Robert Chung
October 13th 04, 07:58 AM
Amit wrote:
> i used one for a while, i'd recommend getting a tacx basic, excel or
> flow for serious training (the new models might have better
> calibration features).
The Basic appears to have been replaced in the Tacx line-up with the Flow.
AFAIK, the roll-down calibration feature from the Flow hasn't been added
to the Excel model. I have no idea whether it's on the I-Magic. As little
as a quarter of a turn on the roller adjustment knob makes an observable
difference to the power readings, so the Flow's roll-down calibration is
pretty handy.
> the most important feature for me is being able to control your
> workload and the cheaper tacx (ergo) trainers let you do that.
Last Spring, some guys on the Wattage list and I did some analysis of the
Flow and the Excel (but not the I-Magic). In both power-measuring mode and
in power-controlling mode (ergo mode), the Excel seemed to do pretty well.
Here's a plot of a 45-minute training session comparing the Excel to the
SRM:
http://anonymous.coward.free.fr/rbr/srm-tacx.png
I've found that my Tacx Flow is quite consistent in power mode. This graph
http://anonymous.coward.free.fr/rbr/flow.png
shows the indicated power at a given speed with four different scale
factors. The point of the graph is that in power mode the Flow seems to
measure wattage consistently, and I believe that if I had an SRM or Power
Tap against which to compare I could find a scaling factor that would
"envelope" the correct power curve reasonably well. That's not to say that
the fit would be perfect; there have been reports (for which I haven't
seen the data, so I can only report it anecdotally) that the Tacx power
curve scales up a bit slower than real power. Still, I was moderately
reassured that the Flow appears to be in the right ballpark.
Alas, despite sharing the same load generator across all three models (the
Flow, the Excel, and the I-Magic) the Flow's ability to control load in
ergo mode appears to be much weaker than for the Excel.
Another guy tried a series of experiments with a Flow in ergo mode. Here's
his description:
http://home.comcast.net/~rwwells/Tacx/DOEFlow.html
from which I calculated this error pattern for the Flow:
http://anonymous.coward.free.fr/rbr/tacx_correction.png
The chart may be a bit hard to interpret so I should probably explain. The
idea of an ergo is that you set the wattage you want (say, 250W) and off
you go (Ron Popeil would say, "set it and forget it"); the ergo will match
the load so that whatever speed you ride, you'll be generating 250 watts.
So in an ideal world you'd want the percent error to be independent of
your speed. The chart plots the error against kph (or, in this case, the
ratio of the wattage setting to the speed), and if the ergo worked well
the pattern would have zero slope.
The observed pattern doesn't have zero slope.
Instead, it says that in ergo mode the Flow's electronic braking system
fails to increase the load by the appropriate amount to keep power
constant as wheel speed increases. If you look only at the black dots
(where the scale factor is set at 100), it says that if you set the ergo
for a constant 250W, the Flow will only have zero error if you're riding
at 23kph. At 28kph, the Flow is only able to produce enough braking force
to generate an actual load of about 200W (about a 20% error). At 36kph,
the Flow was actually producing about 40% less load than it should have.
The scaling factor moves the curve up and down, but doesn't fix the basic
underlying problem.
The good news is that the error is stable and predictable. The bad news is
that the slope of the error is huge, which sort of defeats the entire
intent of the ergo feature.
I don't know whether the I-Magic controls the load more like the Excel or
more like the Flow, but if you're interested in how realistic the "virtual
reality" features of the I-Magic are, I'd think you'd want to know.
Cool graphics, though.
--
Trends and bias in 2004 election polls:
http://anonymous.coward.free.fr/polls/pollbias.html
> i used one for a while, i'd recommend getting a tacx basic, excel or
> flow for serious training (the new models might have better
> calibration features).
The Basic appears to have been replaced in the Tacx line-up with the Flow.
AFAIK, the roll-down calibration feature from the Flow hasn't been added
to the Excel model. I have no idea whether it's on the I-Magic. As little
as a quarter of a turn on the roller adjustment knob makes an observable
difference to the power readings, so the Flow's roll-down calibration is
pretty handy.
> the most important feature for me is being able to control your
> workload and the cheaper tacx (ergo) trainers let you do that.
Last Spring, some guys on the Wattage list and I did some analysis of the
Flow and the Excel (but not the I-Magic). In both power-measuring mode and
in power-controlling mode (ergo mode), the Excel seemed to do pretty well.
Here's a plot of a 45-minute training session comparing the Excel to the
SRM:
http://anonymous.coward.free.fr/rbr/srm-tacx.png
I've found that my Tacx Flow is quite consistent in power mode. This graph
http://anonymous.coward.free.fr/rbr/flow.png
shows the indicated power at a given speed with four different scale
factors. The point of the graph is that in power mode the Flow seems to
measure wattage consistently, and I believe that if I had an SRM or Power
Tap against which to compare I could find a scaling factor that would
"envelope" the correct power curve reasonably well. That's not to say that
the fit would be perfect; there have been reports (for which I haven't
seen the data, so I can only report it anecdotally) that the Tacx power
curve scales up a bit slower than real power. Still, I was moderately
reassured that the Flow appears to be in the right ballpark.
Alas, despite sharing the same load generator across all three models (the
Flow, the Excel, and the I-Magic) the Flow's ability to control load in
ergo mode appears to be much weaker than for the Excel.
Another guy tried a series of experiments with a Flow in ergo mode. Here's
his description:
http://home.comcast.net/~rwwells/Tacx/DOEFlow.html
from which I calculated this error pattern for the Flow:
http://anonymous.coward.free.fr/rbr/tacx_correction.png
The chart may be a bit hard to interpret so I should probably explain. The
idea of an ergo is that you set the wattage you want (say, 250W) and off
you go (Ron Popeil would say, "set it and forget it"); the ergo will match
the load so that whatever speed you ride, you'll be generating 250 watts.
So in an ideal world you'd want the percent error to be independent of
your speed. The chart plots the error against kph (or, in this case, the
ratio of the wattage setting to the speed), and if the ergo worked well
the pattern would have zero slope.
The observed pattern doesn't have zero slope.
Instead, it says that in ergo mode the Flow's electronic braking system
fails to increase the load by the appropriate amount to keep power
constant as wheel speed increases. If you look only at the black dots
(where the scale factor is set at 100), it says that if you set the ergo
for a constant 250W, the Flow will only have zero error if you're riding
at 23kph. At 28kph, the Flow is only able to produce enough braking force
to generate an actual load of about 200W (about a 20% error). At 36kph,
the Flow was actually producing about 40% less load than it should have.
The scaling factor moves the curve up and down, but doesn't fix the basic
underlying problem.
The good news is that the error is stable and predictable. The bad news is
that the slope of the error is huge, which sort of defeats the entire
intent of the ergo feature.
I don't know whether the I-Magic controls the load more like the Excel or
more like the Flow, but if you're interested in how realistic the "virtual
reality" features of the I-Magic are, I'd think you'd want to know.
Cool graphics, though.
--
Trends and bias in 2004 election polls:
http://anonymous.coward.free.fr/polls/pollbias.html