Power transistors are big bulky because they have to deal with a large amount of current in a very short amount of time. They are the fastest of all the transistors.
Heat is the biggest enemy of the transistor. As silicon is heavily dependent on temperature. As the die temperature increases the more sensitive it becomes. And at a certain point, it burns out. At the moment of burning there is a short which generates so much heat that the silicon just burns itself off.
Even the TIP3055 can only handle instantaneous current pulses for a very short amount of time. If that same transistor is put under a continuous load it will burn out in a few instances.
If we look at the derating curve of any power transistor. We will find out that they are only good for up to 1Amps of DC operation. And even at 1 Amp, it will need a sizeable amount of aluminium heatsink.
I recently bought a few Arduino Uno clones. They were not cheap. They utilise the ch34 IC, which is a USB to UART solution. They all have SMD atmega328pb soldered on them.
The problem is that two of them stopped working with the Arduino ide. On inspection, i found out that my software was not updated so firstly I updated the software which now includes a configuration for the atmega328pb.
But somehow the internal bootloader in these SMD chips got corrupted and it stopped working. I then probed the chip with AVRdude prograamer.
Found that the chip is atmega328pb. The configurations were not included at the time of installation. But the avrdude program allows you added other avr parts from Atmel.
I need to monitor battery voltage to check weather my charging system is working correctly or not. But to do that i have get up and walk with my multimeter towards the battery and i have to take these reading in night.
I placed my battery in a corner where there is very little light. So I added a transistor switch which can be controlled using Bluetooth to turn on the led light. Which provides enough light to act as a night light.
The ESP32 also has an ADC built into it. Which is very poor in terms of accuracy. It gets you the idea that there is something to work with but it does not give very precise reading like a multimeter.
Also, the ESP32 ADC is non-linear. The ADC also has an attenuation feature. Which by default is set to 11db. which gives us a workable range of up to 3.3V. But there are flat-out regions which need to be taken into account if you want to measure anything from this ADC. There is a 0 – 0.2V region in which the value read is constant, and there is a 2.9V to 3.3V region which also gives you constant reading values.
Resolution is 12-bit by default.
To measure a large voltage using this device. I made a voltage divider.
TIP3055 BJT is used as a low-side switch. R1 gives a base current of 330uA(=3.3/10000) which gets multiplied by the beta or hFE 70 of transistor to get a collector current of 0.023A or 23mA.
An LM358p is a dual opamp IC which I used in a single supply configuration. I drive the LM358 just shy of its absolute maximum voltage which is +30V. So I gave I supplied 28.8V (+-0.1V). It works flawless but the power supply is a Meanwell switching supply which introduces transient when plugging into the mains supply.
The transient arise when I was plugging in the AC wire to the mains. I used an anchor 6A 3-pin male plug for connection. It was at the brief moment when the transient arise.
The transient which arises apparently has enough power that it blew the epoxy off the silicon and it did with a loud bang.
Doing rough calculation at the moment this happened
Let’s say the transient is double the supply voltage of = 2 x 28.9 V= 57.8V
let’s say the current drawn by the IC is 40mA which is its absolute maximum.
power = voltage x current = 57.8V x 40mA = 2.312W
so for a brief moment, this 2W is radiated by the silicon chip. Which caused the heating up of the epoxy and generating enough expanding force which blew out like that.
but let’s take a more reasonable transient voltage(ISO 7637-2: 2011) of approximately 110V at a rise time of 0.5uS.
I have recently found out that there are FAKE transistors in the market. I have a built a circuit design and when I bought the transistor from the Lajpat rai market in Delhi; I assembled the circuit on a test bench for measurements. But I was surprised that the BD139 did not work with 10mA at 30V and they began overheating. They heat up so much that bubbles appear on the front epoxy and it cracked. I thought that maybe it was damaged so I replaced the whole circuit with a new transistor set. But they also didn’t hold and this time they burnt a resistor along with them. I saw flames engulfing the 1/4 Watt resistors.
So I again made the circuit using the minuscule BC546B. Which neither get warm nor went into thermal runaway.
on the BD139 it was marked that it was manufactured by NXP. And I then checked the NXP website and found out that the NXP closed its manufacturing of this transistor in 2001.
There are a lot of BD139 transistors circulating in the market. It is essential that you must buy from a good source.