what's best to read ESR with a microchip ADC

[dr pepper]

Mundo technology's website shows you how to make a sine with a pic using a r/2r ladder.
Another way would be to use a dds, they are cheap enough and you can specify any freq up to mhz with somethign like a ad9850.

 

[Ian Rogers]

This is coming up on Google now so the heat is on to make one with a pic lol

Ooh! The excitement!!!
I wonder how many members out there are just saying " Why not just buy the one from Bangood"....

I want to see you do it.... Can't be that hard!!!

 

[Ian Rogers]

Mundo technology's website shows you how to make a sine with a pic using a r/2r ladder.
Another way would be to use a dds, they are cheap enough and you can specify any freq up to mhz with something like a ad9850.

The guy in Nigel's post has good results with a square wave.... Pics are good at that..

 

[be80be]

I don't think it cares about what the wave is sine or square. I just want it to work good. Yep could buy one but I can make one too its just getting the math right from the pic. I think I got that part ok
Now I think I got the wiring figured out.
Started a board last night after work stayed up till 4 am playing with the output side. To get a good 100khz output
Tonight I'm hooking up the LCD and the rest of it got my stuff from tanyda in.
So I'm set to go.

 

[MrAl]

Hi,

The Arduino one looks interesting, but without making me read the whole site (ha ha) what is the basic principle of operation? That tells us right away if it works good or needs more good work. There are formulas spewed about so it's hard to tell, and the formulas that are there dont have any theory to go with them, as to why they should exist in that particular form. That, and i see a lot of variation in result accuracy being posted.
Is it just that they are trying to use a high enough frequency such that the capacitance doesnt react that much while the ESR does?

 

[be80be]

Hi MrAi
I'm with you I read the whole thing the last 3 pages some one else kind of took over and added too it but there's no support of there changes without going every where to find the bits a pieces.

 

[Nigel Goodwin]

If you like I can post the original opamp ESR project that I built years ago?, from the long defunct TV Magazine - I OCR'd it all and recreated it as a PDF. It explains how it works quite nicely.

 

[MrAl]

Hi,

Oh yes that sounds interesting too.

 

[Nigel Goodwin]

Hi,

Oh yes that sounds interesting too.

I'll have to do it tonight, I don't seem to have the file here at work?.

 

[dr pepper]

Whats tanyda Burt?

 

[be80be]

I spelled it wrong taydaelectronics.com there awesome parts suppler. I needed two
SCHOTTKY DIODEs . So I order about 300 part's resistors transistors cap's crystals and stuff like that seven days and it's here for 2.89 shipping and $18 for 300 parts
I don't no how the fronts changes on here.

 

[be80be]

I got the board made almost I tested my code for Lcd

The Lcd works now to add some more part's for the tester the thing on the right is a home made pickit2 clone. Glade i made some of them I lost my real pickit2 when I moved to KY.

 

[be80be]

If you like I can post the original opamp ESR project that I built years ago?, from the long defunct TV Magazine - I OCR'd it all and recreated it as a PDF. It explains how it works quite nicely.

Thank's that would be great

 

[Mosaic]

I used pulsed DC to calc. the ESR of a battery with a PIC. Max resolution down to under a milliΩ. Took around 100 x V and 8 x I samples to do it. I used a differential OPA to feed the ADC to be able use high impedance voltage dividers and ferrite beads plus a 10nF cap for noise handling. 8mS settling time (for RLC issues) before ADC sampling measurement. Battery V could be as high as 20V. Currents used are around 16A.
Battery has a steady 'load' of about 2.4K at all times to stabilize open circuit reference voltages.

Sampling:
Voltage divider (4:1) was built from 4 x 133k resistors fed via a type R ferrite bead and de-noised with a 10nf cap at the 1/4v point.
LM324 slow slew OPA (less noise sensitive) did the differential work. Two OPAs did unity buffering into the diff OPA. A 1K calibration pot was used on the Diff gain along with 1%, 10K resistors. All OPA forced into Class A with load resistors, also the OPA is supplied by 12V which is more than double the max output voltage of 5V to avoid crossover distortions.
Result was stable 4 sig. figure voltages and 3 sig. figures currents calibrated against a new Fluke 87V.

ESR results were compared against a dual trace current/voltage trace much like the o'scope ESR example I posted earlier. Practical accuracy/repeatability is good to about 2 milli Ω. The display shows single decimal milli Ω, but while it looks impressive the decimal value has little real value.

I'd post the schematic but its part of a much larger circuit and it's a bit cluttered and difficult to read. I'll have to redraw it .

EDIT: I imagine a genuine multi$, micro offset, nano input current, rail to rail diff OPA would make for true decimal milli Ω accuracy.

BTW the nice thing about a µC ESR meter is that it can produce a reference 25°C ESR result when measuring hot in service parts or need to verify temp. range ESR as electrolytics bounce around because of temps. Once you include a thermistor for temps.

 

[Willen]

Heres the esr tester I made years ago and have repaired many a thing with:

https://ludens.cl/Electron/esr/esr.html

Hi, I made other more simpler ESR meter (look at attached schematic) from here **broken link removed**

Amazing! It is working well. I had to use 33 ohms emitter resistor instead of 100 to get full scale reading because my cheap meter was 500uA, instead of 50uA. Probably you also felt hard to test very high value capacitors like 4700uF etc. Lets say- their well condition ESR is approx 0.002 ohms and very poor (damaged) condition ESR is approx 0.8 ohms. But with such analog meter, it's hard (not possible) to measure 0.002 and 0.5 ohms difference. Both shows full scale on meter and both seems good quality capacitor. How did you repair or decided in the condition?

 

[Mosaic]

To distinguish such detail you'd prob have to add a range switch to shift the DC bias point of the base and permit greater A.C. amplification via the neg. feedback emitter resistor. Perhaps with a 10uF cap across that emitter resistor.

 

[be80be]

Here how Im hooking it up

DC is discharge pin

 

[dr pepper]

Hi Willen,
Theres little diffrence between the ludens schematic and the one you posted, in fact the ludens one has less parts, the osc and the amp are in one 8 pin chip, not having a transistor theres little in the way of biasing resistors.
The scale on mine starts at 30 ohm, going to 2 ohms 1/2 way, and the last quater way is 1 ohm, so you can read down to a fraction of an ohm, the only thing with a log scale is that you couldnt resolve 10 ohm from 10.5 ohm, however its more than acceptable for repair work, I dont read resistance the diffrence between a good and bad cap is usually large, tens of ohms.

 

[be80be]

I'm going to use a tl431 to set my adc vref to 2.50 volts I was thinking maybe just use two diodes that would get me closer to 1.40 volts I thought I I had a 1.25. But are there any ideas on how to get me close to 1.1 volts the adc doesn't need much to work I figuring on setting it to few mA what would be bad about using the tl431 and voltage divider I can't change chips the 18f2550 is soldered in lol

 

[Nigel Goodwin]

Thank's that would be great

OK, here it is: