One of the things I need is an ESR meter, for checking capacitors. Unfortunately being a broke student means I can't easily afford one...
However, I have been thinking of the best way to make an ESR meter.
Why not use the multimeter most people already have? High accuracy on millivolts, high resolution and lowers the cost of the ESR meter as it can use the display, ADC, etc. built into the meter. Plus, some meters offer hold/relative settings, or data recording on a computer, which could be useful.
This idea doesn't require an oscilloscope, like some other test adapters. It gives an output voltage directly proportional to ESR on a multimeter display.
I have been working on a small circuit, which uses an LM339 and LM358 to measure the ESR of a capacitor, using entirely analog electronics. (No microcontrollers, even though it would probably be easier...
)
It works like so. It uses an op-amp + transistor as a constant current source to provide a 100mA pulse current to a cap for about 5µs, every 20-50ms. For caps more than 1uF this causes negligible charging, but it does cause a voltage drop proportional to ESR. With zero ESR, the voltage drop is 0V. With 1 ohm, it is 0.1V. And so on. The output to the meter is a fixed 1V/ohm (gain 10x.)
It's reasonably accurate. The testing method works best around about 200 milliohms and has slight non-linearity, tending to give lower than actual ESR for >1 ohm and higher than actual ESR for <0.2 ohm, however over the full range of 20 milliohms to 4 ohms, it has better than ±3% tolerance, which is probably fine for almost all applications, especially as a good/bad tester. It can measure higher than 4 ohms but has very poor accuracy. This is simply due to the choice of the capacitor discharge resistor (100 ohms), which makes the design simpler by avoiding a dedicated discharge circuit. (The discharge circuit is what gets rid of the charge built up from testing the cap after a few cycles.)
The rest of the circuitry consists of a differential peak hold circuit (for capturing the 5µs spike), a negative voltage generator (for the output amp), an oscillator (used both for the negative voltage generator, and times the output of the pulses), and a low battery comparator (optional.)
I'll post schematics soon, under an open source license. I'm considering making it my college electronics project, either that or a digital audio infrared link...
I have plenty of bad caps to test with. A whole bag full! And many others scattered on my workbench floor.
However, I have been thinking of the best way to make an ESR meter.
Why not use the multimeter most people already have? High accuracy on millivolts, high resolution and lowers the cost of the ESR meter as it can use the display, ADC, etc. built into the meter. Plus, some meters offer hold/relative settings, or data recording on a computer, which could be useful.
This idea doesn't require an oscilloscope, like some other test adapters. It gives an output voltage directly proportional to ESR on a multimeter display.
I have been working on a small circuit, which uses an LM339 and LM358 to measure the ESR of a capacitor, using entirely analog electronics. (No microcontrollers, even though it would probably be easier...
)It works like so. It uses an op-amp + transistor as a constant current source to provide a 100mA pulse current to a cap for about 5µs, every 20-50ms. For caps more than 1uF this causes negligible charging, but it does cause a voltage drop proportional to ESR. With zero ESR, the voltage drop is 0V. With 1 ohm, it is 0.1V. And so on. The output to the meter is a fixed 1V/ohm (gain 10x.)
It's reasonably accurate. The testing method works best around about 200 milliohms and has slight non-linearity, tending to give lower than actual ESR for >1 ohm and higher than actual ESR for <0.2 ohm, however over the full range of 20 milliohms to 4 ohms, it has better than ±3% tolerance, which is probably fine for almost all applications, especially as a good/bad tester. It can measure higher than 4 ohms but has very poor accuracy. This is simply due to the choice of the capacitor discharge resistor (100 ohms), which makes the design simpler by avoiding a dedicated discharge circuit. (The discharge circuit is what gets rid of the charge built up from testing the cap after a few cycles.)
The rest of the circuitry consists of a differential peak hold circuit (for capturing the 5µs spike), a negative voltage generator (for the output amp), an oscillator (used both for the negative voltage generator, and times the output of the pulses), and a low battery comparator (optional.)
I'll post schematics soon, under an open source license. I'm considering making it my college electronics project, either that or a digital audio infrared link...
I have plenty of bad caps to test with. A whole bag full! And many others scattered on my workbench floor.
Comment