Getting force plot data but it looks wrong

Post questions and issues with Concept2 PM3 SDK
Post Reply
ryantrinkle
Paddler
Posts: 3
Joined: December 30th, 2018, 2:55 pm

Getting force plot data but it looks wrong

Post by ryantrinkle » December 30th, 2018, 3:09 pm

I'm trying to pull the force plot data from the PM5 over USB using PyRow, and I'm getting data, but the data doesn't look correct. For example, for a stroke that, on the screen, looks like a relatively smooth arc, I get the following data:
[0, 12336, 6668, 35073, 397, 257, 36353, 65439, 256, 23552, 25344, 41816, 2, 45056, 1, 11, 15]

It seems like perhaps PyRow is decoding the data incorrectly, but I haven't been able to figure out exactly how. I've uploaded the full output (with some additional debugging output including the raw bytes sent/received) of "python strokelog.py" at https://ryantrinkle.s3.amazonaws.com/95 ... ceplot.txt . I'd appreciate any help you can give me in sorting it out!

jamesg
Half Marathon Poster
Posts: 2250
Joined: March 18th, 2006, 3:44 am
Location: Trentino Italy

Re: Getting force plot data but it looks wrong

Post by jamesg » December 31st, 2018, 4:01 am

Make sure your stroke style and sequences are correct, and the force curve will have the only possible shape.
78y, 188cm, 85kg, MHR 155. Last 2k (21 Jan 19) 8.27@24

ryantrinkle
Paddler
Posts: 3
Joined: December 30th, 2018, 2:55 pm

Re: Getting force plot data but it looks wrong

Post by ryantrinkle » December 31st, 2018, 1:19 pm

Hi jamesg, thanks for the response! I'm not sure I understand what you mean by sequences - is that something I need to set in the SDK?

jamesg
Half Marathon Poster
Posts: 2250
Joined: March 18th, 2006, 3:44 am
Location: Trentino Italy

Re: Getting force plot data but it looks wrong

Post by jamesg » December 31st, 2018, 1:54 pm

No, it's how rowing is done: see https://www.concept2.com/indoor-rowers/ ... que-videos

The main action sequence we use is the recovery: hands away, then swing onto the feet, then legs. This gives a strong catch posture so lets us use the legs hard and fast, causing maximum initial acceleration, as needed to reach water/flywheel speed, especially on low drag.

The force curve can be seen on the PM. If we act as above, peak force appears left of center, i.e. early in the stroke.

The curve can contain some scatter, which might be what your numbers show, but does not hide the curve shape. This scatter could be due 50/60 Hz interference with the impulse stream from the flywheel (my hypothesis only). There seems to be no effect on Watt calculations, so interest is academic only.
78y, 188cm, 85kg, MHR 155. Last 2k (21 Jan 19) 8.27@24

ryantrinkle
Paddler
Posts: 3
Joined: December 30th, 2018, 2:55 pm

Re: Getting force plot data but it looks wrong

Post by ryantrinkle » December 31st, 2018, 3:34 pm

Got it! I'm not very well-versed in rowing yet - thanks for filling me in :)

What you're saying makes sense, and lines up with what I was expecting, but the data is quite a bit noisier than I thought it would be - e.g. the chart here: https://docs.google.com/spreadsheets/d/ ... sp=sharing https://ryantrinkle.s3.amazonaws.com/D2 ... yFA-YU.png

I don't mind cleaning that data up if that's the thing to do, but I'd ideally like to do the cleanup the same way the PM5 does so that I get the same result.

jamesg
Half Marathon Poster
Posts: 2250
Joined: March 18th, 2006, 3:44 am
Location: Trentino Italy

Re: Getting force plot data but it looks wrong

Post by jamesg » January 1st, 2019, 5:03 am

What you have is coil and three magnets giving a VLF impulse stream. Surely reducing noise is not a problem if you start there?
78y, 188cm, 85kg, MHR 155. Last 2k (21 Jan 19) 8.27@24

Post Reply