Magnetic Level Probe Mk2

[TerryJC]

Any reason why that wouldn't work?

Doh! Just spotted the error im my spreadsheet.

[Penri]

Terry

Based on what you've said about the A/D I've ordered some 3V0 Zeners plus a selection from 2V0 up just in case. I think it would be prudent to give the A/D a little headroom as Zeners do tend to settle slightly high on their nominal value.

[TerryJC]

Penri,

However, anything over the reference voltage of 3.3 V will saturate the converter and make discerning the 'dead' zones even more problematic. Remember too that the reference voltage that' I'll be using will be the 3.3 V from the Pi (to ensure that we don't exceed the maximum GPIO input voltage), so there could be errors there too.

I've just spotted an error in my statement above! I've included a full copy of the Datasheet for the ADS115 (the IC, not the Adafruit Board), because your last post reminded me of something that I hadn't really paid much attention to; the Programmable Gain Amplifier, that is part of this chip. According to the Datasheet, the PGA defaults to Gain 2 (010), which equates to a Full Scale Range of 2.048 V. This confused me a bit because the potentiometer in Gate Valve Actuator is provided with 3.3 V from the Pi, so on that basis, the input voltage should a have been out of range of the converter at positions of anything more than around 60%. However, it doesn't, so I had a look at the Adafruit Python Module code and it appears that the gain is set to 1 by the driver, which equates to 4.096 V. This still doesn't mean that the input voltage can be allowed to stray above 3.3 V, it simply affects the FSR of the device, which in turn affects a number of other performance factors.

The bad news therefore is that we are not making use of the full range of the converter! This can be rectified by setting the gain back to 2 in the code, but then we would definitely have to use a reference voltage of 2 V or less. That could make the 'All-off' voltage extremely small indeed and might start to get affected by noise. Probably not too important, but the I/P impedance of the converter in Gain 2 is 4.9 M Ohm, and 15 M Ohm in Gain 1. I think neither of those values would have much affect on the reading that we will get from our 'variable-output-impedance' sensor, but clearly the Gain 1 case would have one third of the effect.

So I think that the 3.3 V Zeners would still be OK, as long as we are sure that they don't produce more than around 3.5 V; that voltage would still be within range. However, whatever voltage we use, we may find that we have to calibrate individual probes, based on the voltage generated by their Zeners, if we are not to get errors at the bottom end.

[Penri]

Thanks for that research Terry, I'll sort through a few 3V3 Zeners and see what I get. I also need to tinker with the resistor value to make sure the Zeners get enough current to work properly throughout the load cycle.

[TerryJC]

Penri,

Have you sealed the probe yet? If not, I have a solution to the All-off situation.

It hit me last night; we can improve the voltage difference of the probe when all sensors are off by increasing the value of the resistor R1 to around 640k. In the spreadsheet, which displays 2 decimal places, the output voltage is 3.3 V (with a reference of 3.3 V) giving a voltage difference of 40 mV instead of 10 mV. None of the other values are affected because devices isolate all resistors below the active device.

[Penri]

Have sorted the over voltage Zener issue, the dropped resistor was too low a value so too much current was flowing. I've upped the value to 2k2 and the ALL OFF voltage is stable at around 3.2V.

I'll bring the set up in on Monday for you to see. If you can bring all your voltmeters in too (to add to my 2) we can see how each channel performs.

Mods to the circuit diagram and spreadsheet necessary, will post later.

[TerryJC]

Have sorted the over voltage Zener issue, the dropped resistor was too low a value so too much current was flowing. I've upped the value to 2k2 and the ALL OFF voltage is stable at around 3.2V.

I'll bring the set up in on Monday for you to see. If you can bring all your voltmeters in too (to add to my 2) we can see how each channel performs.

Penri,

Will do. However, I'm still concerned about the low differential between the 100 mm level and ALL OFF. Did you look at my previous suggestion?

[Penri]

Hello Terry

Yes I did read and take note of your suggestion and have ordered some resistors to implement it. In the meantime, I've been reviewing what I built against the diagrams drawn and the spreadsheet model and found a difference.

I've attached a revised circuit diagram and spreadsheet which we can discuss along with trying out the physical probe tomorrow / Tuesday.


Hwyl

Penri

[TerryJC]

OK. See you tomorrow.

[Penri]

Design change to probe to add decoupling capacitors across supply and zener diode.

Amended circuit diagram within updated principle of operation document is attached, together with photographs of the probe interface board and the board when wired.