Sunday, February 22, 2009

Week 8: Review

After a short reprieve last week, it was back to the usual winter weather for the majority of this week. Tuesday's run was pretty good, but when Wednesday rolled around the snow came back and made things difficult again. Unfortunately, this meant that my two longest runs (11 and 18 miles) were performed under relatively hostile conditions.

Due to this, I was pretty tired after my Thursday run so I skipped my Friday bike session to help accelerate my recovery for the Saturday and Sunday runs. Regardless, I managed to get all 50 of the scheduled running miles in this week. While it's going to take some time to get my body used to this kind of mileage, I've gone through this before so it's just a matter of time and training ;)

To illustrate the effect that the weather does have on the runs, however, I've posted telemetry from both my Tuesday run (Sunny, clear roads and -2C) and my run earlier today (Heavy snow, 4-6" of groundcover, -6C). In the former, I managed to comfortably complete 10 miles (16.1km) in a little under 82 minutes. The LT portion (5mi/8km) of this run was done at a 4:36/km pace, and the remainder at around 5:00/km (which took a lot of effort not to accelerate beyond).

The latter, however, was a complete contrast to that. I was plowing through 4-6" of snow-cover and it was difficult to maintain a 5:30/km pace. Significant portions were blocked up by drifts and castoff from the roads, which forced me to slow down even more - resulting in an average pace of only 5:48/km. Naturally, the wind and blowing snow didn't help either and made it quite a fight to get to the end.

It didn't appear that the city's sidewalk plows came out at all as I didn't find a single block of dry concrete anywhere along my route despite the fact that the roads were perfectly clear. Unfortunately, with these longer range runs I have to stick to major roads and with the snow drastically reducing visibility I was renascent to spend much time on the asphalt. Either way, I managed to get the full 18 miles in one way or the other ;)
I also took the opportunity to try out a running gel during this session as I figured it may be necessary for this distance. It was certainly a much better experience than the Shot Blox that I tried last year, although it will take some time to get used to their texture. The one downside, however, is that I couldn't get the package open while running and ended up having to stop and take off my gloves to rip off the cap (visible on the plots just after 13km). As such, I'm going to have to work out a better way to handle them in the future.

As for whether or not it was necessary, I'm still not 100% sure. I haven't really run into any problems with energy depletion on my past runs, although this is my first venture into the range where that begins to become a problem. I normally bring about 8oz. of Gatorade (~50 Calories) on my runs, but left it at home this time around in case there were any potential interactions between the two. I burned about 1900 Calories over the duration of the run (largely thanks to the additional effort the snow mixed in), and only took in a single 100 Calorie gel so that would indicate that I should be able to handle an 1800 Cal. deficit.

Weekly Totals:
Running: 81.6km (50.7mi)
Walking: 1.7km (1.1mi)
Cycling: 50km (31.1mi)
Total: 133.3km (82.9mi)

Year to Date:
Running: 538.4km (334.5mi)
Walking: 122.6km (76.2mi)
Cycling: 615.0km (382.1mi)
Total: 1276.0km (792.9mi)

Next week is another step up, totaling up to 54 miles (87km) with my first stab at a 20 miler (32.2km) in a week. The rest of the week is going to be difficult as well, with a 12 miler on Wednesday and another 10 miles (with 6 at LT) on Friday. My legs are pretty tired right now, so this is going to be a pretty challenging week once again.

Upcoming Week:
Mon 50K Cycle
Tue 7mi (11.3K) Recovery w/6x100m
Wed 30K Cycle + 12mi (19.3K)
Thurs 50K Cycle
Fri 10mi (16.1K) w/6mi (9.7K) @ 15K Pace
Sat 5mi (8K) Recovery
Sun 20mi (29.0K) LSD

Tuesday, February 17, 2009

Polar RS800sd Review


A little over a year ago, I picked up a Polar RS800sd to help analyze my exercise regime more accurately and find ways to improve it. Since then, I've been using this tool extensively for my day-to-day workouts and have spent a lot of time getting to know everything about it. In that time, I started a serious running program and progressed to running a half marathon, so I also have a wide range of experience as both a beginner all the way up to an endurance athlete.

When I was first investigating the myriad of choices in running computers last year, I found it extremely difficult to find any detailed information about them. The manufacturers had basic specifications and user manuals to read through, but the reviews that existed either had very little detail or were simply a re-hashing of marketing information put out by the manufacturer. As such, I figured that it would make sense to share my experiences about this tool in a good deal of detail for anyone who might be investigating them at this point in time.

  • Speed and distance recording via S3 footpod, providing accurate readings of both overall and instantaneous measurements.
  • W.I.N.D. heart rate monitor measures heart rate with single beat resolution.
  • Barometric altimeter measures elevation changes more accurately than GPS-based systems.
  • Measurement of running cadence and stride length.
  • Configurable interval training features with alarms for maintaining speed, heart rate, etc.
  • Lap markers can be added manually or automatically to provide context to raw data.
  • Coded 2.4GHz radio communication is less susceptible to interference than the 5kHz system most Polar running computers use.
  • Compact S3 footpod is significantly smaller and lighter than its predecessor.
  • All exercise displays can be customized, with six individual screens that can be selected when in training mode.
  • Monitors mileage of up to two pairs of shoes in the watch itself.
  • IrDA computer interface for uploading training data and configuring watch.
  • Polar ProTrainer 5 software provides tools for analyzing exercise data, and maintains a detailed training log.
  • Can perform a number of simple tests to determine physiological parameters such as VO2Max.
  • Heart rate monitor, footpod and watch all have user-replaceable batteries (albeit different batteries for each).
  • Optional G3 GPS pod can be purchased for use instead of the S3 (RS800CX can use both simultaneously).


There are three basic components to the RS800sd, the watch itself, the S3 footpod and the HRM strap. In addition to displaying the readings, the watch is also the component that collects and stores the data from the other sensors. The S3 footpod attaches to the shoe and is responsible for recording speed, distance, cadence and stride length. Finally, the HRM strap goes around the runner's chest and monitors the electrical activity in their heart, relaying basic heart rate data as well as R-R data (covered later in this article) to the wrist unit. I'll address each of these three components individually in the next few sections.

The S3 Footpod

Probably the most important component of the RS800sd, the S3 footpod can either be attached to the shoelaces (using a provided accessory) or placed inside of a special cavity in some Adidas running shoes. This small unit contains a set of accelerometers which are used to record the movement of the foot in three dimensions. A microcontroller in the S3 then translates that data into speed, distance, cadence and stride length values relayed to the wrist unit via a wireless signal.

Much smaller than its predecessor, the S3 measures a tiny 55x40x13mm and is barely noticeable when worn on the foot. A small battery compartment on the base of the sensor allows the user to replace the CR2430 battery (~$3.50CDN and lasts approx. three months). Other than that, there isn't a whole lot to say about it - it is turned on automatically by the wrist unit when necessary, so it has no buttons or controls of its own.

Footpod (Inertial) vs GPS

When looking at running computers, most products fall into one of two different categories - footpod-based systems like this, and GPS-based systems like those offered by Garmin. As mentioned above, inertial systems use a sensor attached to the shoelaces to monitor the movement of the foot and convert this into speed and distance information. GPS-based systems, on the other hand, capture and store a runner's location every second and calculate speed and distance using that information. Each of these systems has its advantages and disadvantages, and it is important to understand these characteristics when deciding which system to go with.

As footpods are a closed system, they aren't reliant on any external signals like GPS-based systems are. As such, users don't have to worry about buildings or tree canopies blocking reception of GPS signals (or waiting for the watch to acquire a fix in the first place). Further, as footpods are monitoring the movement of the foot rather than the movement of the runner, they can provide accurate readings when running on a treadmill. Finally, since footpods are measuring the gait cycle, advanced systems like the RS800sd can provide additional readings such as cadence and stride length that GPS-based systems have no way to monitor.

The downside is that as footpods are reliant on the movement of the foot they are only useful for tracking running performance. As GPS-based systems simply monitor the absolute movement of the athlete they can also be used to track performance in other sports (eg cycling). Further, as GPS-based systems log the exact route that was taken they also provide context that is not available when using inertial systems. While not critical, it can be helpful to have this information when looking back on previous workouts (eg a slowdown that might appear random on simple curves makes more sense when one sees that they were crossing a street at the time and had to deal with traffic).


The other major difference between Footpod and GPS-based systems is the nature of the error they introduce into the readings. Civilian GPS systems can only generally provide readings accurate to about 15 meters, which is quite a significant error when on foot. Fortunately, the components of this error that effect speed and distance measurements are random in nature so they tend to get averaged out over multiple samples. The result is that these systems can provide accurate readings over the long term (ie kilometer/mile splits, total distance run, etc.), but they aren't very good at providing instantaneous measurements (ie current pace, short intervals, etc.).

Footpods, on the other hand, can provide extremely accurate short and long term measurements but are dependent on accurate calibration. As these systems are reliant on monitoring the movement of the foot, variances in the gait cycle from person to person can introduce error into the readings. Unlike GPS, however, this error is static in nature - that is, it will repeatedly over/underestimate by a fixed percentage for any given gait. Thanks to this it is possible to eliminate much of this error by running a known distance (eg a few laps of a 400m track) and dialing in the difference as a calibration (ie if it reads 1616m after four laps, the watch will take 1% off each future reading).

The complexity with this is that it is dependent on the gait cycle of the runner, so any changes (due to running at significantly different speeds, aging/new shoes, injuries, etc.) can require re-calibration of the system. In my experience, however, once it was calibrated gait variances never added up to an error larger than 1% during the time that I've used it. I have, however, seen variances as large as 4-5% when replacing an old pair of shoes with a new pair without recalibrating (example). As such, it is still wise to calibrate it every once and a while to make sure that it is still giving you accurate readings. This is especially important for inexperienced runners whose technique may be in flux as they gain experience and strength.

In short, inertial systems have the advantage of significantly more accurate instantaneous measurements than GPS-based systems. When kept calibrated, both inertial and GPS-based systems will provide comparable accuracy over longer distances. When calibration isn't done, however, GPS will often provide higher accuracy for many people so they can be a lower maintenance option. Naturally, the decision of which system is better suited to a runner will depend on their priorities.

Cadence and Stride Length

As this is a relatively unique feature of this particular model, I feel that it is important to make a quick mention of this capability. When enabled, the RS8oosd can record and display these two biomechanical properties which can be quite valuable for many runners. Stride length is simply recorded as an average at the end of each lap, so its usefulness is limited. Cadence, however, is recorded every second and can be displayed in real time.

For a beginning runner this information can be quite useful, as slow turnover is a common problem for many people. Providing a real-time readout of this value allows the runner to carefully monitor their cadence, and makes doing drills to increase one's cadence much easier. When looking over the cadence plot after a run this can also be quite instructive as it gives the runner a better idea of how they dealt with it when they weren't actively paying attention.

For more advanced runners, cadence readings can also be a very useful tool. Over the course of a long run, the muscles in a runner's legs begin to get tired and their cadence will begin to involuntarily fall. This forces us to take longer strides to maintain speed which, in turn, increases the load on the muscles and accelerates the rate at which they get tired. By monitoring the cadence value and correcting for this as soon as it begins, endurance and overall performance can be improved. Further, looking over the cadence plots after the run will give the wearer a good idea of exactly when this began to happen and how it progressed, allowing them to make changes to prevent it from happening in any future runs.

Polar W.I.N.D. HRM Strap

The HRM strap is a relatively simple device which, as its name implies, measures the wearer's heart rate and relays that data to the wrist unit. This unit is basically a very simple EKG, which uses an integrated microcontroller to look for the QRS complex (each one triggering the heart to beat once), and sends a signal back to the wrist unit each time it sees it. The watch then averages out the rate at which these beats occur (on a per-second basis) and, if configured to do so, displays them for the user to monitor in real time. Naturally, this data is also stored in the memory of the watch for post-run analysis.

The strap consists of two basic components - a small sensor pod (shown above) and the strap itself (shown below). The former contains the transmitter, the monitoring electronics and a small user-replaceable CR2025 battery (still on my factory cell, but Polar estimates it lasts about two years). The later contains the sensor pads themselves and attaches to the pod via two metal snaps. Unlike many other heart rate monitoring straps, the sensor pads of this unit are made completely out of a soft textile material. As such, it conforms to the shape of the runner's body and will generally record more reliable data than units that use a hard plastic front with metal sensor pads. Further, the soft material is quite comfortable and one doesn't really feel it when running.

Additionally, as the electronic pod that measures the signal is an independent component, athletes can also buy some Adidas shirts that have the sensor pads built right into them. The sensor unit simply snaps onto the front of these shirts, allowing them to leave the strap at home. They can be a bit difficult to find in Canada, however I was able to find them in the US relatively easily. The little sensor pod in the middle of one's chest does look a little funny, however if it is simply used as a base layer with something else on top that isn't an issue.

As with other heart rate monitoring systems, the sensor pads do have to be wet in order to get a good signal. As such, users simply have to moisten them before they head out for their run. Once one gets going the pads will generally stay wet as the wearer begins to sweat. The only caveat with this is that prior to races they can dry out before one starts, but it is relatively easy to take it off and give it a little squirt of water before beginning. Worst case scenario, it only takes about two to three minutes of running to get dry pads moistened so it doesn't really cause a lot of problems.

R-R Data Recording

In addition to the standard averaged heart rate data that similar products provide, Polar's unit also transmits the time between individual beats of the heart. On the wrist unit, this data is used to provide an optional 'R-R Variability' readout that gives users a rough idea of how stable their heart beat is. At heart rates below the aerobic zone the heart rate varies significantly, but as the rate increases it becomes more periodic in nature. As such, this data is apparently used internally by the watch to determine when runners reach certain training zones, as well as to provide a score ('Running Index') for the quality of the overall run. I haven't found this terribly useful while in the field, as you really have to monitor it pretty closely to see how it is changing, but it is not a default reading so it's easy enough to leave switched off.

After the fact, however, the R-R data can be quite useful when analyzing a run. Primarily, it makes it much easier to see exactly when the monitor was having trouble getting a good signal from your heart. The watch/software seems to have a pretty good mechanism of working the noise out of the averaged heart rate curve, so it can look relatively clean even when the signal is weak. The R-R curve, however, makes it relatively plain to see when there was even the slightest problem, as the curve tends to be pretty noisy if the strap slips out of position. As such, it gives wearers a very good way to determine the reliability of the data shown. Further, this information is invaluable when first using the device as it allows users to get a good idea of how well they've positioned everything.

When the sensor is seated correctly, the R-R data gives users a much higher resolution look at the activity of their heart during a run. The curve records each and every beat of the wearer's heart during the session and plots it out in the included software. As such, if a user was running at 180bpm, they will have three readings for every one in the standard curve. This is naturally not necessary for everything, however it can be very useful for tasks such as analyzing exactly how well one recovers after a hard lap during an interval session.

Polar RS800 Wrist Unit

The centerpiece of the system is the wrist unit itself, as it is the component that users will be interacting with most directly. It is responsible for coordinating the other sensors, recording the data and displaying the real-time readings. In addition to these tasks, the watch also contains a barometric altimeter that records any changes in elevation. Naturally, the watch face and five buttons provide access to the multitude of options and settings offered by the watch.

The various sensors can be switched on and off via menus in the watch itself, and countless settings allow you to specify exactly how it handles the data that it receives. With all options enabled, it has enough memory to store about four hours of telemetry. By bumping the recording rate down to 5 second intervals (from 1 second), this capacity is increased to 11 hours. By disabling R-R and altitude recording, you can bring it to over 30 hours. In its most bare bones mode it can store almost 1900 hours of data, but that provides no speed/pace data and a recording rate of one minute so its utility is limited.

As it doesn't need a GPS radio or antenna, the RS800 watch is much smaller than many of the other products aimed at this market. With that said, it is still a pretty big unit when compared to normal watches. Regardless, it is small enough that it's comfortable to wear and it doesn't really get in the way when running. It is powered by a small CR2032 button cell battery, which is user replaceable via a small cover on the back of the watch. While it is not rechargeable, this battery lasts for about a year and costs less than $4CDN, so that isn't a big issue.

The watch is controlled by five buttons, a large red one on the bottom face of the watch and four small buttons around the edges. The top right button switches the backlight on, and when held down during exercise provides some basic options. The bottom right button pauses the exercise session, and a second press stops it. The buttons on the right side navigate up and down through menus, and during exercise they select which data to show (you can scroll through six different screens with three pieces of (user configurable) information on each). Finally, the red lap button initiates workouts, marks laps during exercise and acts as a general 'ok' button in many of the watch's internal menus.

On the rear face of the watch unit is a small opening for the barometric altimeter used by the RS800sd to measure elevation changes. This method is significantly more precise than GPS-based elevation measurements, and can accurately detect changes of a few inches. The downside, however, is that changes in atmospheric pressure (eg a storm system passing through) during a session will incorrectly manifest themselves as changes in elevation. Naturally, as these changes generally occur pretty slowly (relative to changes in elevation) they are relatively easy to correct for after the fact but one still has to keep them in mind. The other caveat is that the watch doesn't appear to record elevation on a per-second basis like the other values (subjectively looks like 3 sec intervals), so short rises/falls might not be recorded completely (ie a runner may crest the hill between two readings).

Finally, the top of the watch face has an IrDA transceiver that allows it to communicate with a computer using the included ProTrainer 5 software. This allows the user to upload the detailed telemetry from their run into the computer, where it is stored (and can be analyzed) in a detailed training log. Further, all of the settings provided in the watch can be configured from the software as well - making it easier to make detailed adjustments. It should be mentioned, however, that the RS800sd does not come with an IrDA adapter for the computer, so most people will have to buy this separately. Polar does sell an adapter for $70, however as the watch uses the standard IrDA protocol you can buy the necessary adapter (minus the Polar logo) for less than $10 at any electronics store.

Polar ProTrainer 5 Software

The main benefit to using a Running Computer like the RS800sd is the ability to analyze the details of one's exercise regime. This includes looking over the telemetry from an individual run as well as looking over aggregate training logs to see exactly how well one is progressing. The watch and sensors perform the function of providing immediate feedback and data recording, however the last piece of the puzzle is the software that takes all of this raw data and allows runners to examine it in a meaningful way. In the case of the RS800sd, this duty is performed by the ProTrainer 5 software package that is included in the box.


When ProTrainer 5 is first launched, the user is presented with a calendar view (shown above) that provides a high-level view of the overall training program. All of the exercises that are uploaded from the watch will be displayed in the calendar, and the user can manually add any sessions where the RS800sd wasn't used (eg swimming). Double clicking on any day in the calendar will present another dialog that allows the user to drill down for more detail, however I'll get back to that a little later.

Each individual exercise entry includes a user defined name, followed by a line containing the sport (R=Running, C=Cycling, W=Walking, etc.), the amount of time taken to complete the session (if available) and the total distance covered. This is naturally very basic information, however it is sufficient to give a rough overview of what was done on any given day. I should note that older versions of the software (which was used to download the two days shown above) placed R-R data (beat to beat heart rate curves) as a discrete exercise (not counting towards totals), however a recent update now bundles that along with the main exercise.

At the right edge of the screen are a set of panels that summarize the overall activity for each training week (see above). This lists the number of exercises, the total time and distance that was recorded during the week and the number of Calories burned. Further, the graphic at the bottom of the panel charts the amount of time spent in each of the five training zones. These zones are user configurable, but by default they are set to 10% increments of the user's maximum heart rate (ie red=90-100%, yellow=80-90%, etc.). This provides a good deal of information about overall training volume and intensity for each week, and allows the athlete to see how well they've been doing at a glance.

As noted above, to get further information the user simply double clicks on a day in the calendar view which results in the above dialog. The first tab provides the ability to keep track of basic information about the day itself. The user can enter a short note, log the weather/temperature, as well as a few basic physiological parameters. If this isn't manually filled out, the data from the previous day will simply be moved forward so it doesn't need to be modified unless something changes.

The exercises performed on the selected day appear as tabs along the top of the dialog, and clicking on them will bring up the above dialog. The left panel contains basic stats about the workout, all of which can be modified by the user if desired (and are relatively self-explanatory). The top-right panel contains the exact time spent in each of the training zones (mentioned above) during the selected session, as well as the total time of the workout.

For exercises uploaded from the watch, the panel in the bottom right corner is generally the most important. A small thumbnail provides a simple view of the data, with a set of buttons underneath allowing the user to pull up more detailed information. Double clicking on the thumbnail brings up a detailed plot of the exercise session.

The detailed plot (shown above) allows the user to examine the telemetry from the session directly. When all of the options are enabled, the chart will contain plots for stride length (green line), heart rate (thin red line with white fill), elevation (thick red line with red fill) cadence (green line) and pace (blue line). The background is colour-coded with the heart rate zones that were configured for the selected sport.

Laps are marked along the bottom axis of the plot, and hovering a mouse over any of them will bring up a summary of critical stats (time, lap time, distance, average pace, heart rate at the end of the lap, average heart rate, average speed, average altitude, total ascent, grade, vertical ascent rate and cadence average). Manual laps (triggered by pressing the red button during exercise), automatic laps (generally used to mark km/mi splits) and phase markers (used for programmed interval workouts) are designated and numbered separately so they are easy to keep track of. Right clicking on the plot and selecting 'Lap Times/Markers' allows you to get more detail, as well as add/remove/modify laps as desired.

Just to the bottom left of the plot is a set of instantaneous data readings (time, heart rate, Calorie rate, pace, distance, cadence, altitude, ascent and descent) corresponding to the selected point in time. When first opened, this will reflect the starting point of the session but clicking anywhere on the plot itself will move the cursor to provide you with data from that point in time. Pressing the left and right arrows on the computer will move forward/back in one second intervals (or whatever sampling rate was selected) so the user can step through the session.

To the bottom right of the plot is a grade summary for the session, outlining how much of the session was spent on flat ground, ascending inclines and descending. Both time and distance figures are provided for all three, in both absolute and percentage terms. Naturally, the detailed elevation plot is generally more useful but this information can be handy for route comparison purposes.

Finally, the very bottom of the plot includes a summary of the overall workout. This section provides exercise duration, distance, heart rate (avg/max), pace (avg/max), cadence (avg/max), running index and ascent. The running index field specifies a synthetic score corresponding to the overall quality of the session. Polar isn't explicit about exactly how this is calculated, however it seems to correlate tightly to the ratio of speed and heart rate. I'm not entirely sure how accurate it is, however, as the score of 61 that I usually get is supposed to correspond to a 1:30 half marathon which is about twelve minutes faster than my recent race.


Polar ProTrainer also offers the option to customize many areas of the wrist units operation, as well as to create and upload custom exercises. When the user goes to the Tools->Edit Polar Product Settings menu, a tabbed dialog (shown below) is provided that offers a plethora of options.

The first tab (General) provides an overview of the wrist units current state. It includes a summary of the amount of available memory (as well as buttons to manage entries), a readout of the level of charge remaining in each of the three batteries (watch, HRM and footpod) as well as overall totals. The second tab allows basic watch functions to be configured (time and date, adding alarms, etc.) and is relatively simple so I won't cover it here.

The User tab allows athletes to enter basic information about themselves, that will be utilized by a number of algorithms within the wrist unit and the software to calculate values such as calories burned and the running index. All of these values can be updated on the wrist unit as well, however they are much easier to enter in bulk with a proper keyboard and mouse.

The Product tab controls a number of aspects about how data is recorded and displayed by the wrist unit. The Sports Zones button allows the user to customize the heart rate zones that are used for both the reports and for certain real-time displays that the wrist unit can present to the runner during a session. Finally, the 'Customize Exercise Displays' button launches another dialog that allows the user to control which data is displayed when the footpod is disabled (for some reason a different button on the 'Run' tab controls the displays used when it is available).

The 'Run' tab is likely one of the most significant panels available within this dialog, as it controls a lot of the central functionality offered by the running computer. It allows you to enable/disable the footpod, select whether the wrist unit should display pace (min/km) or speed (km/h), modify the calibration factor and a number of other variables. Most significantly, however, is the 'Customize Exercise Display' button, which allows the athlete to control how data is displayed on the watch during exercise.

Each of the six columns in this dialog represent one of the 'pages' of information that the wrist unit will be displayed. The watch face displays one page at a time, and the buttons on the right edge of the wrist unit allow the user to cycle through the pages as desired. Each of the pages can be disabled if desired, and what is displayed in each of the three available slots can be selected by the user. For reference, the available information is as follows:
  • Altitude - A simple readout of the watches current altitude in feet or meters.
  • Ascent* - The gross distance that the runner has ascended during the current session.
  • Cadence* - A readout of the number of steps that the runner/walker is making per minute.
  • Calories* - The total number of calories consumed during the current exercise session.
  • Countdown Guide* - Displays remaining time or distance in the current exercise phase. When running in a phase without a target (or running a 'free' exercise session), it simply displays total time.
  • Distance - The total distance covered during the current exercise session.
  • Exercise Time - The total amount of time accumulated in the current session.
  • Heart Rate - The current heart rate reading (in beats per minute) that is provided by the Wearlink strap.
  • Lap Distance* - The total distance accumulated in the current lap.
  • Lap Time - The total amount of time accumulated in the current lap.
  • RR Variation* - The average variance (in milliseconds) between individual heart beats.
  • Speed/Pace - A readout of the instantaneous pace (or speed) of the runner/walker.
  • Target Zone* - A readout of the current training zone that the wearer is in.
  • Time of Day* - A simple clock readout to provide the current time of day.
  • Zone Pointer* - A graphical readout of where the current heart rate falls within the training zones.
* The entries above listed with an asterisk are only available in the top two rows of the display. The lower row offers a smaller number of selections as it uses a fixed element LCD rather than the bitmapped display used by the upper two.

This provides a good deal of flexibility so that the user can select exactly what information they want available and how to organize it. The above fields cover pretty much anything that runners will likely want access to, and the availability of six selectable screens allows that information to be retrieved without much difficulty. The default configuration works relatively well, however no layout is perfect for everyone so the ability to customize it to this level is quite helpful.

The next major tab allows the user to prepare preprogrammed exercise sessions which the wrist unit will guide them through. The RS800sd can store up to ten exercises at any given time, and if more are necessary the software allows users to save them to disk and call them back when necessary. When an open slot is available, the three buttons across the top of the dialog allow the user to create new sessions - either by a simple zoned exercise (with a single set of targets for the entire session) or a complex phased exercise (with a number of phases, each with their own individual targets).

When the user creates a phased exercise, the above dialog is used to build and prepare each of the phases. In this case, an 8 mile run with ten 100m strides is planned with heart rate targets programmed into each phase. The panel at the bottom of the dialog allows the phases to be prepared and re-ordered, with the following dialog provided to allow customization of each individually:

Using this dialog, the athlete can name the phase, specify how it is initiated (automatically when the previous phase ends or manually by pressing the red button), the duration of the phase (manual, time, distance or when a specific heart rate is reached) and what sort of targets the user would like to aim for (none, preprogrammed training zones, manually specified heart rate range or a pace range). Further, the bottom panel allows the user to specify that one or more phases is repeated a number of times which is helpful for interval training sessions.

Once the phases have been built up, the diagram near the top of the exercise dialog illustrates the overall workout as well as the specified targets. The top right of the dialog provides an estimate of how long the run will take, as well as how much distance will be covered. Once complete, the exercise can be given a name and short description, then uploaded to the watch.

The next time an exercise session is started, the user can then select this exercise and the wrist unit will walk them through the session. When the specified phase is complete, the watch will beep to signal the user to change to the next phase (the name of which is shown on screen momentarily). If targets have been set for the current phase, the wrist unit will also sound an alarm whenever the runner falls outside of the specified range (this can be turned off during the session if desired). After the run is complete, the watch will provide a summary of a number of critical details (time spent, distance covered, average pace, heart rate, cadence, etc.) broken down by phase. Naturally, phase changes are also marked on exercise plots when examining them in ProTrainer.

That covers the main functionality of the ProTrainer 5 software. The package does offer additional features like entering an exercise plan and generating detailed reports of your progress, however this article is already getting pretty long so I'll stick to the basics outlined above. Regardless, the functionality discussed above covers most of the core features that most people will use in a good deal of detail.

G3 GPS Sensor

In addition to the S3 footpod, the RS800 can also be used with a GPS device called the G3 that provides distance and speed data. This model is generally purchased as part of the RS800G3 bundle, however RS800sd users can also purchase the sensor separately and use it instead of the S3 footpod that came with their watch. Note, however, that the RS800, RS800sd and RS800G3 can only use one of these sensors at any given time so the user must select which mechanism they would like to use (although this is a simple menu item, so it can be done on a session-by-session basis). It is also important to note that these models do not store a tracklog of the route like other GPS-based products, instead simply storing the calculated distance and speed data.


Recently, Polar has released a new revision of this training computer offering a few additional features. The core design and operation of the new model is basically identical to the RS800sd, so most of the above is relevant to this model as well. The major differences are as follows:
  • Compatibility with Polar W.I.N.D. speed and cadence sensors for bicycles allows the RS800CX to work for both running and cycling workouts.
  • When used with the G3 GPS sensor, can now record a map of the route taken during the session.
  • Ability to connect to both the S3 footpod and G3 GPS radio at the same time. This provides the route recording capability of GPS and combines it with the accuracy of the footpod mechanism.
  • Automatically falls back to alternative tracking if the connection to the primary sensor is lost (eg if the footpod battery dies, GPS will take over).
  • Ability to monitor up to four pairs of shoes (vs. two in the RS800sd).
  • Allows the ability to append new session onto an existing session when the later is started shortly after finishing the former.
  • A number of other small refinements to the overall design of the system.
As such, the primary advantages of the RS800CX over the RS800sd is for multi-sport athletes who need the addition of the cycling functionality. As I'm currently both a runner and cyclist, the RS800CX would have been a better match had it been an option when I was buying it. For someone who simply runs, there isn't really a lot of material difference between the models (although if buying new you might as well buy the CX).

One thing that should be mentioned is that some Polar distributors appear to be offering the option of upgrading the RS800sd to the RS800CX. There is unfortunately no company-wide policy on this front, however in countries where the service is being offered existing customers can send in their RS800sd and pay to have the electronics replaced to make it the functional equivalent of the new RS800CX. I haven't really looked into what the situation is in Canada as of yet, however I likely will give them a call at some point to examine the possibility.

Technical Details:

As the RS800sd uses wireless transmitters in each of its components, Polar is required to file documents with regulatory bodies in countries where it is sold. In many countries, these filings become part of the public record and thus can be pulled up by those interested - providing internal photos and details of the devices that aren't available from other sources. The American Federal Communications Commission does just this, and fortunately their filings can be pulled up via a simple web search. As such, for those technically inclined readers interested in such detail the appropriate filings for the components are as follows:

RS800 Wrist Unit: INWK1
S3 Footpod: INWJ9

Interestingly, when pulling up the above URLs I also found this entry (INWR7) detailing a bluetooth-based HRM module similar to the WINDLink (named Windlink+ in the filings). Currently Polar uses a proprietary 2.4GHz protocol for communication between their sensors, but moving to an open standard like this would make it easier for third-party products to add compatibility for these products (much like the open ANT+ protocol used by Garmin).


While the RS800sd is an extremely powerful tool, like anything in the market it does have a few rough edges. These aren't really huge issues, and many of them likely fall into the nitpick category, but for completeness sake a quick summary is as follows:
  • The RS800sd uses an integral wrist strap instead of a standard strap coupling, which means that when it wears out you have to send it back to Polar for repair. Further, the lack of a user-replaceable strap means that one has no option to go with an alternative design.
  • The beep signal sounded by the wrist unit when an auto-lap occurs is the same as the phase change signal, potentially making it difficult to differentiate them. When an automatic lap is triggered during a phased exercise, it is easy to get confused and interpret it as signalling the end of a phase. Ideally, it would be nice to have a number of different signals that could be selected for different types of phases.
  • Given the rapid decline in flash memory prices, it would be a significant benefit to have more than four hours of capacity with all features active. While this is generally enough for any individual running session, more memory would mean that it wouldn't have to be synchronized with the computer as often (especially significant when travelling). It was understandable when the RS800sd was released years ago, but it would have been a nice step to add more memory to the updated RS800CX that was released late last year.
Things that Could be Improved

Aside from the direct issues listed above, there are a few refinements that would make the RS800 a better product. These items aren't really faults with the design itself, but instead little things that could be done a little better:
  • When configuring phased exercises, it would be beneficial to be able to specify a total distance value in the duration field of each phase. On many occasions it would be beneficial to have variable length phases (eg the recovery portions of an interval workout) but still maintain a fixed distance for the entire run. At this point, users can only specify a distance or length of time for each phase in isolation so getting a fixed total distance means that all phases need to be distance-based.
  • Switching the IrDA connection to a radio frequency based option (possibly Bluetooth-based given the above filing) would make for a more elegant solution. While IrDA ports may have been common when the RS800sd was released, it has largely been supplanted by RF technologies. Aside from more people having the requisite hardware already, using RF would remove the annoyance of having to maintain a line-of-sight connection when uploading.
  • As noted above, changes in barometric pressure can manifest themselves incorrectly as changes of altitude. Incorporating a second barometer and a simple data logger into the infrared interface (which stays at your desk) would allow the software to completely eliminate this error. Given the extremely high precision of the built-in altimeter, this would yield a nearly perfect elevation record.
  • Adding a mechanism to allow calibration factor to be changed on-the-fly when it is apparent that it is off. Currently, the calibration procedure requires you to initiate it before running the known distance. In some cases, however, it's easy to change shoes and forget to update the calibration. When this happens, it would be nice to simply select a menu item at a known point and change the calibration on the spot.
  • Furthering the above, when using both the S3 and G3 it would be nice to offer an automatic calibration option that will detect any error. While manually calibrating the footpod would be more accurate, a system like this would provide most of the benefits of an inertial system without the user having to worry about maintaining calibration.
  • Adding electronics in the footpod to detect the activation signal from ChampionChip timing pads and drop a lap marker would be extremely helpful when racing. Every finishing photo that I have shows me looking down at my wrist to press the stop button, and adding a simple feature like this would allow me to let the device do all the work. While it would add some cost to the product, the system would only have to detect the signal rather than interpret it.
  • If at all possible, it would be nice to use the same type of battery in all three of the sensors that work within this system. Right now, each of the sensors uses a slightly different button cell battery so it's a bit of a pain to pick up replacements. This isn't really a very big thing, but it's just a matter of a more elegant configuration.


The Polar RS800sd is an extremely powerful tool, and with the associated software it can provide Runners with a lot of critical information to improve their workout regime. It records more data than any other running computer on the market, and does so using some of the most advanced technology available. The footpod-based system does require a certain degree of attention on the part of the user to keep it accurate, but for those willing to expend the effort it yields the most accurate results currently available.

Monday, February 16, 2009

Week 7: Review

Another excellent week weather-wise, making it an extremely pleasant for running. Tuesday and Wednesday saw a lot of rain, and Thursday was pretty windy but I was fortunate enough that my runs happened to fall into times where the weather died down a bit. By Friday, the mixture of rain on the previous days, the sun and the warmth we've been getting had melted most of the accumulated snow and the pavement was bone dry for the first time in a while.

As I had some stuff to do on Thursday, I ended up deferring my 8 miler to Friday so this worked out quite well. After about 6km of running on dry pavement and nearly perfect weather, I figured that I'd give a local trail a shot to make the most of it. During the summer this path is a great running route as it has a good set of lightly rolling hills, decent scenery and the soft gravel surface is a lot easier on the legs. During the Winter, however, it can be a bit hit and miss as it isn't plowed or salted (it passes through a conservation area) so the snow and ice quickly builds up.

Unfortunately, while the roads had been cleared by the conditions on the previous few days the trail was still pretty well covered. The first kilometer or so was pretty good, but after that point there was a good deal of thick snow as well as some icy patches in the low-lying areas that forced me to slow down significantly. As the only other way to get out was to head back the way I came, I decided to continue on. While it was a bit of a fight to get through it, with a little effort it wasn't too much of an issue. Naturally, my pace wasn't spectacular in this segment (averaging about 6min/km) but there wasn't much I could have done about that.

About 2.5km into the trail, however, I came up on a descent and when I looked down I realized that the river had spilled its banks from all of the run-off. As such, the next segment of trail was flooded by about two feet of water and going forward wasn't an option. Given the terrain behind me, however, it would have been a pain to double back. As such, I just figured out the bearing that I had to head to get back to the road and took off through the forest. The terrain between the trees was much better than the trail, as the snow hadn't been compacted by hikers and there was a lot less ice to deal with. Naturally, navigating between the trees, rocks and exposed roots was a little challenging at first, but once I got the hang of it I was able to maintain a pretty stable ~5:00min/km pace.

Once I got out of the forest and back to the sidewalk things naturally picked up, and I finished the scheduled distance with little trouble. Either way, it certainly wasn't ideal but it was nice to inject something a little different than the usual routine into the mix. I'll likely wait until the summer to try the trail out again, however, as proper training requires consistency and trail running doesn't really lend itself to that ;)

Other than that, things largely went as planned this week. I did have to move things around a bit for scheduling issues (aside from Thursday, I also swapped the Sunday and Saturday runs around). Given the discomfort in the leg last week, I held off on the walks again as their training value is limited so I'm more prone to give them up. Things were largely healed up but got a little aggravated again by the uneven terrain on Friday, but the discomfort is pretty much gone today.

Weekly Totals:
Running: 61.3km (38.1mi)
Walking: 1.2km (0.75mi)
Cycling: 80km (49.7mi)
Total: 142.5km (88.5mi)

Year to Date:
Running: 456.8km (283.8mi)
Walking: 120.9km (75.1mi)
Cycling: 565.0km (351.1mi)
Total: 1142.7km (710.0mi)

With the recovery week in the books, I'm now on to the second mesocycle of this training program which is intended to build my Lactate Threshold. As such, I'm scheduled for a total of 50 miles (80.5km) this week and will hit 54 miles (87km) the week afterwards. I'm also looking at my first 18 miler (~29km) on Sunday, which should be an interesting experience.

Upcoming Week:
Mon 50K Cycle
Tue 10mi (16.1K) w/5mi (8K) @ 15K pace
Wed 30K Cycle + 4mi (6.4K) Recovery
Thurs 11mi (17.7K) Steady
Fri 50K Cycle
Sat 7mi (11.3K) w/8x100m
Sun 18mi (29.0K) LSD

Aside from that, one thing I will have to do is start playing around with nutritional intake during my runs. I flirted with it a bit during half-marathon training, but as it didn't really provide much benefit there I didn't really push very hard. As I approach the 20mi mark, however, it is likely to become significant so I will have to take in some calories during my runs. Given the plethora of options on that front, I'm going to have to do some experimentation to figure out what works well for me.

Sunday, February 8, 2009

Week 6: Review

The last three days have been pretty much textbook perfect running weather, and have done wonders in decimating the massive snow piles that have accumulated over the last month or so. There was a bit of cold and snow in the early week, but once we got through that it got up to a high of 8C on Saturday which is quite the reprieve from what we've been dealing with lately ;)

Unfortunately, after the 15 miler last Sunday I began to have a bit of discomfort in my upper right leg. It was likely caused by the fact that I've been spending a lot of time lately running on the roads rather than the sidewalks. As I tend to religiously stick to the left side of the road for safety purposes (going against vehicular traffic), the camber of the road creates an imbalance in my stride (as my right foot is repeatedly hitting ground higher than my left, it hits the ground with much greater force than usual). While this isn't a big deal in and of itself, when the body is continually pounded like this for an extended period it can cause problems. This was likely exacerbated by running long stretches on the larger roads, as the camber at the edge of a four lane road is generally much more aggressive than that of a two-lane residential street.

In the interests of letting it heal, I skipped the cycling session on Monday. By Tuesday it was improved, but there was still a bit of tightness there. It wasn't really bad enough to throw off my running schedule, so I went out for the Tuesday run anyway. Fortunately, it didn't really cause much issue with my running and actually improved a bit over the duration of the session. It was still a bit uncomfortable to walk on it, however, so I skipped my walk that afternoon to let it get better.

On Wednesday, I elected to give the scheduled 30K ride a shot to see how it would work. As cycling doesn't have the impact of running or walking, this worked quite well and any discomfort completely disappeared while I was riding. It even stayed away for a few hours after the ride, although it did start to come back afterwards. I did my 5mi recovery run at the Running Room as usual, and like the day before it wasn't really an issue.

As the previous runs went well, I did my 10 miler on Thursday but skipped the scheduled walk in case that would help. On Friday, I did my ride as usual (although I scaled it back to 40km due to scheduling issues) with the same results as I had on Wednesday. At this point, it was significantly improved but again wasn't completely gone so I elected to skip the 5 miler on Saturday morning. This does mean that I lost a bit of running mileage, but I figured with the long run scheduled the next day it was worth it to ensure that I'd be able to make the most of it. It was certainly a shame as the weather yesterday was incredible for this time of year, but it's better to be safe than sorry on this front.

Which brings us to this morning, where I set out for my first run with a chunk at marathon pace. The schedule called for 16 miles with 10 of them done at marathon pace. I'm still not 100% sure what that should be, as my fitness has improved significantly since September so my previous race times don't really mean a lot at this point. As such, I looked over my data for the last little while and settled on giving a 4:45/km pace a shot (which works out to a little over 3h20m for a marathon distance).

Given the issue with the leg, I carefully plotted a course that would allow me to stick with the sidewalks for the majority of the run. Thankfully, the heat over the last few days had melted a significant amount of the snow so the problems that I ran into last week wouldn't really be an issue this time around. What I didn't properly anticipate, however, is that as the temperatures had dropped below freezing overnight there were a good number of icy patches that I had to deal with. As the snow banks didn't completely melt, in many places they formed a dam on either side of the sidewalk trapping the water and allowing it to refreeze.

Fortunately, the clear portions were pretty much bone dry so these ice patches were easy to see, however I had to slow down significantly when passing over them. In some areas (especially on hills, where the water ran down the sidewalk) these patches were quite large, so they had a significant negative effect on my ability to maintain my desired pace. For a normal long run this wouldn't be a big issue, but when trying to maintain race pace it made it difficult to keep my heart rate up where it needed to be.

Despite these issues, however, the run went quite well. I averaged an overall pace of 5:05/km, with the 10mi segment at 4:52/km. Running the raw numbers through a spreadsheet however, my median pace worked out to a perfect 4:45/km so when the obstructions are factored out I stayed pretty close to where I needed to be. Unfortunately, my heart rate was much lower than it should have been during this segment and it's hard to tell if that was the fault of (a) the periodic slowdowns or (b) if the pace was a little slower than it should have been. Either way, I'll have to do some more experimenting with different speeds to figure out where I need to be.

As I missed the 5 mile recovery run on Saturday, my mileage is a little short this week (40mi vs. 45mi). Naturally, the two walks and cycling session that I skipped also mean that those numbers will be a little lower than usual. Either way, the leg is feeling a lot better today and as next week is a recovery week hopefully it will get back to 100% soon.

Weekly Totals:
Running: 64.8km (40.3mi)
Walking: 2.3km (1.4mi)
Cycling: 70km (43.5mi)
Total: 137.1km (85.2mi)

Year to Date:
Running: 395.5km (245.8mi)
Walking: 119.7km (74.4mi)
Cycling: 485.0km (301.4mi)
Total: 1000.2km (621.5mi)

As noted above, the upcoming week is a recovery week so mileage will drop down a bit (to 37mi) before jumping up to 50 miles the week afterwards. Other than the bump in the road this week, the training program has been going quite well so far and as long as I'm a little more careful things will hopefully continue along that path ;)

Upcoming Week:
Mon 50K Cycle
Tue 8mi (12.9K) w/10x100m
Wed 30K Cycle + 5mi (8K) Recovery
Thurs 8mi (12.9K) Steady
Fri 50K Cycle
Sat 4mi (6.4K) Recovery
Sun 12mi (19.3K) LSD

In other news, I'm now officially registered for the Mississauga marathon so I'm committed to seeing this program to the end one way or the other. There is still a lot of training to be done which is only going to get harder and harder, so I figure that locking myself in will help to keep me motivated ;)

Sunday, February 1, 2009

Week 5: Review

The weather this week was pretty brutal, with heavy snowstorms on Wednesday and Thursday that made running quite difficult. On Wednesday, we ended up cutting the 5mi recovery run a bit short (5.7km/3.5mi) as the 6' of slush we were pushing through made it somewhat pointless to keep going. Accumulation was fortunately not as bad on Thursday, however it was snowing heavily while I was doing my 10 miler which made the session somewhat unpleasant.

Fortunately, this weekend has been much better. Yesterday was a little cold, but other than that the roads were mostly clear and the snow didn't come until the late afternoon so it was a non-factor. As such, I increased my distance from the planned 4 miler to a 10km in order to make up for the lost mileage on Wednesday. The mixture of the nicer weather combined with the clear roads made this a much more pleasant run than I've had for a while.

Further, the weather today was as close to perfect as it gets around here. A nice warm 6C, little to no wind and sunny skies let me keep most of my winter gear at home and run in much more comfortable clothing than normal for this time of year. After running for a couple of months now wrapped up in winter clothing, it was certainly refreshing to be free of the toque, mittens and neck warmer ;)

Unfortunately, while the roads were in near perfect conditions many of the sidewalks weren't. While this isn't an issue on my shorter runs, as I can generally just switch to the road when necessary, with the longer runs I need to use some 80km/h arterial roadways that make it dangerous to leave the sidewalks. As such, portions of this run were much harder than usual (ie the choppy portions of the cadence/pace plots below), and in some parts I had no choice but to go out on the road if I wanted to keep running. Either way, once I got around the obstacles it worked out to be quite a nice day for a run, and plodding through the snow makes for good strength training ;)

With respect to my other sports, I managed to get most of my walking in this week, as well as all but one cycling session. Hopefully things will slow down a bit next week, and I'll be able to pick up the slack on this part of my training ;)

One other upside this week is that my copy of Pfitzinger/Douglas' Advanced Marathoning 2nd edition finally showed up. For the last few weeks I've been using a copy of the first edition borrowed from the library, so there will be some minor changes in my schedule as the new book makes a few tweaks in the plan. The books are mostly the same, with a few minor differences (eg expanded supplementary training section, small tweaks to the schedules and the addition of an 85 mile/week program) but I figured with the update coming out it made more sense to get a copy of the new one.

Weekly Totals:
Running: 68.8km (43mi)
Walking: 32km (20mi)
Cycling: 80km (50mi)
Total: 181.8km (113mi)

Year to Date:
Running: 330.7km (205.5mi)
Walking: 117.4km (72.9mi)
Cycling: 415.0km (257.9mi)
Total: 863.1km (536.3mi)

For the upcoming week, I'll be adding my first long run with a section at race pace. That is likely going to be a good challenge to see how well I've progressed over the last few weeks, although there will likely be some learning involved. The one thing I will have to do, however, is find a better route for that run as plowing through snow will be a much larger problem when higher speeds are necessary.

Upcoming Week:
Mon 50K Cycle
Tue 9mi (14.5K) w/5mi (8km) at LT
Wed 30K Cycle + 5mi (8K) Recovery
Thurs 10mi (16.1K) Steady
Fri 50K Cycle
Sat 5mi (8K) Recovery
Sun 16mi (25.8K) LSD w/10mi (16.1km) at Marathon Pace

As the prices for the races that I'm looking at go up this month, I'm in the process of registering for them now. I registered for Around the Bay a few weeks back, and added the Achilles 5K last night. I'll probably be officially registering for the Mississauga Marathon in the next couple days as well, as it's price goes up and having that one locked in will likely help to motivate my training that much more. I'm still flirting with a few other options, as well as trying to work out what I want to do after this training cycle.