I was considering a home-built unit. It shouldn't be too difficult to build such a unit youself, although I've seen industrial type units (but are very expensive!). You could take a straight bar of stainless steel, at the required length, which is 1mm thick, and 7.5mm wide. Twist it, say one full twist every 72mm. Put a float on it, with a slit in the center of the float that accepts the twisted bar, and put guides on the float that keep it from twisting. As the float travels up and down the bar, the bar is twisted 1 degree for every .2 mm the float moves. Put a 72 position rotary encoder at the top and you get 1 pulse per mm. You can work it from the other end by finding the encoder you want to use, and then twisting the bar to give you your 1mm resolution, but the steaper the twist, the more likely the float will bind, and your resolution will go out the porthole. Use a 360 degree pot instead, and you can get as good a resolution as your A/D can handle (excepting mechanical limits), and you won't have to constantly track a rotary encoder, either. If you want to trade resolution for ease of assembly, just mount a float on an arm connected to a pot. It is a non-linear relationship, As the arm nears vertical you'll see the pot change little for a large change in water level. If you think it through, though, you can probably find a way to mount it such that it measures full scale from, say 45 degrees below horizontal to 45 degrees above horizontal. -Adam Andy Shaw wrote: > > Thanks Adam, > any advice on which float system to use and where I can get hold of one in > the UK? > > Andy > > ----- Original Message ----- > From: "M. Adam Davis" > To: > Sent: Monday, October 16, 2000 3:42 PM > Subject: Re: [EE]: Help water level > > > Given that it is for a sub, the water shouldn't slosh around all that > much. Why > > not use a float? It is probably a little larger than you want, and > perhaps > > you'd rather have a non-mechanical sensor, but for ease of assembly and > reading > > and simple float setup with a pot or rotary encoder can't be beat. > > > > When you do go to another expiremental method, it would be benificial to > have > > the float working as well so you can compare the readings and improve your > > design, before getting rid of the float. > > > > -Adam > > > > Andy Shaw wrote: > > > > > > Hi Folks, > > > I'm trying to put together some sort of sensor to measure the amount of > > > water in a model sub ballast tank. The requirements are: > > > 1. Must be small (not a lot of space). > > > 2. Easy to seal against water. > > > 3. Continuous range from say 1 cm to say 30 cm deep, with 1mm accuracy. > > > 4. Possibly best not to use anything that depends upon pressure (since > this > > > may vary depending upon sub depth). > > > 5. Easy to hook up to a PIC (either digital or analogue OK - I have > both). > > > 6. Cheap! > > > Not asking much! > > > > > > I seem to remember seeing a rain gauge project that used two (insulated) > > > wires separated by a fixed gap as part pf some sort of oscillator. But > guess > > > what I've lost the mag. Anyone got the details of know if this would > work > > > for me? > > > > > > Thanks > > > > > > Andy > > > > > > -- > > > http://www.piclist.com hint: The PICList is archived three different > > > ways. See http://www.piclist.com/#archives for details. > > > > -- > > http://www.piclist.com hint: The PICList is archived three different > > ways. See http://www.piclist.com/#archives for details. > > > > > > > > > > -- > http://www.piclist.com hint: The PICList is archived three different > ways. See http://www.piclist.com/#archives for details. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.