Ok, i got a bit strange situation. I have one analog gas probe that has range from 0 - 180 Ohm ( 0- 2V ). The device that reads the values of this probe awaits values from 0 -
24V so i'v done this:
formatting link
(Sonda means Probe )
When the float is in the top position output has 24V so this is OK. When the float goes slowly down the voltage drops and this is also OK.
But then at the end of the probe ( few cm till the lower end )voltage raises again to 24V. Now I do not know if this has something to do with amplifier but i gues that my schematics is wrong...
That amp is designed for a split supply -15v and 15v so you will run into commom mode problems and magic smoke @ 24v. You could put a resistor in the X1-2 point to ground such that the sensor X1-1 voltage point never goes below 3v (EG -12 for a dual supply). Id pick one so that you'd get a range of 5v-15v.
Or pick another opamp with better common mode rejection and 24v vcc.
why the 24V supply? op-AMPS like this like a little headroom.
Float? when you say "gas" you mean gasoline?
Adding a small capacitor in parallel with R2 could stop oscilations a larger capacitor will also smooth out "slosh"
Is the ouput sourcing or sinking current? your claim that 24V works suggests that it's sinking. so, how much curretn flows from the load when grounded?
Is the 24V supply to the sender kosher? if the sender develops an intermittent open it could spark if there's enough cable capacitance.
He needs to employ a bridge on the input to keep the op-amp in the center of the rail voltage.
When the probe goes the full stroke either way, It can do the 2 volt offset swing and it'll stay with in the common mode range.
Also, looking at the PDF, it looks like the output on the low side will not get close to the low side of the rail, or maybe that was the high side, either way, a full rail swing on the output is not possible with that op-amp
There are ways of course to get around that if he wishes to continue use of that op-amp..
Seems the circuit is similar to a pot +24 to ground with wiper as the measure point. But modify it as +24 to resistor to measure point, then rheostat with wiper to ground. But one end of rheostat, the low R end is incomplete and the wiper in continuing its motion, "drops off" the resistive "track" which opens the circuit (no more ground). This is sort-of common when a (cheap) wire-wound rheostat is used. Have seen this maybe 3-4 times in the past 50 years.
Change the op-amp to a LM358 or LM324 type. These are so called "Single Supply" op-amps meaning that the inputs function ok down to the op-amp's most negative supply terminal (in this case gnd).
You will find that this will only work well for loads that draw any load current in the direction from the op-amp output to GND, like a resistor or an old style meter. Loads that feed current from a positive voltage into the op-amp's output will cause the op-amp to have difficulties for very low output voltages.
Not that it is directly related to your question, but in the hope of saving you another question later: You will also find that there are problems (due to oscillation) if your load is capacitive, like a long shielded cable. If, and only if, you think your load is capacitive and/or experience/suspect these problems, then please read the rest below:
The LM358 has a high open-loop output impedance for very low output currents (sometimes happens when the output is near ground). I suggest that if your load is capacitive: (1) try to make your load not be capacitive e.g. by removing capacitors, but if you can't, then (2) if possible add a few kOhms resistance in series with your load to make it not so capacitive, and fiddle with the gain and supply voltage if necessary to correct it for the resistance that you had to add, and (3) try drawing some extra load current from the op-amp output to ground to make it have lower open-loop output impedance which lowers the phase shift within the feedback loop resulting from the load capacitance. The simplest way to draw some load current is to add a resistor from the op-amp output to GND, but it is difficult to choose a resistor that doesn't draw too much current at 24V yet draws enough current at e.g.
0.1V, so a better option might be a current sink. If you need one of those thne post back here. or (4) look for some fancy new-fangled op-amp that likes capacitive loads (rare) can take 24V (rare these days) and is "single supply" or "rail to rail". This will probably be more expensive and quite likely will only come in a package so small that it disappears when you sneeze.
Yes, I've seen this several times as well. Best to check the performance of the device with an ohmmeter. Check wiper to each end, one at a time. Rotate knob to see if the resistance becomes infinite (open) at either end.
---------------------- We switched the op-amp to LM358 and on the amp exit we put one 100k. I don't know if this is OK, because you have written that some power should be drawn from amp itself ( if i get this right ). With 100K i can do that but i don't know if this is effecting the rest of the schematics.
In our case, we are connecting amp exit to one external device that has no other function then to read voltage level. Those voltage levels are controlled over SW on the device and translated as gasoline tank values.
You could try with and without the 100k. If it works ok pretty much all the way down to 0V without the 100k fitted then you could leave it out. From your description I can't tell what current the "external device" load draws (or sources) but hopefully it will work ok without any extra load.
------------------ OK. Am facing one strange problem but i dont think that it has something to do with the gasoline probe. When you are moving floating "part" in some faster way, you can not see any problems but when i move it slowly then we got a problem.
It looks like that the electronics dont recognize the values in 10 mm range. To be more exact:
- probe length is 15 cm.
- if i pull the float from the start to top i will get values from 0 -
23.9V but if i pull it slowly then we can see the problem
Just an example:
0cm - 0V
1cm - 0,5V
2cm - 1V and so on...
This would not be any problem but in the real situation, if you have gasoline tank with let's say 400 liters and if the float "part" goes down for 8mm ( when the car is still, not on the hill etc.), nothing will be measured and this big problem because on such big tanks someone can steal few liters and nothing will happen, because the electronics will not recognize any changes...
Yes, sounds like you may have a mechanical problem with the device generating the reference. I don't know what it is but, if it's a wiper of some kind, it maybe sticking then snapping to the next position, doing it fast would make it move constantly.
The other possibility is you have some low resolution reference on the probe and in your circuit you have an integrator to slow things down in the response time. When moving fast, it may appear to look smooth because the integration is slowing it down between 1 cm steps and thus looks smooth. Doing it slower, allows for the integrator to settle faster than the probe movement, and then you see the problem.
I would measure the probe output directly before it goes through your circuit. That is the only way I could see where the problem may exist.
I'd be curious as to what kind of apparatus is attached in the probe to detect that? Or are you using a capacitive probe? Seeing that it's gas, not knowing what type, I can only envision flames!
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.