Back to Basics: ADC and Impedance

Recently, in a discussion with a few folks, I found that ADC concepts were not that clear. Esp. with regard to the source impedance part. Some were under the impression that you connect a voltage within ADC range and measure, thats it. This might be a very basic thing, but I thought I should explain.

ADC Impedance

Source impedance is simply the open-circuit resistance seen looking back into whatever is driving the ADC input. An ideal source is zero ohms. A real sensor, voltage divider or source output has some finite output source resistance. That resistance, the ADC’s input resistance and capacitance form an RC low-pass filter. Check images for the equivalent circuit.

ADCs have a sample‐and‐hold capacitance (Approx tens of pF). When a voltage source is connected to ADC, this capacitor charges through the source resistance and an internal switching resistance (approx 1-2kΩ). This RC circuit needs a finite time to rise up. If you have a 2bit ADC for the last bit to settle, it would take approx 10x the RC time constant. To put it approx numbers, let’s say Rsource=1kΩ, Ron = 1kΩ, C= 10pF, RC time constant is 20ns, 10x that would be 0.20us. So you need to sample and take values only after this settling time for the LSB, else it will give you are wrong reading. This becomes a major factor if you are taking only 1 ADC reading of the sample and then going away. If that value is not settled, you get a wrong reading. Especially applicable when sampling fast changing AC signals. Now if Rsource is larger, say from a large MΩ voltage divider, your sampling time is going to 1000x this number. So always aim to keep source impedance less than 1k or so if you can afford the current.

What to do if you can’t? Then use a buffer so that your flow will be,
Sensor ->(High-Impedance) -> Op-amp (Unity gain buffer) -> (Low-Impedance) -> ADC input. 
This guarantees the ADC always see a low-impedance source because of the op-amp and your input is effectively shielded.

Please do take care of the source impedance case. There is lots to discuss regarding ADCs. If there is genuine interest, I can spend some time creating a series in the future. Do let me know.

If you liked the post, Share it with your friends!

Emergency or Inverter Bulb Working Principle

Some of you may have seen these bulbs in your home. At first glance, they seem like your regular LED bulbs, with a string of LEDs powered by AC input. But here’s the twist: they hide a rechargeable battery inside, often an 18650 lithium cell. When the power goes out, this battery springs into action, illuminating a subset of LEDs.

You can get a pretty good idea of what’s packed inside these LED bulbs. It typically includes a standard bridge rectifier and an LED driver for the regular lighting component. There will be a battery charge controller, ensuring the battery gets juiced up during operation. When the power goes out, the battery takes the wheel, supplying power to the LEDs. It’s a straightforward setup, nothing too complex

But here in lies the interesting part, how will the LED know to turn ON when there is no power? Because it can also be that its switch to it is turned OFF. How does it differentiate? It uses a clever mechanism, it can turn ON only when the input AC terminals Live and neutral provide a momentarily closed path. This cannot happen in the normal case and can occur only if another load in the same circuit is also turned ON. It could be any device. This will provide a path for the battery to turn ON power to LED till the actual AC power kicks in.

Consequently, you can perform a seemingly magical feat by merely touching the LEDs with your hands (a reasonable resistance between the AC input pins triggers the battery to power the circuit). It’s a neat party trick that can impress your guests when you effortlessly illuminate an ordinary-looking LED bulb with no external power source, using just your hands! ?

If you liked the post, Share it with your friends!
1 2 3 20