I need a sensor for soluble oil % concentration. Will run to an LED display

[fastline]

Long story short here, we own a CNC machine shop. They all run soluble oil as cutting fluid. This is a special oil designed to go into water. We have to control the concentration pretty accurately for proper performance. We also need to watch the PH. I want to make a meter that can show our concentration as well as PH. Problem is each coolant may have a different refractive index or SG which can affect how it is tested. I am curious if anyone knows what type of sensor we might be looking for? i have do idea how concentration is even tested. I assume PH would be straight forward. These would read back digitally to diplay real time.

 

[JimB]

The only technique I know for measuring oil/water concentrations is a capacitive probe, this can be extended to measuring water in gas using resonance methods. I digress.

The relative permitivities of oil and water are very different.
Water has a permitivity of about 75, but oil (at least the hydrocarbons which come out of the ground) has a permitivity of about 2.

Now, I do not know what is the permitivity of the soluble oil used in machining, is it a mineral oil or a vegetable oil?
You could maybe do some tests for your self with a home made capacitive probe.

JimB

 

[fastline]

These oils are typically mineral based. Do you know what type of component we would be looking for if we were going to build the sensor? In my experience, buying ready made sensors is ridiculously expensive.

 

[Sceadwian]

Have you tried contacting the maker of the oil directly, the maker of the oil should gladly tell you how to go about testing for concentration, the main problem being there may be no way to do it with something like a simple sensor the chemistry is more complex than that, you may have to do a titration for it, depends on how frequently you need to check it, there are automated titration machines out there, but they cost an arm and a leg for the unit let alone the reagent.

By the way, in case you don't know titration means you would have to take a sample of a fixed volume and apply various other reagent chemicals to it for various lengths of time generally one or two reagents are used with a lot of agitation and sometimes heat, generally the results are shown color metrically (the finally color of the solution determines the concentration)

 

[Sceadwian]

Jim, how is that going to show him how to determine soluble oil? Such a sensor wouldn't care if the oil/water was emulsified or not.

 

[fastline]

I guess my thought was since the permittivity is so different, testing it could be somewhat accurate for concentration. I am still researching but there probably is a simple way to do this. Just have not found it yet.

 

[JimB]

These oils are typically mineral based. Do you know what type of component we would be looking for if we were going to build the sensor? In my experience, buying ready made sensors is ridiculously expensive.

Probably the simplest thing would be two parallel rods 5 to 10mm apart.
It may be an advantage to have the rods insulated to eliminate the possibility of corrosion.
The rods form a capacitor which is connected to some circuit which measures capacitance.

I just did some quick calculations and rods 100mm long, 2mm diameter and spaced 3mm, will have a capacitance of about 3pF in air.
Not much is it!
However, in water the capacitance will be about 220pF, and in oil about 6pF.

So from 100% water to 100% oil there will be a change of 214pF which is quite easy to measure.
In an experimental setup, I would measure this by using the capacitive probe as part of the tuned circuit in an RF oscillator, so the change in capacitance changes the frequency.

Jim, how is that going to show him how to determine soluble oil? Such a sensor wouldn't care if the oil/water was emulsified or not.

The oil IS soluble, that is the whole point, it is mixed with water as a coolant and lubricant in machining.

JimB

 

[Sceadwian]

Yes Jim but I think you're missing the point, the sensor isn't going to know the difference between globs of oil near water and actual emulsified oil/water. It will be able to do jack point diddly to detect the amount of oil that is still actually emulsified unless it can also detect the emulsifier concentration itself as well as every possible component that could tie that emulsifier up. If the flow is turbulent it will read 50% water and 50% oil even if only 1% of the oil is emulsified. He has a VERY specific need of detecting soluble oil, not bulk oil/water ratio. The only way I know of detecting the actual solubility percentage is by titration. I've used this stuff before in a production environment, I can tell after handling the solution even briefly if its 'good' or 'bad'. If this weren't the case then they would just simply mix water with oil, they don't cause it doesn't work.

 

[fastline]

Sceadwian, I am not sure I understand your post. Obviously the solution we are testing is 100% soluble oil in water. There is always tramp oils from other lubrication points of the machines that gets in the coolant but it will always stay at the top and we would not be testing at the surface for that reason. I guess in my mind, Jim's methods seem they might work for what we are doing. We would only need about +/- .25% accuracy or maybe more. It is very hard for people to just "handle" the fluid and know if it is 9% or 10% solution. The oil product goes into water very easy and stays in solution. The typical measurement method is refract but I have a feeling the inline methods would be a huge overkill for this. Keep in mind some operators do NOT care enough to watch coolant so giving them a meter to watch is a safe alt.

 

[cobra1]

hi fastline, the oil you are refering to is this white water???

i used to work in a machine shop and we never had any problems with the concentrations that the manufacturer specified, on occasion the white water went brown, at which point it would just be drain off and replaced.

if there is a reason for your needing a probe then i would go with the capacitive method detailed above.

 

[JimB]

My interest has been awakend by this thread and I thought that I would try to measure the change in capacitance caused by various fluids.

I made a probe which would lay in a shallow plastic tray.
The probe consists of two enamelled copper wires, 2mm daimeter and about 240mm long, bent into a Z shape.
The wires are supported by two paxolin blocks, the wires pass through holes in the blocks, the holes are spaced 3mm apart.

I connected the probe to my capacitance test jig, the capacitor in the photograph is there because the enamel on the wires is damaged and there is a DC leakage path between the wires. My test jig has a small polarising voltage between the terminals, and the electrolytic effects can be seen as a drift in the readings when the probe is in water.

The results I got are as follows:

Probe disconnected      4.7pF
 
Probe in air                16pf
 
Probe in tap water      416pF
 
Probe in sunflower oil  25pF
 
Probe in mineral oil    21pF

So we see there is a big change in capacitance between oil and water.

I dont have any soluble oil to test this out any further, but as a basic concept for a measuring system, I think that it is in with a good chance.

JimB

 

[bryan1]

Hi Jim,
I have found this thread very informative and your method does seem to prove the difference a capacitive levels in fluid. If one is to measure the capacitance on new fluid then when a change occurs it can be shown. My mate owns a cnc machineshop and he did or still does have a major problem with rust forming on his machinery. He has had the coolant rep around and the last time I saw him the problem was still there.

With your experience I feel both myself and the OP will benefit if you could draw a schematic of a circuit to measure the capacitance of fluid it would be great.

Regards Bryan

 

[ronv]

oil

Are you sure the problem is with the concentration of oil or is it in the break down of the oil. Either way it still might work!

 

[Sceadwian]

Depends on his setup ronv, he's not given a lot of details, if he's taking readings from the bottom of a non turbulent holding tank then he'll get the concentration percentage he's looking for but that's not the only determining factor of the 'health' of the solution. These solutions can not be upkept indefinitely unless you're using submicron filtration for particulates. The primary problem we had with the stuff is that it was being used as a post dip after phosphating to leave a very thin almost dry to the touch oil film on the part after the water evaporates(which is what provides the real corrosion protection the phosphate is primarily a matrix to hold the oil) But over time the buildup of phosphate in the oil solution cause it it do so some odd stuff, he's not going to have to deal with phosphate contamination just heavy metal particulates, and keeping the PH neutral to keep the solution from dissolving the metals.

 

[JimB]

Fastline contected me by PM and after a few exchanges I agreed to show details of my capacitance meter test fixture which could possibly be adapted to feed one of the analogue inputs of his PIC which he is using to deiplay various parameters.

So, here goes...

JimB

 

[JimB]

JimBs Capacitance Meter

Most of the detail is as per the attachments.

The output voltage seems to saturate at about 2.1 volts which is OK as I normally use a voltmeter on its 2v range.

When there is no test capacitor connected, there is a small output voltage, this is backed off by the potentiometer circuit R6, R7, which means that the -ve side of the meter is floating above circuit ground.
This is OK for me as I use an isolated DMM, but for a PIC this may be a problem as the input is probably referenced to PIC 0v.
However it may be just as easy to forget the backoff circuit and sort out the zero with a calculation in the PIC.

JimB

 

[earckens]

Is there an alternative for the 74121? I have a hard time finding this monostable multivibrator with Schmitt-trigger input. Would the 74hc4538 be a possible stand-in? Anything else?

EDIT: old thread, sorry, I wil post as a new thread

 

[JimB]

The discussion of 74121 replacements continues here:

https://www.electro-tech-online.com...t-retriggerable-alternative-ic-wanted.153063/

JimB