5V Insulation tester [detecting if material started to conduct]

[oduleex]

Hello good, smart people.

I am trying to make 5V insulation tester.

I have thought I might use op-amp to do so in this configuration: https://tinyurl.com/292qspxs

From simulation it looks that at around 62M Ohm of resitance OpAmp with change it's state.

I am looking for an OpAmp that will trigger on higher resistance. Like 100M+ (has to run on 5V or 3.3V, Arduino ) or anything else.

I only need to know that material started to conduct electricity (using only low voltages to detect that). I don't need it to be triggered at any exact high value (but nice if it will trigger at highest value possible - faster reaction) 

What I need is just an I/O information that insulation has weakened, or other way that CURRENT started to flow - but TINIEST amount of it.

What do You think about that? And suggestions?

As I understand correclty I should look for lowest input bias current opamp. Am I right?

61M resistor is DUT (Device Under Test) - insulation material I am checking.

Thank you in advance.
Oduleex

Notes:

IQ vs Accuracy Tradeoff In Designing Resistor Divider Input To A Voltage Supervisor - https://www.ti.com/lit/an/slva450b/slva450b.pdf
https://www.electronics-tutorials.ws/opamp/op-amp-comparator.html

[Zero999]

That circuit won't work in real life because it relies on the input impedance and bias currents of the amplifier model, which are uncontrolled and temperature dependant in real op-amps.

What resistance to you want it to trigger on? 5V across 62M is 80.6nA, so it's certainly doable with a FET input op-amp.

[oduleex]

Hello Zero999. Thank You for Your response.

I want to check if my insulation is degrading. That's my goal.
My insulation is  100M - 1G [or more] Ohm range.

50M-100M - super low value I want to detect

If I trigger on higher Ohms value it's good as well. Only thing is I don't want to trigger randomly. I do not exactly care about exact value. I am just looking for an information that something wrong is going on with insulation.

I have 5V/3V3 to use. It would be easier to generate high Voltage but that will might cause lot of noise which I want to avoid. Or maybe there are some modules already designed for that?

From Ohms law:
 50V over 1G = 50 nA
 500V over 1G = 500 nA

Edit: I have just found this project https://www.eevblog.com/forum/projects/high-voltage-leakage-breakdown-tester-project/ - checking videos

[Zero999]

Insulation testers use higher voltages than 5V, not just because it results in a higher current, which is easier to measure, but also because it puts the insulation under more stress and the test is more often than not done as a safety measure.

It's trivial to make a circuit which triggers when a resistance falls below a specified value. Use a comparator with two potential dividers, with one of the resistors in one of the potential dividers being the resistance under test.

A FET input comparator such as the LMC6762 is mandatory. Don't use the LM393. It won't work.

I've just remembered a 100nF capacitor should be connected across the 5V supply and you probably also want 1nF across R4, to stop it from triggering from mains hum.

[jonpaul]

insulation R is non linear.

All "insulation testers" have a Vout 50/100/500/1000V from inverter.

You are NOT making an "insulation tester" but a megohmmeter.

Jon

[oduleex]

@jonpaul / @Zero999 Thank You both.

Yes, I am trying to build megohmmeter - I am just interested in LOGIC information if my medium started to conduct (resitance lower than  something).

Do you think I can easy "scale it up" to detect even higher resitances? Like 1G/10G ?

I will design this circuit in simulator and make same tests.

[Zero999]

@jonpaul / @Zero999 Thank You both.

Yes, I am trying to build megohmmeter - I am just interested in LOGIC information if my medium started to conduct (resitance lower than  something).

Do you think I can easy "scale it up" to detect even higher resitances? Like 1G/10G ?

There is a limit. At some point the input impedance of the comparator will dominate.

I will design this circuit in simulator and make same tests.

You need to be aware of the limitations of simulators and how to set various model parameters, otherwise it's easy to design something which works in simulation, but not in real life. For example in the LTSpice file I attached, I set the input impedance of U1 to 10T to model a high input impedance MOSFET input device. The circuit wouldn't have worked with the default 500Meg impedance. No doubt, it would have been possible to get it to work by tweaking the resistor values, but it would then fail if I build it using a high impedance comparator.

This is not a good project for a beginner.

[oduleex]

I just want to check if there is water in my garden .... Jesus...

Edit: There is also one more usage for this solution, as I have mentioned I just want to know if my insulator starts to conduct.

But with discussion happening I am learning and others reading as well! Thank You for that.

[jonpaul]

Water..garden:

Use finger...wet or dry

Easily avail water sensors, just search

Jon

[tooki]

I just want to check if there is water in my garden .... Jesus...

You chose to use the phrase “insulation tester”, which is an existing term for a specific type of test equipment. You chose not to give any more information about your actual application until deep into the thread. You don’t get to be annoyed at people because you’re not getting the reply you wanted: you got the correct replies for what you actually asked.

[Zero999]

I just want to check if there is water in my garden .... Jesus...

Edit: There is also one more usage for this solution, as I have mentioned I just want to know if my insulator starts to conduct.

But with discussion happening I am learning and others reading as well! Thank You for that.

50M to 10G is far too high for a moisture detector.

Ideally you want to use AC for the test current,rather than DC, for a moisture detector to avoid corrosion.

[oduleex]

Don't get me wrong, sorry for that. I didn't mean to be rude I just thaught It's going to be easier to implement.

Water testing is just one of application I want to use this device for.

I also have material that can absorb moisture, once there is enough moisture it will start to conduct. I can than turn on dryer which will get rid of that moisture [perfectly fine]. Thing Is I want to be informed that we have to much moisture in our insulating material [turn on dryer].

I know how to use Arduinio and I can controll dryer ON/OFF. This works just fine. I just want to build something that will tell me that our insulator starts to conduct - It can be triggered at 50-100M / 200M  or even 1G Ohm (the higher the better but I am aware it's going to be harder to implement).

Water detection in soil - I have found materials online, also about AC to avoid corrosion. Thank You for pointing that!

Sorry, if You got me wrong, I didn't mean to be rude or not nice. I Just wanted to show my emotions and that I was amazed that it's so hard to make it.


Note to myself: https://www.ti.com/lit/an/slva450b/slva450b.pdf [IQ vs Accuracy Tradeoff In Designing Resistor Divider Input To A Voltage Supervisor]

[Zero999]

If you measure the resistance of damp soil with a multimeter, you'll find it has a much lower resistance than the numbers you give, probably under 1M. It depends on the dissolved salts.

[oduleex]

If you measure the resistance of damp soil with a multimeter, you'll find it has a much lower resistance than the numbers you give, probably under 1M. It depends on the dissolved salts.

Yes, that's correct. I have already received parts (simple water prescene testers) and I will be having a play with them. One is op-amp based as I can see, so I will play with it (thankfully I have accesss to an oscilloscope and not bad DMM).

Check soil conductivity is one thing but as I have already tried to mention  I will be testing it with other situation I was trying to write about.

Fingers crossed! Thank You again.