Measuring seat position - rowing machine

Projects General Guidance
1

Hi All,

I'm putting together a replacement computer for my cheapo rowing machine. One of the things I would like to do is measure the seat position. See attached pic.

Does anyone have any clever ideas about how to do it? I assume things like infra red distance sensors need large targets. Reed switches and a magnet on the seat would do it I guess, but would lead to lots of wiring if I want more than a couple of positions.

I am considering an alternative approach using an accelerometer (e.g. MPU6050 etc), but I guess these also would need a lot of wiring.

Cheers

Brutha

rowingmachine.jpg
rowingmachine.jpg1636×663 253 KB

2

Belt attached to the seat that runs over two free running wheels, attach quadrature encoder (so you can sense direction and amount of movement) to one of those wheels. Add a limit switch of sorts on one end for easy zeroing of the start position.

3

Brutha:
One of the things I would like to do is measure the seat position.

.... lots of wiring if I want more than a couple of positions.

When I was a cox back when I was a lad, I used to call 1/4, 1/2, 3/4 and full for the first strokes from the start, so I'm wondering why you need any real precision? Wouldn't knowing 5 positions suffice?- both ends and 1/4, 1/2 and 3/4? (But of course it depends why you want to know the position... it's a looong time since I was on the water.)

edit: Does a Concept2 "Dynamic" @1200GBP give the seat position?

4

wvmarle:
Belt attached to the seat that runs over two free running wheels, attach quadrature encoder (so you can sense direction and amount of movement) to one of those wheels. Add a limit switch of sorts on one end for easy zeroing of the start position.

Ah, that's a nice idea, thanks! Together with the limit switch I could possibly make it even simpler, and just have a wheel attached to the seat, running against the support just behind it. Although the belt idea means less wiring and the electronics staying in one spot which is nice.

I'm also wondering now if I could use some kind of optical sensor arrangement, perhaps with tippex on either the support or one of the axles under the seat might work.

5

elvon_blunden:
When I was a cox back when I was a lad, I used to call 1/4, 1/2, 3/4 and full for the first strokes from the start, so I'm wondering why you need any real precision? Wouldn't knowing 5 positions suffice?- both ends and 1/4, 1/2 and 3/4? (But of course it depends why you want to know the position... it's a looong time since I was on the water.)

edit: Does a Concept2 "Dynamic" @1200GBP give the seat position?

Hehe, no I'm sure it doesn't in fact! And I know very little about rowing, so any input on what is actually useful is welcome!

It was prompted a bit by someone on Reddit, where I posted to see if anyone had any good approximations for turning the stroke speed into actual distance and calories. They were idly reflecting that the seat movement would tell you if you are "shooting the slide" (some fault in your rowing technique that I have yet to fully understand!!!).

The info that I get from the sensor in the actual rowing bit is very basic (just 4 or 5 pulses as you pull the cord out and the same as you return it). So, the seat position would also be handy to identify when a full stroke has actually taken place.

As you say, a limited number of positions would be enough, but even 5 reed switches would be a reasonable number of wires and mounting.

6

Brutha:
turning the stroke speed into actual distance and calories.

If you mean the stroke rate (strokes/minute) then I don't think there is a correlation. In fact if you look at the rules for ergo regattas, there's actually a limit to the rate of (iirc) 19 strokes/minute. A good rower can get huge distance per stroke at low rates and a crap rower like me could flay around like crazy and hardly move the boat.

Maybe that's what Concept2s cost what they do: the dyno is very sophisticated and carefully calibrated so that it gives a good indication of distance and speed over the water, and of course the energy consumed. As far as I know, ergo times are quite a good indication of water times.

If there's no capture of seat position on a Concept2-level ergo (and bear in mind these things are used for training at Olympic level under the eye of serious sports scientist boffins) then we can safely say that such data will have no meaning in an Olympic level training regime. (But I don't know if they do capture that or not; last time I was involved with my son's rowing was 10y ago on a Concept2 ModelD with the PM3 computer. It certainly didn't capture the seat position.)

7

Yes, I'm sure you're right.

What I'm really looking for is a way of giving a plausible distance and calorie count, rather than accurate ones. It's so I can compare figures between days essentially, and give me something to work at!

It's very difficult to accurately count the strokes based simply on the sensor in the machine (the current computer is rubbish, giving several strokes I've only done one most of the time). It's because the output varies between 4 or 5 pulses depending on how far you pull the handle.

So, the seat movement was originally about counting the strokes accurately, and I'd really only need to have one reed switch in the centre of the support for that.

But I thought I might be able to come up with something that gave me better distance with better technique, and the seat position I thought might be helpful in that. Perhaps not though!

8

Brutha:
It's because the output varies between 4 or 5 pulses depending on how far you pull the handle.

Ok so if you got 4 pulses, you can't tell is it's 2x short 2-pulse strokes or 1x long 4-pulse stroke, hence the need to check the seat position?

But of course the nature of rowing is that you can sit with your legs stretched and not move the seat, yet still do a full stroke on the rope with your arms and waist...

Won't there be a gap in the pulses while the rope rewinds, so 2x short strokes might be pulse-pulse, gap, pulse-pulse while 1x long stroke would be pulse-pulse-pulse-pulse?

(Interesting that the PM5 computer on a Concept2 costs 155GBP which is about what I think you said your ergo cost :wink: )

9

elvon_blunden:
Ok so if you got 4 pulses, you can't tell is it's 2x short 2-pulse strokes or 1x long 4-pulse stroke, hence the need to check the seat position?

Yes, essentially that's the problem!

elvon_blunden:
Won't there be a gap in the pulses while the rope rewinds, so 2x short strokes might be pulse-pulse, gap, pulse-pulse while 1x long stroke would be pulse-pulse-pulse-pulse?

No, unfortunately not: basically there's a reed switch on a mounting inside the housing, and a magnet mounted on the wheel the cord is wrapped onto (not on the flywheel). So, it pulses no matter which direction the wheel is running. BTW, when I say pulse, I just mean the reed switch opens.

And the number of pulses is quite variable; either 4 or 5 in each direction (and sometimes the magnet will be over the switch at the end of the pull, so I get 9 in total).

elvon_blunden:
But of course the nature of rowing is that you can sit with your legs stretched and not move the seat, yet still do a full stroke on the rope with your arms and waist...

Yup, and this was the thinking behind tracking the seat position and maybe using it somehow in the distance calculations - I guess there is some optimum sequence of movements. By measuring the time between pulses I could get the speed I'm pulling the oar at 3 or 4 points in the stroke, and by measuring the seat position I can tell whether I'm doing it with my legs or arms/body. The idea would be that if I do the stroke "right" I would get better distance.

elvon_blunden:
(Interesting that the PM5 computer on a Concept2 costs 155GBP which is about what I think you said your ergo cost :wink: )

It is! I think all mine does is count the pulses from the cord wheel, and multiply by fixed factors to get strokes, calories and miles!

10

elvon_blunden:
If you mean the stroke rate (strokes/minute) then I don't think there is a correlation. In fact if you look at the rules for ergo regattas, there's actually a limit to the rate of (iirc) 19 strokes/minute. A good rower can get huge distance per stroke at low rates and a crap rower like me could flay around like crazy and hardly move the boat.

Maybe that's what Concept2s cost what they do: the dyno is very sophisticated and carefully calibrated so that it gives a good indication of distance and speed over the water, and of course the energy consumed. As far as I know, ergo times are quite a good indication of water times.

If there's no capture of seat position on a Concept2-level ergo (and bear in mind these things are used for training at Olympic level under the eye of serious sports scientist boffins) then we can safely say that such data will have no meaning in an Olympic level training regime. (But I don't know if they do capture that or not; last time I was involved with my son's rowing was 10y ago on a Concept2 ModelD with the PM3 computer. It certainly didn't capture the seat position.)

Having thought about this a bit more, clearly the easiest and best approach is going to be to mount some magnets on the flywheel itself, add a reed switch and simply ignore the existing sensor. I was given this advice before, and your info above amounts to the same thing, but for some reason I ignored it, not sure why!

I've ordered a cheap (£20!) 2nd hand monitor from eBay, and a raspberry pi, will be interesting to see what I can do with it!

If anyone is interested, there are some useful formulae in the link below: of course, I won't be able to calibrate like the concept2, but at least I should get properly proportional speed and calories etc.

http://eodg.atm.ox.ac.uk/user/dudhia/rowing/physics/ergometer.html

11

Brutha:
I should get properly proportional speed and calories etc.

That sounds good yes. It may not return a calibrated distance vs time as a Concept2 would, and you may have it set so that you do 2000m in about 7 minutes which would earn you some seriously shiny tin at a regatta, but you would know that you're getting better or worse I guess, over time.

When my son was competing, the first thing his buddies said to each other when they hadn't seen each other for a while was "so how's your ergo time?' That's why one make has pretty much taken the market; the numbers are transferable across machines.