Drag Factor on Model A

[walterpump]

Hi Gang: Is there a way to calculate Drag Factor for a Model A depending on which gear is being used. I would love to be able to compare my performance with everyone else. I am loving my Model A! It has the original PM1 Monitor which works great.
Cheers. Jim

[JaapvanE]

I don't know if a PM5 can be retrofitted. You could attach a Raspberry Pi (even a Zero is known to work) to it, and install Open Rowing Monitor on it. That can calculate it based on some assumed values for flywheel inertia, which you can estimate based on the resulting speed. A challenge is the electrical connection, although it probably is a simple reed switch.

Alternative is to measure the flywheel decay by hand and calculate it from there (see https://github.com/JaapvanEkris/openrow ... e-flywheel).

[walterpump]

Thanks for the info JaapvanE; I will research the links.
Yes a PM5 can be fitted to the Model A, but I have no intention of doing that. the monitor is more money than my Model A - ha, ha.
Even if a PM5 is fitted it will not provide the correct drag factor as the Model A is too different from all the newer models.
So calculating the DF is the way to go, and I will check out the links you provided.
having fun with my new-to-me Model A. The PM1 monitor on it works great also.
Cheers & happy rowing.

[gvcormac]

The PM5 is programmed to calculate drag correctly for older models. All it needs to do that is the angular mass of the flywheel. The gearing and blade setting is irrelevant.

I understand not wanting to spend the money on a PM5, but a Model A is definitely a collector's item. Whether it is more valuable to you original or PM5 I can't say. If you were to upgrade, you should keep the original speedometer, in case somebody in future wants to restore it to original.

[Citroen]

The PM5 is programmed to calculate drag correctly for older models.

Except, it doesn't do drag factor on a model A.

[MPx]

While drag factor is "of interest" it doesn't really feature in comparing performance vs others. Some erg their best stuff at DF 100 while others chose 200 - DF itself is therefore largely irrelevant. So if Jim's interest is comparison of "performance", then a quick flick through the relevant rankings should provide the info sought. My understanding is that many of the shortcomings of the PM5 are essentially legacy as C2 have been diligent in making sure performances can be compared across all models of their erg - though I have to admit I've no idea if that's true for the PM1 as my journey started with a PM2! A worthy stance in may ways, but annoying that no improvements can be made on the clunky stuff - rounding anyone!

[walterpump]

That is very interesting MPx.
So I can go ahead and compare my performance with others and disregard the drag factor.
That is good to know.
Cheers.

[Ombrax]

That is very interesting MPx.
So I can go ahead and compare my performance with others and disregard the drag factor.
That is good to know.
Cheers.

Walter, this ^^^^ is an important point.

The DF (or on other C2 ergs, the damper lever setting - I don't think the Model A has a damper shutter like the later models) is not a "degree of difficulty" measure. For example, in races participants can put the damper lever anywhere they like, at 1, or at 10, or anywhere in between. If you set it at 1 (and therefore have a low DF) it will be much easier to pull the handle, and the flywheel won't spin down very quickly, but you won't get as much "credit" for the pull because the PM knows that it was at a low DF. If you set the lever at "10" it will be much harder to pull the handle at the same speed as you did for "1" and the flywheel will coast down much faster because the drag is higher (due to more air getting in there). It's much harder to do, and the PM will reward you with a lower time/500m pace because it knows that you worked harder (assuming you did).

Bottom line, folks using the erg set the damper lever (the models that have one) wherever they like for the workout they're about to do and go for it. Typically big huge weightlifter guys who are super-strong (but don't have great cardio) who like to do 500m sprints use a DF that's as high as they can get it. Normal folks rarely do that.

Good Luck

[gvcormac]

The PM5 is programmed to calculate drag correctly for older models.

Except, it doesn't do drag factor on a model A.

Really? Do they advertise that? I'd be major pissed off if I spend $160 not knowing that.

[gvcormac]

The PM5 is programmed to calculate drag correctly for older models.

Except, it doesn't do drag factor on a model A.

Really? Do they advertise that? I'd be major pissed off if I spend $160 not knowing that.

Or do you just mean it doesn't display drag factor? Surely it uses it to calculate pace. If not, what use it a PM5?

[Citroen]

It says it here: https://www.concept2.com/service/indoor ... /retrofits

[JaapvanE]

Or do you just mean it doesn't display drag factor? Surely it uses it to calculate pace.

Nah, not really. In the end, all you need some conversion between angular velocity and linear distance, which can be done easily without any dynamic calculation of drag. Especially as the Model A lacks any method of setting drag (changing gears does NOT change the work done, so it shouldn't change speed either) there is no actual need to calculate anything dynamically.

You could go the complex way and predetermine a fixed drag factor and use that in all calculations following the science as described by the physics of ergometers. Or you could take the very easy approach and ignore all complex science bits and make a glorified pulse counter that kind of displays the right speed based on the feel of one of the Dreisigacker brothers, similar to tuning a bike speedometer to a wheel radius. I assume they did the last, as it is build quite similar as a bike speedometer, and it is even the approach of 90% of the current market (explicitly excluding the more advanced brands like Concept2, RP3 and BioRower).

[gvcormac]

Or do you just mean it doesn't display drag factor? Surely it uses it to calculate pace.

Nah, not really. In the end, all you need some conversion between angular velocity and linear distance, which can be done easily without any dynamic calculation of drag. Especially as the Model A lacks any method of setting drag (changing gears does NOT change the work done, so it shouldn't change speed either) there is no actual need to calculate anything dynamically.

You could go the complex way and predetermine a fixed drag factor and use that in all calculations following the science as described by the physics of ergometers. Or you could take the very easy approach and ignore all complex science bits and make a glorified pulse counter that kind of displays the right speed based on the feel of one of the Dreisigacker brothers, similar to tuning a bike speedometer to a wheel radius. I assume they did the last, as it is build quite similar as a bike speedometer, and it is even the approach of 90% of the current market (explicitly excluding the more advanced brands like Concept2, RP3 and BioRower).

Air pressure, bearing resistance, bent vanes, dust, all affect drag. I don't see why Concept 2 would invent a whole new, inferior, algorithm for Model A, rather than just plugging in the moment of the A flywheel.

[gvcormac]

It says it here: https://www.concept2.com/service/indoor ... /retrofits

I see. What they state, "Please note that the drag factor function will not work on a Model A," is ambiguous. I suspect -- but without proof -- that they simply mean it won't display drag factor, not that they don't calculate it. I suppose they made that choice because -- as has been noted -- it is not adjustable. Perhaps they thought it would be confusing to display a quantity over which the user had no control.

[JaapvanE]

Air pressure, bearing resistance, bent vanes, dust, all affect drag. I don't see why Concept 2 would invent a whole new, inferior, algorithm for Model A, rather than just plugging in the moment of the A flywheel.

This article suggests that the model B's monitor could be fooled, by changing the damper setting (as they use a running average of 15 strokes, you get a advantage by dropping the dragfactor mid race). But the article is unclear if it relates to a PM2 or PM1.

You need quite accurate recovery detection to get a decent drag calculation (drive detection is much easier, which typically is used for stroke detection). I seriously doubt the magnetic disk used by the PM1 retrofit is accurate enough to do meaningfull recovery detection and calculation, as the decay of the flywheel needs quite detailed modelling to get to the dragfactor. To clarify, on OpenRowingMonitor we use Linear Regression on the C2's 200 datapoints to get to a useable number. In my practical experience with calibrating rowing machines, reducing the number of datapoint has quite some adverse effects on dragfactor stability. Using a running average for 15 strokes migh fix that, but it is a crude method.

Please note that you shouldn't overestimate the ultimate effect of the drag factor onto pace. When you look at the formula for linear velocity, it is:
LinearVelocity = Math.pow((dragFactor / (magicConstant * 1000000)), 1/3) * AngularVelocity
Where the magicConstant traditionally is 2.8. So a variation across dragfactors is divided by 2800000 after which the cube root is taken, which renders small variations in dragfactor largely insignificant. Especially as the 2.8 has one significant digit, rounding is a big issue in this.

As an example: a dragfactor of 115 will result in a constant of 3.5, and a dragfactor of 125 will result in a constant of 3.5 as well (admittingly, rounding is an issue here). When your angular velocity is 100 radians/sec, with a constant of 3.5, you'll end up with 3,5 m/sec, or a pace of 2:23. With a constant of 3.44, you'll get a pace of 2:25 for the same angular velocity. With a constant of 3.55, you'll end up with a pace of 2:21. So the visible effect is quite limited, even with this very extreme example.