Designing for high-volume manufacture

When designing for high voltage on FR4 it’s important to use PCB design rules to ensure adequte clearance. FR4 when new and clean typically has a dielectric rating in the range of 800-900V per mil (that’s thou!), but due to ageine effects, a 300v limit is  more realsitic. With a typcial PCB stackup, that’d mean about 2.1kV for a sinlge later of 7 mil pre-preg – which is pretty astonishingly high for good ol’ FR4! Of course, thats layer-to-layer, and not track separation…

Better common-mode rejection

Common mode rejection is best obtained using a differential signal into common (matched) impedance. But what if the impedance source is inherently (and uncontrollably, at least by us) mis-matched? So we have to resort to thought/simulation and conisder the voltages generated (and dropped) by the various impedances in the circuits. If the impedances are poorly defined, we can’t hope to achieve very high rejetion ratios, but worthwhile rejection can still be achieved.

 

Low noise, high sample rate

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vestibulum nibh sem, convallis ac nisi at, placerat pretium lorem. In vitae tincidunt lectus. Etiam orci augue, venenatis eu vulputate eget, interdum in felis.

adipiscing elit. Vestibulum nibh sem, convallis ac nisi at, placerat pretium lorem. In vitae tincidunt lectus. Etiam orci augue, venenatis eu vulputate eget, interdum in felis.

Lead-free PCBs

What is a ‘lead free’ PCB? OK, it’s one that’s got no lead in it – that’s pretty obvious – but lead-free reflow process is typically run at higher temperature that for a leaded process. So can you run leaded solder on a lead free PCB? Sure – but sometimes there are packages with solder attached already – e.g. Ball Grid Arrays – if those have lead-free balls, and you try to mix with leaded process, then there’s a danger that the solder balls on the BGA won’t metl completely resulting in poor connections. So you can sometimes run leaded solder on lead free, but caution advised. When we do in-house prototype runs through our own reflow oven, we typcially used leaded solder as it’s a bit more tolerant to temperature variations – with lead-free, often have to push the temerature to avoid cold spots, and the variations in the hot spots can cuase discolouration (or in exreme casers, warping, delamination and even charring!) – so for a ‘on the lab bench’ process, running leaded is usually the best way.

Low noise amplifier success

Building the lowest noise amplfier circuits can be challenging. For years the good old 5534 (and 5532) were the go-to amps. With their combination of voltage and current noise, they’re right in the sweet-spot for the kinds of impedances typcially used in audio circuits. Sure there are amps with better voltage noise and amps with better current noise, but for the money, the venerable 5534 still performs quite well. Newer amps can best it though, some of our favourites that we now keepin include  ADA4898 (an astonishingly low noise amp), OPA1612 – another fantastic audio amp – the good ‘old’ LM4562 (though not as old as the 5532, still been around a while), and OPA2350 for single rail work.