ESR Meter (Kinsa) Questions

[Wbdsgnr]

Dear people/Kinsa,

Using Kinsa's ESR meter design, i've created my own PCB and re-wrote the software in C. I'm still stuck at calibration, since i can't reach the 0x4CF value. Anyways, i've got a couple of minor questions regarding the design (analog electronics is not my strongest point).

First off, the PWM signal. We're outputting an 100 KHz signal with a duty cycle of 90% (What i extrapolated from the JAL code). However, I thought ESR meters use small pulses? So should it be 10% by any chance?

Next question is the first stage push-pull amplifier. Why a push pull amplifier with a PWM signal? The PWM signal will never trigger the lower PNP transistor correct? I'm just curious cause i'm trying to understand the circuit so I can solve my problems

Finally, the trim / pot in the AC feedback loop of the common emitter amplifier has to go all the way down to zero for me to get the highest possible output reading. I zero the meter with an HEX value of about 0x1070 and with a 10 ohm resistor at input i'll max get a value of 0x3D0, so I guess my gain isn't as good as it could be. Should I adjust the emitter resistor?

Hope you can help!

[kinsa]

Hello, welcome to the forum.

First off, the PWM signal. We're outputting an 100 KHz signal with a duty cycle of 90% (What i extrapolated from the JAL code). However, I thought ESR meters use small pulses? So should it be 10% by any chance?

ESR meters are actually AC ohmmeters. The duty cycle of the source is not important, what is important is that it is an AC signal.

The 90% duty cycle was chosen due to the inverting amplifier at the input. The PIC actually measures the inverted 10%.

Next question is the first stage push-pull amplifier. Why a push pull amplifier with a PWM signal? The PWM signal will never trigger the lower PNP transistor correct? I'm just curious cause i'm trying to understand the circuit so I can solve my problems

Remember there is a capacitor at the output.

Finally, the trim / pot in the AC feedback loop of the common emitter amplifier has to go all the way down to zero for me to get the highest possible output reading. I zero the meter with an HEX value of about 0x1070 and with a 10 ohm resistor at input i'll max get a value of 0x3D0, so I guess my gain isn't as good as it could be. Should I adjust the emitter resistor?

I assumed you have tried different values. The adjustment procedure was an iterative process. You have to repeat the "zero" step after you have adjusted the potentiometer.

Note that the calibration using the trimmer resistor is only useful if you have used identical components. If not, then there will be issues with the amplifier gain especially if you have used a different type of transistors.

What the circuit does is basically a digital AC ohmmeter; you have an AC current source, an amplifier, a rectifier and a meter (the PIC adc).

All that needs to be done is correlate the resistance with the ADC readings. This step is performed by the "full calibration", where a set of lookup tables are made. Please take a look at the original analog version of the meter, this is where I got the idea from.

Feel free to ask questions if this is not clear.

[kinsa]

Next question is the first stage push-pull amplifier. Why a push pull amplifier with a PWM signal? The PWM signal will never trigger the lower PNP transistor correct? I'm just curious cause i'm trying to understand the circuit so I can solve my problems

This is a simplified explanation:

The lower PNP will turn on when the input is at 0v and if there is a positive potential on the output.

This happens when the capacitors are charged by the upper NPN during the positive cycle of the pwm signal.

[Wbdsgnr]

But shouldn't the base of the lower transistor have an higher potential then the emittor (which is connected to ground). I tried to simulate it with the falstad circuit simulator.. (cant post links yet)

Or is it that in real life the PWM signal will never reach GND exactly?

I did indeed use different transistors then the ones mentioned in your circuit... I guess i'll replace them with different ones and i'll do the full calibration method. Thank you for helping me out Kinsa, appreciate it!

[kinsa]

I used multisim.

A PNP transistor turns on when the base is 0.6v lower than the emitter  (the emitter is the one with the arrows).

[Wbdsgnr]

Hmm ok. It does work when I add the 4u7 Capacitor right after the amp... it converts the signal to AC... interesting This wasn't in the first design was it? Well ok, thank you I
 also gathered some 2n3904/6 transistors... Lets see if i'll get it to work this time.
When everything works I don't mind posting the code/circuit here as well. Might be a help to some people I did use some random phone LCD... And I had to translate the JAL 'tasks' into interrupts and flags.

[Wbdsgnr]

Ok, I replaced the transistors, and the output range works now. Now i've got a question regarding the full calibration:

I connected a shielded wire to the PCB for the test probes. Isn't it best to calibrate the unit using these wires? (So the added resistance is is basically cancelled). Or am I missing some basic AC vs DC difference here?

[kinsa]

There is no harm trying

I did it that way since it is more convenient. Also, the readings are affected by the test lead inductance (try making a coil of 1-2 turns).

If you plot the calibration data, you will notice that the slope is steep near 0 ohms. To improve the results a bit, a small resistor in series with the test leads is sometimes necessary. Note that this resistor must not be present during the calibration

[big t]

Hello to all,

Haven't check the topic "I Make by combining materials and parts of ESR METER - Help Plz" long time ago. Now I see that this topic is closed, I used "ESR" search and found this topic 

1. The question to Kinsa regarding schematic:

Kinsa, maybe you can specify/check C3 value in your schematic? Because it's marked as .1uF (point before value, I used 0.1uF for C3) or maybe it should be a 1uF value? Thanks.

2. Question to Wbdsgnr:

How are your ESR meter, is it working? If yes, please post your schematic and code. Thanks.

[kinsa]

0.1 uf

[pEN DoTA]

pa subs..... sa mga genius.