opamp AC square wave to to DC

Go To Last Post
111 posts / 0 new

Pages

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Switch mode power supplies do not generally provide isolation. You have to use a topology, usually with a transformer, that does provide isolation. If there is significant capacitance from the circuit common to "earth", you will have major problems. For this kind of application, you really need to have it battery powered so that it is genuinely isolated.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Jim very sorry,

 

What i meant to say was that i will have an analogue switch , was an error say smpt.

 

Yes it is battery powered.

 

My switch will be SN74LVC1G66

 

Once again sorry for the error

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What is the function of that "switch" and what drives it? The output of an MCU  is just as good.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As its analogue switch it will isolate the probe connectors and the known resistance.  

 

This way only one of them will be getting the outputted square  that will be feed into the op amp.

 

But yes i agree that i could have two separate GPIO pins outputting a pulse 1. to Connector 2. resistor, both then feeding to the op amp.

 

It just accord to me that any GPIO can be used as i no longer using a PWM dedicated pin.

 

Thanks

 

 

 

 

 

 

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Again, if you don't have exactly (or very near so) 50% duty cycle, you will destroy your sensor.  Maybe adding some capacitive coupling in the drive would be helpful--that ensures no DC component (assuming good low leakage cap).

In direct drive methods, sometimes an integrator is used (as a feedback ) to force the DC average to zero.

When in the dark remember-the future looks brighter than ever.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Will keep it 50% duty cycle ,

how would it destroy the senor? I will be using steel probes. 

 

Yes I will be adding coupling cap on pins that is outputting the wave. Is there away to calculate best value for my applications?

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The sensor cannot tolerate DC, so an exact 50% (perfect balance) is required.  Also, if you add caps, then that wave will curve itself to enforce a zero  dc condition.  

When in the dark remember-the future looks brighter than ever.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes my signal is 50%. I have made sure with the timer/counter.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi 

 

I managed to get my a prototype working to a certain degree, it seems that if i was to use resistors to simulate the EC at the two contact points its works fine, but once i place its in some form of substance for examples water from the tap which i should produce a low EC about 0.4V but i am getting about 3.3V.

 

Am i missing something trivial or do i need condition the GPIO pins?

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Schematic, please!

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry, thanks for your reply.

 

Please see it below.

 

If i was to place a resistance for example 1M between the the probe terminals U9 and U10 i get an output of 0.45V which is very close the calculated output 0.46V for 1 EC, which i believe 1M resistance is . The reading from ADC is about 35-36.

 

I have placed the terminal in tap water as well as distilled water thinking there should be no conductivity, but my ADC is producing a 255.

 

Please see my schematic below.

 

I am outputting a square wave from net PWM_OUT , but its not a PWM signal but a GPIO pin as previous recommend . 

 

 

Thanks

Regards

DJ

Last Edited: Sun. Oct 29, 2017 - 10:11 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry, I do not comprehend net labels "CLK_OUT" and "PWM_OUT". What are they and how are they related? And, where in that circuit is the resistance being sensed? 

 

I am concerned about an op-amp powered from 33V connected to something with an absolute maximum input voltage on any pin (with possible exception of /RESET). Why not use a rail-rail op-amp powered from 5V?

 

Actually, that circuit is (sorry to say) totally crazy. If the CLK_OUT is 0-5V, the op-amp output will be (about) +30V to +25V with open circuit where ever the sense resistor is.. Your ADC (and maybe the micro, totally) will be fried. I am surprised that you get anything at all. BUT, if the CLK_OUT is the CLK_OUT of a micro running at more than 1MHz, the op-amp, what ever it is, may be so severely slew-rate limited that you see little or none of that signal on the op-amp output. Your circuit is in serious need of some knowledgeable design effort.

 

Thanks

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Last Edited: Sun. Oct 29, 2017 - 10:37 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

CLK_OUT and PWM_OUT are outputs from GPIO pin.

 

PWM_OUT is outputting directly to the probes or there terminal at the moment.

 

CLK_OUT outputs to known 200K, used for testing purpose . Simply there to measure a known resistance.

 

Sorry 33V is 3.3V, labeling error. 

 

The op amp is 3.3V, with a voltage divider the create a 1.65V to represent the + and -, if i am correct. 

 

MY CLK_OUT is 10Khz signal , which is being output from the GPIO pin. Has this got any effect.

 

As i am not using 33V as previously thought but 3.3V,do i need to change anything?

 

 

 

 

Thanks

Regards

DJ

Last Edited: Sun. Oct 29, 2017 - 11:50 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Jim

 

MY CLK_OUT is 10Khz signal , which is being output from the GPIO pin. Has this got any effect.

 

As i am not using 33V as previously thought but 3.3V,do i need to change anything?

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OK, both of those changes make a lot more sense. Now, where does the resistance you are trying to measure go? That is not obvious, yet. 

 

And, how are CLK_OUT and PWM_OUT are related. Even more to the point, why do you have two signals?

 

What happens to R12 during an actual measurement? E.g. Is it removed?

 

Your op-amp needs to be specified for 3.3V operation. Is it? What is the output swing under that condition. Or, better yet, what is the op-amp?

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What i am measuring is water with additional substances and the EC gives us an idea on the concentration level of it. So the aim is place probes into this water substance,and as the substance will have some resistance effect, which will be higher then tap water and distilled water. 

 

When you say where does the resistance go, i am not sure what you mean by this? 

 

A clock is generated in my AVR..

 

The PWM_OUT will be used to output directly to probe connected to U9 and the second probe is connected to U10. This is the signal which will be used in the actually application.

 

CLK_OUT is another GPIO pin that used 200K resistor, this GPIO is kept in high-Z during when the main application is active. The 200K is never removed, as its purpose is simply to check if the EC is being produced with known resistance. The idea is to simply check is the ADC value is correct. 

 

The op amp i am using is TLV341 and it for 3.3V .

 

 

At the moment if I place resistance between U9 and U10, i do get a response related to EC, but once water is introduced is preforming as the EC meter i intended for.

 

What could be wrong?

 

 

 

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The use of U9 and U10 for what might be connectors is highly confusing. Connectors are usually J or CNNR.
Convention for representing the decimal point is to use the the units letter -eg: 3V3 is 3.3V. 0R15 would be 0.15 Ohms.

Using 10kHz means that small capacitances may have a significant effect. The capacitance of your cable might be significant. Is the sample in an insulated vessel ie beaker? Mains pickup will also be significant methinks. Time to get the ‘scope out and measure.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry , will make the changes.

 

The PCB will be water insulated and connectors will have probe elements attached to it. For this test, i am embedding the part of the PCB with the connectors. 

 

Yes its a glass beaker like container.

 

Using the scope its does show the resistance in the water is very low, and square wave beginning outputted on connector pad "U9" is being received very close to the original on  connect pad "U10" .

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DJ - Many more questions:

 

1. How and where are you measuring signal? Is it at the op-amp output or somewhere else? Is a microcontroller ADC involved or are you measuring with a meter? Tell us a lot more about this "measurement". For example, how are you converting this varying amplitude signal into something that a micro or a meter can measure?

 

2. What do you measure at the op-amp output with CLK_OUT driving and with CLK_OUT floating (no probe connect in either case)?

 

3. Is it correct that ONE probe plugs into what YOU call U9 and U10? And the effective resistance being sensed is between U9 and U10?

 

4. What do you measure with the probe installed but out of water?

 

5. What do you measure with the probe in a beaker full of water or non-conductive equivalent?

 

6. What do you measure with the probe in a larger body of water?

 

This circuit is designed for the case where the probe resistance is 200K or larger. What is the specified resistance range for the probe you have?

 

Best wishes

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Jim

 

I might have a solution which I will discuss, but I will firstly answer your questions.

 

  1. Yes I am measuring the voltage that is being output from the opamp using my AVR ADC. Using a sample rate that is 4 times higher them my desired output frequency I preform the following logic:

 

Tick 1: GPIO LOW TO HIGH

TICK2: READ ADC –Save to variable adc_high

TICK3:HIGH TO LOW

TICK4:READ ADC  –Save to variable adc_low

TICK5: IS TICK 1

 

The difference between the two variables them become is my ADC reading.

 

  2.   Yes when no probes are connected or emerged in any substance, I use “CLK_OUT” to output a pulse that uses the 200K, to produce . The ADC produces 178 which is about the 2.31V.  But CLK_OUT was used for          testing purpose only and not used in actually application, even though the resistor is still there on PCB with CLK_OUT set to high-z.

 

3.    Yes that is correct I am measuring the resistance between U9 and U10 . The probes  are screw type terminal , which the PCB be will be eventually water protected the terminal contact will preforms as the probes.

 

4.  Eventually it will be compost based substance that will be added to water.

5.  Yes I have tried both water and distilled water which I believe is not conductive.

 

6. Yes I do .

 

 

Why do you say that the circuit is designed for probe with values of 200K+, is it because of the 200K resistor.

 

For now I have removed the 200K. As my probes are simple terminal in theory I do not think it will have much resistance  

 

What I was thinking of was the following, and was wondering if it makes any sense?

 

  • Replace the 1K R10 with maybe 150K or more . This insures there is always a fixed resistance to a certain degree.
  • Then measure the ADC when terminals U9 and U10 are shorted , this represents when there is no resistance.
  • Then measure water and distilled water.

 

What i do not understand i placed some de-ioneised water in a small container, embedded my terminal U9 and U10. Outputted a pulse on U9 and using scope monitored both U9 and U10, and it showed U10 was getting the conductive of the signal. Should this be happening?

Could there be a chance the current consumption is slow low, that the circuit is too sensitivity? 

Thanks

Regards

DJ

Last Edited: Mon. Oct 30, 2017 - 03:14 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

First, some terminology which I do not think you are using correctly. Conductance is the reciprocal of resistance. 10 Ohms resistance is the same as 0.1 Siemens conductance. 1 KOhm is same as 1 mS. You frequently seem to use the term conductance when you appear to mean resistance.

 

The net conductance depends on a number of factors. Classic physics. 

 

1. The specific resistance - Ohm-meter - or specific conductance Siemens/meter. This is the property of the material through which the current flows.

2. The length of the current path Resistance is proportional to the length and conductance is inversely proportional to the length

3. The cross-section of the current path. Conductance is proportional to the cross-section area while resistance is inversely proportional to the cross-section area. 

 

Measuring your sensor in deionized water in a non-conductive container should be about the most perfect measurement you can get. There is no current flow in or out from external sources (because the container is insulating). There is a minimum frequency that you need to pay attention to. If the frequency is too low, you get ionization at the electrodes. If this is happening, I suspect that the reading will change over time as the electrode surface changes.

 

If the resistance is too low (or the conductance is too high), the amplifier output will saturate and you will not get a valid reading. Try dropping a few gains of salt into the DI water. The resistance should drop very quickly. The largest reading you can possibly get, assuming Vref = Vcc = 5V and assuming that the op-amp output swings all the way to ground and to 3.3V (false, by some millivolts, in both cases)  will be 1023 * 3.3V/5V = 675. Actually, because your virtual ground (R13 & R14) is NOT at the center of your logic swing (e.g. NOT 2.5V), the op-amp output will be asymmetrical about the midpoint of the output swing range, and one phase will saturate before the other. You would be much better off powering the op-amp from 5V and to set the virtual ground to be 2.5V.

 

Best wishes

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok, as conductance is the reciprocal of resistance, should the probes have some resistance as when i measure my sensor in de-ionized water in a plastic container i am getting the clock pulse that i am outputting on U9 appearing on U10 as well?

 

Its seems that when i am measuring de-ionized water the resistance is low so placing salt is not making much difference.  I am sending a 10Khz signal.

 

What should be expecting when measuring in de-ionized water?

 

My problem is that my sensor is detecting the clock on U10 even with de-ionized water.

 

I am thinking my probes which are the contact pads, as low resistance could this be the issue? 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

If you are using the actual circuit board pads as the electrodes, you may be getting unrealistic values. Do you have an actual probe? That is what you should be testing with! Anything else will be little or no use.

 

Jim

 

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

I once worked for a company that made water purity sensors, we used stainless steel probes to reduce any possible corrosion issues, screw terminals as probes seems problematic at best!  There is also the issue of the pcb getting wet, contaminated, etc... 

IIRC, the circuit used an AC signal at the probes to also prevent corrosion issues too.

 

 

Jim

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am not using the PCB as electrodes. The U9 and U10 are screw terminal and stainless steel connections pad will be connected to these terminal. Eventually the PCB will have epoxy compound giving it the necessary protection.

 

I think I might have solved my issue. What I have down now is made the GPIO pin connected to U10 generate the CLK and I have removed the 200K resistor.

 

U9 is NOW going to be GND even though for this PCB, it will be connected to a GPIO pin forced low. So one probe is the CLK and the Other is GND.

 

I think this was the part with the  error. I have attached an update diagram.

 

 

 

This change has made my readings correlated to what should be happening in theory.

 

When the probe is not placed in anything I am getting a ADC reading of 179

 

When placed in Di-ionised water I am still getting 179

 

When placed in tap water  my readings reduce to 176. Which I believe reflects low EC.

 

When placed back into De-ionised the reading is back to 179, as I placed salt into the container this reading started to drop, still about 120. I guess I would need to further mix the water to get an increase in EC reading.

 

 

That being said,

 

I am getting an ADC reading of :

 

 

217 When Clock is high

38- When Clock is low

 

Giving a 179 when low is subtracted from the high, as suggested in the earlier post.  This reflect an ADC reading of 2.31V when not connected, why does this happen?

 

Thanks 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You should have HIGHER conductance (lower resistance) with tap water than with DI water. Your change makes little sense. Please, it appears that you are making random changes with little idea of what you are trying to do. 

 

When there is no sensor, you are saturating the output of the op-amp because the gain is 140X. Any reading is junk = crap = useless at this point.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Jim

 

The change was making one probe the CLK(also connect to OP amp) and the other a GND. Instead of keeping once CLK and the other the connection to Op-amp 

 

If i am correct this is how it should be ? previously it was one probe was the CLK and the other was the receiving end. 

 

Yes i am getting a higher conductance . As the conductance increases, the reading from the ADC are dropping. And as i add salt to DI water , the resistance are dropping, you see the readings from the adc correlates to that. 

 

I think tomorrow, i will need to do a bit more test, against the reading against salt increasing in the DI water. 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If one electrode connects to CLK_OUT and the other to ground which is really another GPIO out, the ONLY change you will get is due to the finite resistance of GPIO output pins. Each one has an effective series resistance of about 100 ohms (maybe 80 at 5V). Pardon me, but that is a terrible way to design a circuit. 

 

Lets start with your "sensor". What is the maximum and minimum expected resistance of this sensor? How do you know what the expected resistance is? Or, are you just guessing?

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree, in regards the GPIO pins, this was an error prior to making the PCB. Today i did not want to spend much time modding the PCB to get it correct, but from what you mentioned its might be worth cut the track between GPIO pin and connector and then connecting it  directly GND.

 

In regards to resistance not sure to be honest, but i know that based on my calculation it should be able to measure up to 7EC, due to the 140K resistor if i am correct. 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What is 7EC?

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry the electrical conductivity , the circuit is been set to work with a max which is 7. 

Thanks

Regards

DJ

Last Edited: Mon. Oct 30, 2017 - 11:43 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

djoshi wrote:
What should be expecting when measuring in de-ionized water?

 

Wikipedia wrote:

Purified water is suitable for many applications, including autoclaves, hand-pieces, laboratory testing, laser cutting, and automotive use.[4] Purification removes contaminants that may interfere with processes, or leave residues on evaporation. Although water is generally considered to be a good electrical conductor—for example domestic electrical systems are considered particularly hazardous to people if they may be in contact with wet surfaces—pure water is a poor conductor. The conductivity of sea-water is typically 5 S/m,[5] drinking water is typically in the range of 5-50 mS/m, while highly purified water can be as low as 5.5 μS/m (0.055 µS/cm), ratio of about 1,000,000:1,000:1.

 

https://en.wikipedia.org/wiki/Pu...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think that DJ as confused the specific conductance of water (seawater 5 S/m) with the actual conductance/resistance exhibited at the terminals of the sensor. I strongly doubt that any actual sensor would exhibit a conductance of 7S (7EC, as he calls it). Imagine a sensor with a path length of 1cm and a path cross section of 1cm**2 with water that maybe 50mS/m. If I do my computations right (and I might not, used to resistance rather than conductance computation), that is 50mS/m * 10e-4 m**2 / 10e-2m = 0.5mS. That is a LONG ways from the 7S suggested by DJ. 

 

Further, if that 7S were right, the resistance would be (1/7) ohms with which that circuit could not possibly work.

 

Things just do not add up, either with the electronic design or with the assumptions upon which the design is supposed to be based. They conflict with each other and with reality. Sorry, but you cannot call a black horse as white!

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
a path cross section of 1cm**2

There is a button in the toolbar to give you a proper superscript:

 

 

So you can have a proper 1 cm2

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Jim

 

Yes it is getting confusing, 

 

My aim is to measure the conductivity in water when mixed with other substances e.g. salt or fertilizers or chemicals. Can you point me the right direction in terms with schematic 

 

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

djoshi wrote:
Can you point me the right direction in terms with schematic 

Sounds like the schematic is not the Big Issue here?

 

Sounds like you really need to spend some time reading-up to get a solid understanding of the principles involved ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Analog Devices have an app note for a conductance meter using an aducm360 and a handful of Analog chips (of course). It uses 24bit adcs. Might be worth a read. They have a board as well.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

I have a question, maybe silly, but anyway...

Water is a good dielectric, therefore the electrodes immersed in water form a capacitor. So deionized water is a poor DC conductor, but it can still conduct AC. Since the OP is using a 10kHz wave, will it not be able to go through and result in measurable conductance?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks I will be reading the article, once i finish this post.

 

In simple terms if i am correct i am measuring current, depending on the substance of what i am measuring there will be a resistance . Therefore current will increase and drop based on this resistance .

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The circuit would have worked (with a few tweaks) provided a suitable probe is used. DJ seems not to have understood the relationship between the fluid conductivity, the probe geometry, and the probe's terminal resistance. For a given fluid conductivity, the resistance depends strongly on the size of the electrodes and the geometry of the electrodes. It is NOT a simple matter to predict. Normally, folks use a known geometry that results in a resistance in the range they can use. Though I have not read it, the Analog Devices note likely discusses all of this (they do a pretty good job of that in their technical material).

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Jim

 

I will be reading the documents as my it seems to have a lot of information. 

 

From brief reading i see that e probes must be 1cm3 and spaced apart 1cm as well.  But will read it in detail by tomorrow 

 

If i am correct the article is:

http://www.analog.com/en/analog-dialogue/articles/fully-automatic-self-calibrated-conductivity-measurement-system.html

 

My fluid conductivity will be the only varying element, as this fluid can various things.

 

Once you know the size and geometry of the electrodes , how does that impact the resistance what what can be measured? Does it set the max limit or minimum? How would i then tweak my current design?

 

I will read the article so i get a better understanding the theory behind it .

 

Thanks

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I did the computation above. 5mS/m medium between electrodes that are each 1cm2 and spaced 1cm apart gives you a terminal conductance of about 50mS = 20 ohms. YOU choose the electrode area, spacing, and geometry to get the terminal resistance that you want. Actually, 1cm2 is a bit large, that was just for a reference point. If you halve those to 0.5cm2, then you double the terminal resistance. If you double the separation, then you double the terminal resistance, again.   But, if your medium is 5S/m, then the terminal resistance is reduced to 1/1000 of the value used in the example.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

Last Edited: Tue. Oct 31, 2017 - 11:50 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I couldn’t see any mention of the pwm frequency - i dare say it is in the code. I am interested in the actual frequency - your use of 10kHz seemed a bit too high.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Don't have a good sense of frequency effects. I know that if it is too low, then chemical reactions begin to take place at the electrodes.

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
I couldn’t see any mention of the pwm frequency - i dare say it is in the code. I am interested in the actual frequency - your use of 10kHz seemed a bit too high.

 

No i am not using PWM, I am using 10 Khz, but this can be changed. I have been reading other articles i came across that 10Khz could be to low and something in range of 40Khz.

 

 

So lets say my probes are 0.5cmand are spaced about 1.5cm apart, this terminal resistance

ka7ehk wrote:

I did the computation above. 5mS/m medium between electrodes that are each 1cm2 and spaced 1cm apart gives you a terminal conductance of about 50mS = 20 ohms. YOU choose the electrode area, spacing, and geometry to get the terminal resistance that you want. Actually, 1cm2 is a bit large, that was just for a reference point. If you halve those to 0.5cm2, then you double the terminal resistance. If you double the separation, then you double the terminal resistance, again.   But, if your medium is 5S/m, then the terminal resistance is reduced to 1/1000 of the value used in the example.

 

Jim

 

Thanks will read the article tomorrow as its late at here now but once the terminal resistance is known, than how is that incorporated into a a reading of the substance e.g. water ?

 

Once i get soothing working, would there need to be any additional calibration?

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Once you know the terminal resistance with a material that has a known specific conductivity, it is simply a matter of linear proportion (inverse proportion, actually) to determine the resistance at any other conductivity. You can only know the terminal resistance with some known material (e.g. material with known EC). Terminal resistance is not some abstract quantity, it is a combination of material conductivity, electrode area, and electrode geometry. 

 

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Jim 

 

Ok Say for example electrodes are 1 cm apart and are 0.5 cm. If it was to be place in a substance with a certain conductivity and  make a note of the reading and then i can use that to determine any other conductivity and resistance.  If i am correct you can purchase conductivity substance as calibration reference ?

 

What must i do on my circuit as at moment as when the electrodes are merged in de-ionized water it seems like its being detected on my other probe contact, even though De-ionized water has very high resistance? 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Your first paragraph is correct, no matter what the electrode area and spacing is. Let me provide an example. Suppose that you have a calibration reference that is 0.5EC (that is, 0.5S/m) specific conductivity. And, suppose that you read 100 counts (your peak-peak ADC reading). Now, suppose that you put the same sensor into an unknown solution and read 50 counts. Then, you know that the EC of the unknown sample is 0.5EC * (50counts/100counts) = 0.25EC. Pretty simple, right?

 

I do not understand what you mean by the phrase 

when the electrodes are merged in de-ionized water it seems like its being detected on my other probe contact

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes that is simple part and easy to understand. In addition to that i might need to add a temperature compensation. 

 

I think there is something with my design, as the when i place my electrode into DI water, i notice that the CLK pulse on U9 appears on U10 when using the scope, as if there is no resistance.  Reading is 255 from the ADC.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Deionized water is not the same as distilled water.

Deionization does not remove organic material.

Depending on the original water source, deionized water may have a tendency to become acidic upon exposure to CO2 in the air.

You may want to check the pH of the water you are using...

David (aka frog_jr)

Pages