Calibration method for ESR meter

[lpkf2018]

Hello, I am very interested in your ESR meter. I plan to make one myself, but I found that there is no calibration method for the meter. Could you please publish a brief document on ESR meter calibration? I really need it. thank you!

[latchdevel]

Hi, I'm glad you're interested.
I'm sure you won't regret making one, it works great.

You have to keep in mind that there are two stages, first you need to set the "hardware-based parameters" after assembly. This is extremely simple, only two resistors and a reference capacitor are needed.

To obtain the ESR parameters, pressing the "minus button" shows the RAW values ​​of the ADC, both for the 1 ohm range and for the 10 ohm range.

• First, open the probes, both readings should saturate the ADC and show 1023 and 1023. If not, there is a problem.
• Second, short the probes, a low value will be displayed for the 1 ohm range, for example 29, and zero for the 10 ohm range, and without releasing the minus button, press the menu button to set this ADC value as zero ERS.
• Third, place a 1 Ohm reference resistor as DUT, and note the value for 1 Ohm range, for example 570. This is the approximate value that will be set as the value of M_ESR_1 in the code.
• Fourth, place a 10 Ohm reference resistor as the DUT, and note the value for the 10 Ohm range, for example 890. This value divided by 10 will be set as the value of M_ESR_10 in the code.

To obtain the Cx parameters, pressing the "plus button" shows the RAW value of the Timer1 tick counter.

• Place the reference capacitor, for example 100 uF, and note the counter value. This value divided by the capacitor capacitance in uF is what should be set as the value of M_Cx in the code.
• Cx fine tuning is done by varying the value of add_Cx in the code.

You can then perform a fine calibration adjustment which is stored in the EEPROM and can be modified from the user menu, accessed by holding the menu button at power up.

And the ESR zeroing is always available and easily accessible, since it depends a lot on the connected probes.

But you are right, it is necessary to publish a complete document on this topic.
I have just made a small update to the circuit, it may be the right time to publish the rest of the documentation.

I hope to be able to do it in the next few weeks.
Regards.

[mockendon]

Many thanks for updating this project and the excellent PCB design. Please may I ask about troubleshooting in relation to calibration. I have calibrated with the 1R, 10R and known Capacitor using the calibration menu on the meter (no values changed in the code). The test points on the PCB all measure correctly (within 0.1V). Capacitance measures reliably but I always get an ESR of 0.000. Is it mandatory to set the values in the code or have I made some mistake in my assembly? My scope shows 40mV ptp pulses of 40KHz on the meter outputs.

[latchdevel]

Hi @mockendon,

I'm glad you decided to try it. You won't regret it; it works very well and is very easy to assemble.

First of all, just a couple of initial considerations.

• The 'REF to GND jumper' on the bottom layer must be soldered.

• The test probe connections must be in 'operational conditions', meaning there must be continuity between P1 and P2, as well as between P3 and P4. This setup is referred to as the 'OPEN' state.

Setting the values ​​in the code is mandatory, as it is necessary to adjust the instrument to the actual values ​​of the components used, and this adjustment is done by software.
But don't worry, there are only three important parameters, and they are very easy to calculate.

• M_ESR_1
• M_ESR_10
• M_Cx

All the components are common and easy to obtain; just be careful with two of them:

• The AD620 instrumentation op amp, which must be "original"; there are many copies that don't work well. INA129 or even INA128 could be used, although I haven't tested them.

• The 4066 must be selected carefully, and above all, the resistance in "on" of the third switch must be measured when operating at 5 V, pins 8 and 9. This resistance must be added to R2 to obtain a total as close to 167 Ω as possible.

Here are some that I have tested:

  - TI CD4066B Original    143 Ω -> Very good and easy to obtain, but make sure it's "original".
  - Philips HEF4066B        87 Ω -> Very good, but hard to get.
  - ST HCF4066BE           160 Ω -> Good.
  - TI CD4066B Aliexpress  204 Ω -> Bad.
  - Toshiba 74HC4066A       55 Ω -> Not usable because they have a certain resistance in "off" that alters the range of 10 Ω.
  - Philips 74HC4066N       30 Ω -> Not usable because they have a certain resistance in "off" that alters the range of 10 Ω.

Setting the M_ESR_1, M_ESR_10, and M_Cx values ​​is part of the instrument's construction. They must be fine-tuned until precise values ​​are obtained for their corresponding references. They are directly related to the components used, so once set, they should not be changed.

In normal use, the "ESR set zero" must be updated for each probe used, and the "calibration menu" is always available to correct any deviations.

I hope this helps.
Let me know how you're doing, and if you need anything, I'd be happy to help

Best regards.

[mockendon]

Fantastic, thank you so much for this extra detail - I can see you are hard at work on the manual! The 'ref' jumper was the key, I now have plausible M_ESR_1, M_ESR_10 & M_Cx values and the ESR values are also realistic. I have a working ESR meter - this is a significant contribution to the Open Hardware and Software community. You deserve my thanks and that of many yet to come for your excellent work here.