Magnetic Levels Probe

[Penri]

Hello

Have made more progress on Magnetic probe, see photographs attached. The float assembly floats, the magnets trigger the Hall effect switches and the bar graph display shows the resulting level.

I found that increasing the supply voltage to the Hall effect switches improves their sensitivity so, if it can be accommodated by the buffer circuit, I'd like to run the sensor off the 19V supply, Terry can the buffer deal with that? Logic low should be around 0.4V.

You will note that the float is not the most elegant of things, it was made from stuff I had at hand that worked. It is made it in 2 parts, held together by tie-wraps so it can be removed and replaced when I source some better material. The float is made of wood covered in a number of coats of yacht varnish. I tried magnetic tape as the magnetic source but the stuff I could get proved too weak to be detected by the Hall effect devices, as mounted, so I resorted to using the button magnets I had. The magnetic assembly is made up of a ring of 6 lolly sticks, each with 3 button magnets mounted 50mm apart, held in place by hot glue and heat shrink sleeving.

I need to say that although the probe does work, that the level indications are dependent on magnet(s) and sensor(s) being adequately aligned, there are a few too many variables for it to be anywhere 100% reliable at the moment. This is entirely down to the float assembly and I'm sure it will be improved with better magnetic design.

That's it for now, happy to take comments and questions.


Hwyl

Penri

[TerryJC]

Penri,

It looks good!

As it stands, the Interfacing board can't cope with a 19 V I/P without a bit of modification. Unfortunately, the 74HC4050 buffers have an absolute maximum I/P voltage of 16 V. The 2N3904 / 2N3906 trannies used to translate the 3.3 V from the Pi could do up to 40 V, but at the moment, both are wired for 5 V, one for the Butts Pump and the other anticipating it's eventual use for driving the main pump via PWM and the Motor Drive Board. We therefore have some choices:

1. There is an input to the Interfacing Board which is pulled up to 5 V and at the moment is designated for the Float Switch at the Butts end. We could modify this by adding another resistor to drop the voltage to between 5 and 15 V (perhaps with an onboard link or two to reconfigure). This might be the best way, if you believe it would work because it requires the least number of changes to the Interfacing Board, which would remain a common item. If we eventually use a Hall-effect Probe at the Butts end, there are still a couple of unused pins on the 4050 Translators so we could easily add another pot chain there.

2. I could find space for another dedicated level translator (19 V to 3.3 V), but that would mean major changes to the board, which may mean that the board is not common to all locations.

3. Something else ????

Let me know what you think.

[Penri]

Terry

Just spotted this post, the easy answer is to leave out the led indicators and pull up the open collector o/p at the buffer end, let me think through the options.

Hwyl

Penri

[Penri]

Terry

Can you clarify what the maximum input voltage the buffer can cope with as it is configured at the moment, thanks.


Hwyl

Penri

[TerryJC]

Penri,

I'm just about to go out, so this is a quick reply.

1. According to the spec, (attached), the absolute maximum input voltage is 16 V.
2. I was guilty of much woolly-mindedness when I posted yesterday. For some reason I was thinking that there was only one input from the probe, when in fact there will be about 10. I can still add potential dividers to the Interfacing Board, but it won't be the easy fix which I thought it would be.
3. I don't think pull-ups at the probe end will suffice, because we still need to reduce the voltage to within the range of the buffers or do something completely different.

Collecting all those points together, my quick and dirty solution (which needs some thinking through) would be for the probe to provide the Interfacing Board with it's open collector outputs. The Interfacing Board would provide 19 V to the probe and then pot down the signals to less than 16 V (hence my original suggestion of 15 V).

Beyond that, we would need to buy level translators that can do better than 19 V, but I would need to spend some time to source them.

[Penri]

Terry

Thanks for the information.

What I've decided to do, for a trial at least, is to create a 12V rail in the probe for the pull-ups, I'll do this with a Zener and dropper resistor. This would allow me to keep the 19V for the Hall effect devices and provide you an output voltage within a tolerable range. Zener and associated resistors on order.

Have a good day out.

Hwyl

Penri

[TerryJC]

Penri,

So I don't really need to do anything?

[Penri]

Terry,

No I don't think so, provided it works of course.


Hwyl

Penri

[hamishmb]

Hi,

I've set the software up to use the same 10 pins as the resistance probe, just so you know. We'll have to define unique pins for each sensor to use at some point, but right now it's fairly low on the priorities list.

Are the intervals still from 0mm - 900mm inclusive?

Hamish

[Penri]

Yes, intervals as before.

Penri