Source Sealed Potentiometers?

Hi John,

Have you considered ant other technology as Hall sensors ? Not as straightforward maybe, but they keep working in harsh enviroments.

I'm using :

Stan

IP 67 is flush water, IP 68 is under water. You want to measure an angle ? What is the angular speed of the sensor ?

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net

No problem using Hall effect sensors, but I need multi-turns.

Thanks,

JM

I need to be able to survive immersed in fresh water for several minutes. I think I might be able to get away with IP67.

I need to be able to measure an angle 0-360, & can't directly couple to the shaft/axis of rotation. I can put a gear on the object to be measured, & due to space constraints I need to reduce at least 5:1.

Angular speed would be slow, probably no more than about 30 RPM.

JM

Is cost an issue or are you able to look at synchro/resolvers as a means to measure the angle. There are some around that, due to their construction, are OK for immersion. The solution is more expensive than a pot for measuring angle but is quite dependable.

--
********************************************************************
Paul E. Bennett ....................
Forth based HIDECS Consultancy .....
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************

I'd like to keep the cost down, for sure. I haven't found any synchros or resolvers that can deal with multiple turns, & I don't have a way to tell which turn it might be on after power-up/reset. A multi-turn pot. would work very nicely, but all I've been able to find that are immersable are pretty pricey (i.e. oil-filled, $US80 & up).

JM

You stated that you could use a gear for reduction. The way resolvers are usually used for keeping track of multiple turns of a shaft and how far the shaft itself has gone would be to use two. However, as you have stated that direct drive is not possible and you would be gearing anyway you could just gear the synchro/resolver appropriately to gain the desired revolutions in one turn of the synchro/resolver shaft. Naturally, the applicability of this woul depnd on the accuracy and resolution you require of the measurement.

We are all groping a little with such scant detail of the situation.Perhaps a little more about the nature of your requirement here would help us help you.

--
********************************************************************
Paul E. Bennett ....................
Forth based HIDECS Consultancy .....
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
********************************************************************

What I need to do is rotate what amounts to a lazy-susan/turntable/camera mount (pan only). I need to rotate up to

360 degrees, but not continuous. The device sits on a bearing so there is no center shaft that I might couple to. OD is 3". My plan is to attach a gear (I've got that part worked out OK) to be driven by an electric motor. What I need is some sort of position feedback. The gear I've chosen for the turntable has 156 teeth with a pitch diamter of 3.25". I figure that with a mating 32-tooth gear, I can attach a 5-turn pot. (ratio = 4.875:1, leaving a little travel at either end). What I DON'T have, besides a shaft to couple onto directly, is space for another 3.25"-dia. gear to keep the ratio of turntable to position sensor at 1:1 or below, & I'd like to avoid using a multi-stage reduction gear if I don't absolutely have to.

I also have to be able to survive immersion in fresh water for up to 30 minutes.

Thanks!

JM

... snip ...

What sort of positional resolution (in degrees, or even in gear teeth count) do you require? Something like a pair of leds for tooth resolution, and a reset mechanism via a microswitch could provide the position. The pair of leds can detect rotation direction, and the whole schmeer should be immune to over-rotation, unlike a pot. Counting the 156 teeth should give you roughly 2 degrees of positional resolution.

Your only worry would be the effect on the light path of the immersion, which should not be excessive in fresh water.

--
 Some informative links:

2 degrees of resolution is enough.

A prior prototype used a pair of reflective optical sensors, with black & white strips on the rotating assembly. One sensor read a narrow white "index" stripe that indicated the unit was in its "home" position, while the other read either black or white, depending on whether "home" was CW or CCW. That worked well, when I needed to do a realignment such as on power-up. I had a couple of issues that have led me away from this solution:

- I don't have 100% confidence in being able to count teeth reliably. Once I believe there has been a missed step, it's easy enough to do a re-alignment, but I'd rather not have to do that. The prior prototype used a stepper motor, & kept a running count of steps after it was aligned to its "home" orientation. I didn't feel like I could gurantee that the unit couldn't get physically forced out of position, to where its actual orientaton would be different from where I thought it is, based on the step count.

- I don't have a lot of confidence in the use of optical sensors in what I expect to be a rather dirty environment.

I need to know, with a high degree of certainty, the orientation of the platform at all times (within the limit of about 2 degrees). Counting teeth, I believe, amounts to a guess & once the certainty of position is gone, it's gone.

Thanks.

JM

45 years ago we controlled a gamma ray spectrometer by counting teeth. The sensors were microswitches that were operated by a cam on an idler gear. Quite reliable. You would need two such, with appropriate phasing to tell direction. We used the idler gear teeth proper as the cams. The whole system was a collection of relays and an ordinary motor to do the actual moving. Modern systems would need positive switch debouncing, which can easily be done with a spdt break before make switch and a set/reset flip-flop. You would want the sensors out of phase by 1/2 tooth, but that would not be a critical adjustment. The idlers would have 6 to 8 teeth.

--
 Some informative links:

Gill Instruments has some non-cotact sensors that might work for you. You would need to be able to add to (or modify your existing turnable) so that a varying amout of metal passed near rhe sensor depending on angle.

Robert

I'd still need a method to tell when I'm in some exact position, & some way to tell whether that position is CW or CCW if I'm not exactly there already, so that I can have a reference angle from which to start counting teeth. I'm interested in absolute position, not relative.

Thanks.

JM

I'll give that a look.

Thanks.

JM

--snip--

--snip--

--snip--

--snip--

You can purchase "absolute encoder" discs for optical encoders. These have black/white or opaque/clear segments that provide a binary value to a row of sensors so you can read the disc's current position "directly" (possibly through, say, a Gray Code lookup table).

The same principle can be applied to a magnetic or mechanical encoder. Put bumps on a disc, or lay down axial rows of tiny magnets to be read by a row of Hall Effect sensors.

Is there any way you can apply this approach? You'er still faced with the problem of choosing sensors (optical, Hall Effect, microswitches) that can operate reliably underwater in fresh water.

Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut minds pring dawt cahm (y'all)

-- "Perfection [in design] is achieved not when there is nothing left to add, but rather when there is nothing left to take away." -- Antoine de Saint-Exupery

--

You add a third switch operated by some sort of detent position on the main turntable. You then rotate until this operates in a pre-specified direction (to avoid backlash effects). Bingo, master location is set.

You can also get gray coded rotation sensors. They will probably cost more, but are absolute. Then your software will have to do the gray to binary conversion, which is not a big deal.

--
 Some informative links:

I know about absolute encoders, but I still have the problem of needing to track multiple turns, since I don't have a shaft to directly couple to or the space to add a full-sized 1:1 gear.

Magnetic strips might be a possibility, to tell me which segment of the circle I happen to be in, with the resolver providing more-than-sufficient resolution within that range. If I could determine magnetic polarity, I should be able to reduce the number of sensors to 2, with 4 strips staggered around the perimeter of my turntable:

High/North, Low/North, none, Low/South, High/South.

Then, with 5:1 gearing on the resolver, that I know fits within the confines of my space, the combination of segment sensor/resolver shold tell me what I need to know.

So, there's another question for somebody: How can I detect not only the presence of a magnetic field, but its polarity? Off the top of my head, I don't know if that's something most Hall-effect sensors are capable of doing.

Thanks.

JM

I can do that easily enough & have done so before on a prototype, using optical sensors. The only trouble is, I won't necessarily know which direction to rotate to find that index. I can go 360, but not past.

Right. That can be done easily enough in a table.

Thanks!

JM

--snip--

And you can't add optical or magnetic sensors under the platform or around its perimeter?

Ah. I misread the "2 degrees" (e.g. 1/180 revolution) resolution requirement, so I was picturing someone embedding _lots_ of littly teeny magnets in a clear resin disc.

Take a look at these linear Hall Effect sensors. Their output is a voltage that varies in proportion to the magnetic field, rising for one pole and dropping for the other:

Linear Hall-Effect Sensors

A1301 and A1302 Continuous-Time Ratiometric Linear Hall Effect Sensors

For examples of other types, look at:

Hall-Effect Sensors

Another approach would be to use two sensors next to each other and bias one with a small magnet to reverse the polarity it responds to.

But, as you point out, if you use a "sensor wheel" that is making multiple turns while the platform is making its full limit of 360 degrees, none of this _guarantees_ that your monitor didn't go to sleep and so miss a whole rotation of the sensor wheel.

So: is there any way of putting limit switches (or some sort of sensor) on the platform?

Frank McKenney, McKenney Associates Richmond, Virginia / (804) 320-4887 Munged E-mail: frank uscore mckenney ayut minds pring dawt cahm (y'all)

-- You can't tell which way the train went by looking at its tracks.

--