B.L. Jinks

I'm having to throw darts to find a solution to my situation. Maybe y'all will know of some options.

I currently have a wired Sigma 1009. This computer has been in use for ~20 months, logging 1,400+ hours. It is connected to a 12" diameter treadwheel for an African hedgehog. It has worked fabulously well. (I did have to double the wheel size to get a number that I could program, but I just halve the data to get average speed and whatnot. Once in a while, the data would be skewed on account of her not keeping enough speed to collect accurate data, but that was a fairly irregular occurrence.)

I have completed a prototype 18" diameter treadwheel. The larger size will be good for her. Initial testing indicates that Sophie will not turn this larger wheel at a high enough speed to collect reliable data. (This wheel is large enough that I don't have to fudge the programming.) Treadwheel construction has worked out very well. We have mounted an inline skate wheel to the body of the treadwheel, which mounts to a frame with a 1/4" bolt as a spindle / axle. Super-smooth action.

Problem #1 -- The Sigma 1009 auto-offs after 3 seconds. At lower speeds, she doesn't turn this larger wheel quickly enough. Manual testing indicates that she must be turning the wheel at a minimum speed of 1.2 mph for computer function to be smooth and reliable. Her nightly average speed is only about 1.35 mph.

Problem #2 -- Actual use testing shows some issues that I didn't have before, and I don't know what to make of them yet. She has been showing some good speeds, up around 1.6 mph, yet the computer is not recording average speed at all. The distance function seems fine.? That has thrown me a bit, but I am using the computer for 'off label purposes', so figure that it is connected to not being sensitive enough for low speed use when connected to a larger diameter wheel? (The wire is run through a small hole that we cut in the wall of her room, and passes into our living room, which lets us peek at her activity in real time.)

At any rate, I need a computer that will will auto-on and auto-off --- one with a longer timeout than the 3 seconds that the Sigma 1009 provides. It needs to be accurate at very low speeds, providing reliable data for average speed from around 0.7 to 2.0 mph. Distance calculation is less important, but accurate data for that would be nice.

She runs from 30 minutes a night to 2 1/2 hours, and averaging about 1 1/4 hours these days. The data is very useful for monitoring health.

Any ideas? Thoughts?

ItsJustMe

I've never had a computer that auto-offs after so short a time. My Planet Bike is at least 5 minutes. No idea on how well it would perform at very low speeds.

One thing you could try is to mount two or more magnets and do whatever creative thing you need to in order to make it work right. Since your circumference should be 958mm, maybe mount 3 magnets, giving a pulse every 319mm, and enter 3190 as your circumference? Then you could just divide everything by 10, and you should get perfectly adequate responses on everything since you'll be getting more frequent pulses on the sensor.

20_700c

The Sigma manual for this unit suggests battery life is ~2 years - could be a weak battery issue.
Try this link for the FAQ/to download the manual.

B.L. Jinks

Two good suggestions already. I have all of the documentation / manuals. It isn't showing the low battery indicator, but the age and run time are certainly maxed out. It just uses a CR 2032 battery, which I have those on hand; her thermometers use the same battery. I'll put in a new one tomorrow. It sounds odd that only the average speed would be affected by a low battery, but who knows? The battery should be replaced anyway, so well worth a shot.
.....................................................

A second magnet sounds like a really good idea. I'll have to think it through - see what I might be missing or not accounting for. But a second pickup would double the amount of time she has to complete a full revolution and that would be plenty. I'm using small rare earth magnets because they are a lot easier to use for this application. I have a couple more here, so that should work. I'll have to offset placement left / right, rather than top / bottom, but that shouldn't be a problem. I can't have a magnet too near the sensor position because if she hops off the wheel it will swing back and forth (having a natural 'top' and 'bottom' because it isn't trued). I have to avoid having a magnet within that sensing zone or she would rack up a lot of unearned revolutions.

With two magnets per wheel revolution, I think that I would just need to halve the data. For example, average speed would be reading as double what it really would be.

With a second pickup, the computer would also think that the speed was much higher, so if reliability at very low speeds is an issue, that could eliminate that issue at the same time. I will think this through further and probably try it out tomorrow. She has put on 1 hour, 5 minutes tonight and it is showing 0.00 average speed, with a distance of 1.65 miles (max speed 4.75 mph).

Her old wheel was 12" > 958 mm. Her new 18" wheel, after deducting for the track added to it is 1,400 mm. I could change the wheel size and do it that way with an additional pickup/magnet, but just halving the data is probably the easiest thing. I was having to do that before anyway because the 1009 won't accept a wheel size smaller than 1,000 mm, so I doubled the size and have always halved the data, which we record in a spreadsheet.

ItsJustMe

If you are working with a 958mm, as I said, use THREE magnets, enter 958/3*10=3120mm and you're reading exactly x10.

You can extend this all you like. Continue to use the /2 method, but if the pulses are not coming fast enough to make your comp happy, put on 4 magnets, enter 1400/4*10=3500mm and again you're reading x10.

Nothing's going to be easier than dividing by 10. Now let's convince the US of that, and maybe we can finally go metric.

B.L. Jinks

UPDATE:

Changed the battery; a good idea but no help.

Tried the multiple magnets trick, which I had really high hopes for. That didn't work out though. The new wheel runs so smoothly that when she jumps off, although the 'natural bottom' quickly orients itself, the wheel swings so smoothly that there is no configuration of two magnets that will work. When she jumps off, the wheel will swing back forth a pretty significant distance - and a whole lot of times. She racks up bonus miles like crazy. Ha! I'm, not seeing a workaround to make the two magnet trick work for her. Too bad, it was an awesome idea.

Also, no progress on finding a computer with a calculation timeout of longer than three seconds. I need one with a calculation timeout of more like six seconds. Does anybody know of a computer where this timeout can be programmed? Too much to hope for probably.

HOWEVER -- I realized that although the average speed calculation is waaaaay off, the elapsed time calculation should be pretty darned accurate and that the distance traveled function should be fully accurate. Then I started thinking about using this data to calculate the correct average speed.

Machine Head Software, UK to the rescue:
https://www.machinehead-software.co.u...time_calc.html

Works like a charm.

I'm still on the hunt for another computer, but it looks like I am back in business.

The new treadwheel is looking like a success. After acceptance difficulties, she has had several runs now, including her best run yet on the new wheel last night: 2 hours, 20 minutes / turning the wheel 3.71 miles / at an average speed of 1.59 mph / with a maximum speed of 5.36 mph. This data is within the range that I would expect to see. A strong run for her. Yay! It is too soon to declare victory on her accepting the new wheel, but this is a good thing to see. The first six nights she refused to get on the treadwheel at all! Three of the last five nights have been pretty decent. Two of them pretty mediocre. Now that I am seeing some acceptance, time is probably on my side.

Looigi

This is way too cute. Does she have a STRAVA account?

How about posting some pix.

ItsJustMe

Trying to think of a quick solution to the swinging wheel problem - How about the wheel is mounted on a spring so that when her weight is on it, it moves down a little, maybe half an inch, and the "orbit" of the magnets then lines up with the sensor. When she jumps off, the wheel moves up and the magnets no longer come close enough to the sensor to register.

gsa103

Add a car tire balancing weight at the outer edge of the wheel. The weight will provide the natural "down" which you previous had.

Looigi

A weight will reduce the frequency, but very likely won't reduce the amplitude or number of swings.

B.L. Jinks

I'm working on a set of videos about the new treadwheel because there are people that would be interested in a wheel this size for small animals. This looks like it is going to be way better than the best wheel currently available for hedgehogs - which I have one of them (and it's good, just too small). This new wheel is only an option for people that can commit to a generous cage size. Sophie has two 26" x 68" cages, which are connected with a bridge.

Once we get a little further into identifying and making needed adjustments I'll come back and post links. We're still a little early into testing to proceed with the videos just yet. Height from floor, angle of running surface . . . there are a lot of details that can impact acceptance by the animal. One little thing can make a big difference. Identifying those details with Sophie is especially difficult because because she considers her running to be a private, solitary activity.

This will only be night twelve, so plenty yet to observe and think about.My first choice for an upgraded solution is a different computer. Changes to the wheel design need to be very carefully considered. One little change could be a lot of trouble, and then there is the question of how many things will need to be adjusted because of the adjustment.

Here is a link to where we were last week, when we had nowhere to go but up. Last night was a great night for her. I am hoping tonight will build on that.
https://youtu.be/dih9VTV-KZc


She's a bit happier in this video taken during one of her mini-baths.
https://youtu.be/w_m2NfrJaWU

.

Looigi

Hmmm. I wonder if some visual patterning on the wheel would help? Lines or something else to make the surface and motion more evident. I see it's textured, and I imagine she can see that but....?

B.L. Jinks

Hedgehogs have terrible eyesight. Her old wheel was clear - not window glass clear, but pretty clear - and the surface was smooth - not that she could see it. Plus, it is blacker than black when she is active and running. Her cage has five tunnels, a house, plus food and water dishes. Partly, she knows her environment. That is one reason that making changes can be met with a lot of resistance. They don't like changes. She navigates with her whiskers. They aren't just cute - they are a vital sensory tool. Her whiskers give her the equivalent of excellent night vision.

Another good night last night. Two runs: 27 minutes; plus 1 hour, 2 minutes. Using the average speed calculator tool at Machine Head UK is working out great. The computer, because she is inconsistently turning it fast enough to trigger reliable calculation, has been showing average speeds between 3.5 and 4.5 mph. Converting elapsed time (which WILL be a little bit off, but probably not terribly so) and total distance (which should be exactly correct) -- is showing her average speed the last few nights to be between 1.53 and 1.61 mph. This consistency is typical for her. She may trend up or down a couple points, but those tend to be long-term trends, not nightly spikes.

If I find another computer, that would get me back to having an exact elapsed time. I suspect that the variance is minimal. Without data, however, that is an educated guess.