve shocked myself plenty of times and I remember the pain from going through compliance testing more vividly. There are several topologies to choose from (flyback, forward, half-bridge, full bridge), regardless of topology they will all have to meet a variety of compliance standards to be sellable domestically and abroad. Depending on your market segment you may have slightly different or additional requirements, but in almost all cases you will have to pass some sort of regulatory standards. These include at a minimum safety and EMC - conducted and radiated electromagnetic emissions. Again, talk to your compliance engineer or your customers to figure out what standards you must comply with and make sure you plan for it from the outset of the design. The ease and peace of mind of buying a pre-certified and tested offline converter makes for a compelling argument. You don’t have to care about topology or cost involved in R&D and certification time and testing. Sure you will pay a slight premium but in a lot of cases unless you have a very high volume your per unit cost may actually be cheaper when you factor in time and engineering dollars. Another potential benefit from buying a power entry module or external adapter is that it might allow you to shrink your product or reduce thermal relief requirements by removing that heat from your core product.

So Now What?

Ultimately each designer needs to do the analysis and balance the risk, long term cost, upfront cost, engineering resources available and more to really decide which direction to go. I just encourage anyone considering designing their own offline power system to make sure they research the requirements outside of just design calculations that you will face before you can get your product to market. Also take a look at what precompliance testing you are able to perform yourself. You could potentially be looking at a hefty price tag if you submit your product for testing and it fails, requiring a second round. Avnet has application engineers around the world that can help you with this analysis and also help you find the resources you need to make an informed decision.

Here are some links that might be helpful to explore options -

AC/DC modules - http://www.newark.com/c/power-line-protection/power-supplies/ac-dc-converters 

AC/DC converters (chip down design) - http://www.newark.com/c/semiconductors-ics/power-management-ics-pmic/ac-dc-converters 

PWM controllers - http://www.newark.com/c/semiconductors-ics/power-management-ics-pmic/pwm-controllers

PFC (power factor correction) controllers - http://www.newark.com/c/semiconductors-ics/power-management-ics-pmic/power-factor-correctors-pfc