… … … This compendium is rather big (±24MB), please take your time while it is downloading ... … …
If, for easier navigation, you desire a clickable Table of Contents in a separate frame, then use this temporary link.
Fora on SUUNTO t6:
□ Yahoo WriststopTrainers forum: http://health.groups.yahoo.com/group/WriststopTrainers/
□ SUUNTO Training world discussion forum: www.suunto.com/training
□ T6 Usergroup forum in SuuntoSports (becoming obsolete for obvious reasons): http://suuntosports.com/Default.asp
WHAT ABOUT the new release of STraM?
SUUNTO launched a new release of STraM (version 2). In the ReadMe.txt the main improvements w.r.t. version 1.3.4 are enlisted:
· Automatic data conversion from previous version while installing the new version.
· Automatic full / update version detection. One installation file for both cases.
· New improved analyzing module (FirstBeat body model) with very fast analyzing.
· Increased startup time.
· Support for new activity levels
· Automatic activity level. Algorithm changes slowly, but should adjust every users activity level into correct status within a few week or month. Depending on the starting level.
· New printing layout. Multiple graphs available.
· New XML export. Several logs may be included in the same file.
· New R-R graph.
· Log splitting. Splits logs to two or three separate logs and removes the original to keep the summary values correct.
· Downloading wizard. Optional downloading interface to your device.
· Compare your plans and logs from a calendar with analyze view.
· Calculate Day / Week / Year summaries of the values your desire with the analyze tool.
· Online summary calculations from a single log using a graph with movable selectors.
· Internal architectural improvements.
· Support for Suunto Monitor
Unfortunately, STraM2 is overwhelmed with bugs, of which some are considered as critical (damaging your log data) or as of high importance (useability is highly affected) .
Therefore I decided to pool all the bug reports in a separate excel sheet ‘STraM version 2 (update month day, year).xls’. You will find it here (Files section, folder ‘STraM version 2, what about it? ‘).
In that excel sheet you will find 4 TABS:
· BUGS: all the bugs reported are described and assigned to an impact level (critical, functional High, functional Middle, functional Low, cosmetic). Advices (workarounds) and bug status are reported as well;
· IMPROVEMENTS: a description of the noticeable improvements are enlisted as well;
· OTHER CHANGES: a list of other changes is also given;
· WISHLIST: a list (of most wanted) new features is reported as well.
People are encouraged to post new bugs, improvements, other changes or whishes in one of these two user groups.
This document discusses all aspects of the SUUNTO t6:
1. it explains the basic working principles of the hardware and the training software (Chapter 2);
2. it focuses on managing accurate speed and distance recording and calibration for running and cycling (Chapter 4, 5 and 6);
3. because the data measurement (heart rate, speed, distance) is expected to be accurate, a separate chapter is included to tackle weird measurements (Chapter 11.2);
4. it explains all the programming features (Chapter 7);
5. it focuses also on the functionalities of the SUUNTO Training Manager (STraM) accompanying software;
6. it guides you in the tuning process of the software to achieve the highest accuracy of the calculated oxygen consumption (VO2) and other body parameters such as EPOC and the related Training Effect (Chapter 8);
7. it contains an extensive discussion on the application of the ventilation, oxygen consumption and heart rate data to find valuable threshold values like the lactate threshold (Section 10.6). This all is based on the ability of STraM to provide these instant body states by analysing the instant heart rate variability. This makes the t6 and STraM so unique.
This document also contains the t6 Swiss Army Knife manual (see Chapter 10), which can be used to automate the retrieval of several useful data:
1. the accurate LAP- and SPLITdistances;
2. the accurate calibration chart;
3. the relevant interval and recovery data;
4. relative accurate estimations of the tuning parameters.
5. relative meaningful threshold values HR.LT, SPD.LT, HR.MT
For the Dutch speaking people: Deze uitgebeide samenvating is gemaakt aan de hand van een uitvoerig testdocument (in het Nederlands opgesteld) dat ik gemaakt heb voor de harttslagmetershop (http://www.cardio-systems.com/). U kan daar eens aankloppen. Wellicht kunnen ze u daar verder helpen in het Nederlands.
About Speed and distance the following items are covered throughout the document:
1. Reliability and robustness: Does the PODs (Peripheral Observation Devices) have a reliable and robust data transfer to the t6?
2. Accuracy: In the technical specifications of the manual it is claimed that the accuracy of the speed and distance sensor is “97% when calibrated”. Is this correct? In which cases this accuracy is not met? And what should be done in order to get the highest accuracy?
a. Are distance based interval workouts really feasible in real life?
b. Are the SPEED LIMITS really functional?
c. What is the nature and the resolution of the speed graphs in the STraM?
4. Consistency: Is the speed and distance data provided in the t6 and the STraM consistent?
5. Bugs: Are there major design flaws, and are there workarounds?
75% of the content herein is not available in the manuals, Suunto website or documentation. However, I really feel this knowledge is very useful for your day-to-day use of the t6, or to evaluate the t6 before a possible purchase.
If you don’t have time to go through the report, here is my general conclusion on the t6:
1. It is a long lasting investment because the t6 uses a reprogrammable flash memory. This makes ‘firmware upgradings’ during a factory service perfectly possible.
2. Extremely short learning curve: In general the t6 is a very easy to handle wriststop personal trainer. It also provides support for the most essential workout types (but there aren’t numerous programming features).
3. The t6 is extremely accurate, highly reliable and robust.
4. The t6 footPOD can be used for steady endurance running as well as for workouts where speed changes are part of the exertion (e.g. interval, fartlek), or for workouts where stride variations often occur (e.g. hilly terrain running).
5. The STraM is a step into the 21st century:
a. It is the first WristStop Trainer that calculates the instant VO2 oxygen consumption, and is dealing with an integrated EPOC (oxygen related) training effect. But as with other methods used in the WristStop Trainer business –OwnIndex test, Orthostatic Overtraining test, - the user has to trust the outcome. I haven’t found scientific proofs of its accuracy. Once tuned however, I have found that the estimation of VO2max is quite similar to the outcome of the Daniels formula;
b. STraM is database oriented. This makes it possible to manage and manipulate all kinds of measurement data in an integrated way.
c. Based on several intelligent mathematical models, STraM is able to calculate instant varying body parameters. These workout related body parameters can be exported and analyzed further in separate softwares. An example is the detection of the body heart rate or running speed at lactate threshold or at VO2max (see Section 10.6.4).
6. As result, classical single workout analysis tools are not very much supported, but a HR and speed curve observation as function of stopwatch time and distance is available.
Discussions and comments on this document are welcome in the Yahoo WriststopTrainers discussion group (http://health.groups.yahoo.com/group/WriststopTrainers/ ).
Good luck with your training mate!
I am neither a SUUNTO technician nor expert. So I take not any responsibility whatsoever on the content of this document and all possible effects this could have.
The t6 is not cheap. Therefore it is important you will have value for money. Nothing is ‘perfect’. Also the t6 is man made, and it has its design flaws. In this regard it is important to know what kind of solutions are offered if you encounter such flaws:
1. the accompanying software (Suunto Training Manager) is actually at level 1.3. The next level 1.4 will have additional features. At regular times bug fixes are also downloadable. Therefore it is important to report bugs and suggestions for improvements to SUUNTO helpdesk.
2. The t6 uses a reprogrammable flash memory. This makes ‘firmware upgradings’ during a factory service possible. At this time this update can’t be done at home though the serial port, since more pins are to be used for that.
The next firmware will be released before the end of this year, and will cover several firmware bug fixes (e.g. the INT2 beep design flaw).
3. After 5 months of using the t6, I feel that the weakest point of the equipment is the footPOD. This is probably due to the fact that the match of electronics and fast movement (accelerations) is not straightforward. Hence I really feel that it has its benefit to buy somewhere where the after sales service is excellent (like here). If you footPOD is broken after 6 months, just 2 weeks before your target race, you will understand that value.
The running and cycling speed is measured with wireless PODs (Peripheral Observation Devices). First of all a correct signal reception is to be obtained from the hardware PODs. Then a correct signal transfer from the PODs to the t6 and a correct data management and recording in the t6 is to be achieved.
All these steps are discussed in short.
The footPOD has to be placed transverse to the running direction. This position is critical towards correct speed measurement (3% measurement error and more! See Section 11.2.1). Once the footPOD is placed in the plastic cover and the laces are tied, it should be impossible to move or slide the footPOD in any direction or angle. Correct attachment guidelines are provided in Figure 1.
Figure 1 Correct attachment of the footPOD. The upper middle photo shows the correct transverse position. Be sure the laces are tied correctly and firm under the plastic cover. In 3 and 4 the plastic cover should bulge between the tied laces (It is recommended to have a distance of about 2cm between the lace crossings attaching the cover). Then the footPOD will be attached correctly in the cover (5 and 6). Some people attach the pod to the laces, not using the method of sliding it between the laces, but thread the laces through the pod holder. In any case, when the sensor isn’t fixed firm and correctly, instable recordings (a variation in distance recordings for the same run) will be the outcome.
Extensive tests show very clear that violating one of these guidelines will for sure result in very erratic speed and distance measurements. See Section 11.2.1 for some data.
Some people report that their footPOD is still too loose with this kind of attachment. They can attach their footPOD as is shown in the next figure, by crossing the laces (between the clips) ‘on top’ instead of ‘under’ the footPOD.
Figure 2 Special firm attachment of the footPOD.
POSE runners should take care that the footPOD is positioned as close as possible to their toes (see Section 18.104.22.168 for reasons why).
In the manual, the attachment guideline of the bikePOD isn’t clear. Special attention is to be taken for the following items (see Figure 3):
1. (Bottom photos) the best position of the bikePOD is the one with the fewest obstacles between the POD and the t6. So if you attach the t6 on the bike mount on the handle-bar, be sure no parts of the frame or fork is between the bikePOD and t6
2. (Top photos) there are only two correct locations where the magnet should pass the sensor. These locations are denoted with white colored bullets on the bikePOD. (I have had feedback of a t6 user indicating that his bikePOD works correctly in any position between the colored bullets.)
3. (Middle photo) choose one of the two positions to be able to manage the maximal distance between the magnet and the bikePOD.
Figure 3 Correct attachment of the bikePOD
Extensive tests show that the whole measurement chain is robust and stable. Basically this is due to the DYNASTREAM proprietary ANT wireless Personal Area Network (PAN) technology used in the PODs:
· It is a ultra low power network solution
· It is a digital signal transfer on a 2.4GHz RF transceiver (almost no interference problems with other sources)
· It is coded (user crosstalk immunity)
The technical specifications say that the ANT technology is able to achieve a correct transmission up to 10 meters. In practice it can be shown that this only holds in clear weather conditions and without any obstacles between the ANT devices. So in practice, the effective distance is most often much smaller, as can be observed when a proper position of the bikePOD has to be found. Remark also that ANT can not penetrate water (hence also not your body) because it uses radio frequencies (electromagnetic waves).
Before the t6 is able to receive a proper POD signal, a single and specific ID should be assigned in the t6 to every POD. This is called ‘pairing’.
Pairing is only necessary the first time a POD is used, or when the batteries of the t6 or the PODs are to be replaced.
One should take care of the following when one is about to pair a POD:
1. if you start to pair a POD, be sure you first short-circuit the battery contacts in the device;
2. then go into the proper menu of the t6 and start the paring;
3. When the t6 starts seeking the POD, then enter a fresh battery in the POD (not earlier).
Feedback of SUUNTO on this topic:
“…Pairing is not needed after replacing the battery of t6 or any of the pods. Pairing while another pod is on does not lead to erratic values, usually pairing just fails…”
When you want to display POD sensor measurement data, you first activate (make awake) the corresponding sensors:
1. The heart rate sensor is activated as soon as a heart beat is sensed by the sensor. So (1) the contact surfaces must be sufficiently wet, and (2) they must be in tight contact with the chest skin.
2. the footPOD is activated when the button on the back is pressed until the red led starts blinking;
3. The bikePOD is activated as soon as the magnet passes one time the bike sensor.
Then you synchronize the t6 with the PODs that are activated (awake). This is called ‘Connect’:
1. The shortest way is to navigate the t6 into the TRAINING MODE or SPD/DST MODE. In either of these two MODES you can press-and-hold the lower left button (ALT/BACK) to start searching the active sensors. Then ‘Searching’ is displayed, and each of the sensors found is declared on the display.
2. It is also possible to select the ‘Connect’ menu item I either the TRAINING MODE or SPD/DST MODE.
Once the sensor is connected, the received sensor signals are displayed.
If one also wants to record the sensor signals too, then one supplementary need to start the StopWatch by a press on the upper left button (START/STOP).
If one wants to stop recording the sensor data, first the t6 recording process is paused by press on the upper left button (START/STOP), and then the record file will be closed by press-and-hold on the upper right button (UP/LAP). Once the data file is closed properly, the StopWatch displayed indicates a zero time, and the distance displayed indicate a zero distance.
If one wants to continue the recording process after pausing the StopWatch, just presses once again on the upper left button (START/STOP).
At this time there is no real time recording of the measurement data, i.e. the STraM software is not able to show real time gaps due to pausing the StopWatch. Actually only the real time is recorded at the time the StopWatch is started, not at each time the StopWatch is continued.
In normal working conditions it is rather unusual that the t6 looses the connection with the PODs. However, such situations are not unusual. This can occur:
1. When the effective transmission distance becomes too big. Since the ANT data transmission can not penetrate water, you will for example not measure your heartbeat when you hold your wrist with t6 on your back.
2. when a sensor is de-activated:
a. When you press-and-hold the footPOD when she is activated.
b. After the magnet didn’t passed the bikePOD for half an hour (auto-off).
c. After the heart rate sensor does not receives a heartbeat for a while (auto off). This can be the case when the heart rate sensor looses contact with the skin, or when the contact surfaces become too dry.
In such situations, the t6 shows with a searching indicator which device isn’t received any more:
1. when the heart rate isn’t received any more, minus signs ‘- - - ‘ are displayed on the location of the heart rate on the display;
2. When the footPOD or bikePOD isn’t received any more, minus signs ‘- - - , - ‘are displayed on the location of the speed value on the display.
Feedback of SUUNTO on this topic:
“…t6 keeps searching "lost" sensors 15 minutes if stopwatch is on and 1 minute if stopwatch is off/paused…”
The t6 then assumes that the measurement device is deliberately deactivated and hence will not search any more that device.
When the t6 assumes the device is deactivated, it will stop to search the device and shows a zero value ‘0’ instead of the searching indicator (minus signs).
Before an ANT transmission channel is established between the device and t6, the t6 needs the full battery power to seek for the corresponding time slots in which the data will be transferred. Also, when the connection with a device is lost (searching indicator appears on the display) the t6 also needs its full power to seek the missing device. Therefore, other power consuming tasks (light, sound) are disabled during that period to prevent the t6 from being malfunctioning. For additional information, see Section 22.214.171.124.
As long as the t6 is searching the device (searching indicator is displayed) it will re-connect the corresponding device seamlessly without any manual interventions. If such a situation occurs during the recording process (StopWatch running), it is interesting to know what data are recorded in the t6
In order to show this, one should be aware of the different recording principles of the heart rate and the other POD signals. The specific difference is that the heart rate recording is R-R based (beat-to-beat), which allows for subsequent EPOC analysis in the Suunto Training Manager software package (STraM). Such an R-R recording does not record the average heart rate are the recording times (or sample times), but it records every time difference between each heartbeat up to one millisecond.
Except from the heart beat recording, the other sensor values are recorded at the sampling time the t6 is set, or when the manual LAP is taken (this also happens when the StopWatch is paused).
With the STraM software it is possible to set the sampling time recording rate to 2 sec or to 10 sec (See Section 3.4.2).
Figure 4 shows the result in the STraM of a recording process where the signal reception of a bikePOD and the heart rate sensor between minute 2 and minute 3 was deliberately malfunctioning (a searching indicator was enforced).
Figure 4 STraM visualisation of the recorded signals of speed sensor and the heart rate sensor. Between minute 1 and minute 2 the t6 was searching these devices. This was achieved by the deliberate action of separating the t6 more than 10 meter away form the active bikePOD and heart rate sensor. Before 1 minute there was a low heart rate and bike speed. After 2 minutes there was a high heart rate and bike speed.
From this figure one can conclude that during the time a searching indicator is displayed, the corresponding sensor signals are recorded differently:
1. the t6 records zero speed values for the time the t6 display showed minus signs
2. the STraM shows interpolated heart rate values during the time the t6 display shows minus signs
Feedback of SUUNTO on this topic:
“…STraM shows some interpolated value for heart rate if signal is lost during training (stopwatch is on). t6 actually do not record speed but distance between recording interval (2 or 10s). There can be seen speed spikes after connection failure, this is because during no connection t6 records 0m distance increments and when connection is ok again t6 gets one far too big increment and STraM cannot handle that at this time…”
According to other feedback, future StraM will handle such occasional spikes.
When a POD is not connected, or when an active POD becomes disconnected for a longer period than 1 (stopwatch paused) / 15 (stopwatch on) minute, then a zero sensor value is displayed on the t6. This last situation can occur in the following real live situations:
1. if you step off your bike (t6 on your wrist) and walk a certain distance to make water, the displayed speed can turn from ‘ - - - , -‘ into ‘0’;
2. If you step off your bike (t6 on the handle bar) and walk a certain distance to make water, the displayed heart rate can turn from ‘- - - ‘ into ‘0’; (I really do not advise to leave the t6 unattended).
3. Assume long duration relief races (team race). If you only want to record your personal running time, and not the recovery period where other mates take over the run, one can pause the stopwatch during your time of inactivity. During these pauses it is possible that the contact surfaces of the heart rate sensor dry up, which make this sensor de-activated. If then the contact with the t6 is broken for more than a minute (stopwatch is paused), the t6 will display ‘0’ bpm HR.
However I do not recommend to pause the stopwatch for that purpose. Your training level will then show wrong values because you cut out important rest times out of the HR recordings.
4. If you take a break just aside your bike for longer than 30 minutes. Then the bikePOD will be deactivated automatically.
In such cases the t6 will not re-synchronize automatically. Hence, once the missing sensor is active again or when the active sensor is again in the vicinity of the t6, the t6 should be informed to seek all the active sensors again. This can be done by another press-and-hold on the ALT/BACK button.
Figure 5 shows the result of such a process.
Figure 5 Effect of a manual re-synchronization of sensors from which the minus signs were turned into zero values after one minute of no signal reception. At time1 minute the stopwatch is paused (see marker). Then the t6 is taken away from the active sensors for about 2 minutes. This leads to zero sensor values displayed on the t6. In this state the t6 StopWatch is continued for another minute, and paused again (see second marker). Then the t6 is set again in the vicinity of the active sensors and the t6 is forced to resynchronise to the active sensors again. Once this is achieved, the StopWatch is continued for another minute. Before 1 minute there was a low heart rate and bike speed. After 2 minutes there was a high heart rate and bike speed.
From this figure one can conclude that, once the t6 stops searching to the active sensors (the minus signs are then replaced by zero sensor values), the corresponding sensor signals are recorded significantly different:
1. The t6 records zero speed values for the time the t6 display showed zeros.
2. The STraM cuts out possible interpolated heart rate values, and just shifts the heart rate values to the right. This results in two weird effects:
a. The displayed heart rate signals beyond the resynchronization are all shifted on the timeline. This makes interpretation of the heart rate ambiguous;
b. At the end of the timeline the missing heart rate values are displayed as zero values.
From this one can conclude that a disconnected speed sensor does not pose problems during a re-Connect. A disconnected heart rate sensor however kills the heart rate data if one re-Connects the sensors.
Hence the following PRECAUTIONS are recommended:
As long as you plan to record heart rate data in the same workout data file:
1. Keep the heart rate belt tight (do not put it off), keep its contact surfaces wet, and keep the t6 in the vicinity of the heart rate belt.
2. If it is necessary to re-Connect a device, be sure the heart rate displayed on the t6 does show correct heart rate values, no zero values. Then the re-Connect does not invoke the heart rate shifting effect at all.
These precautions are very effective, as can be seen in Figure 6.
Figure 6 Manual re-Connect. The scenario is identical to the one shown in Figure 5. The only difference is that the PRECAUTIONS mentioned above are followed strictly.
If you are unable to show correct heart rate values on the display, then you will need to start a new data file instead of continuing to record into the same data file.
As already mentioned in the previous chapter, the heart rate sensor data is recorded as R-R data (beat-to-beat time interval recording).
The other sensors (speed and distance sensors, temperature, and altitude) are sampled automatically or manually:
o Every 2 or 10 seconds.
With the STraM software it is possible to choose one of these sampling rates (See Section 3.4.2);
o when a lap is recorded automatically:
§ when the ‘Autolap’ function is set ‘on’ (SPD/DST MODE menu), every time the autolap distance is reached, a laprecord is taken automatically;
§ when the ‘Timers’ function is set ‘on’ (TRAINING MODE menu, 2 carrousel INTerval TIMERS, 1 WARMUP TIMER, 1 COUNTDOWN TIMER) ), every time a timer value is reached, a laprecord is taken automatically;
§ when the distance ‘interval’ function is set ‘on’ (SPD/DST MODE menu, 2 carrousel INTerval distances), every time a distance value is reached, a laprecord is taken automatically;
· when a lap is recorded manually:
o when the LAP button is pressed;
o When the Stopwatch is paused.
The registration of the altitude data will only be done if the air pressure sensor is set in the t6 to ALTitude. If it is set to BAROmeter, this measured sensor value is not recorded.
Up to 100 LAPtimes can be recorded. After that, the t6 will not record a LAP any more.
By starting the StopWatch, the t6 stores a workout header, and starts to fill the t6 memory with sensor data. So, it depends on the number of sensors to be recorded, the workout time, and the number of heartbeats, how fast the memory of the t6 is filled. For a very small workout the following memory allocation holds:
· HRbelt connected: about 1% memory used
· HRbelt and footPOD connected: about 1,25% memory used
· HRbelt and footPOD connected, USE alti: about 1,65% memory used
So theoretically, after recording 80 tiny workouts with HRbelt and footPOD connected, the memory is full.
Feedback of SUUNTO on this topic:
“…Maximum number of logs is 30 and you'll get memory full information if trying to start stopwatch when there's already 30 logs even if there's still memory left…”
One can conclude that the memory allocation depends in practice on the following items:
1. Heart rate: The technical specifications say the memory can allocate 100.000 heart beats. For an average of 150bpm, a single workout can last 11h5min. If 11 workouts are to be stored, the header information makes the total workout time 9h43minutes. It also should be clear that due to the unknown mechanism the real number of storable workouts will be less than 11.
This all is only valid of no other sensor values are recorded.
2. sampling period of the other sensor values: 10 sec recording rate will consume less memory than a 2 sec recording rate;
3. the number of activated and connected sensors;
4. The number of LAPrecords: as soon as there are more than about 30 LAPs recorded, the t6 has to add another memory page to the header for each subsequent 30 LAPs (about. See MSG 34).
To get an idea, in Table 1 some realistic memory allocation data is given.
Number of active sensors recorded
Number of LAPrecords
Memory usage per hour
HR; bikePOD; ALT
HR; bikePOD; ALT
Table 1 Realistic data of the memory allocation
For long time speed recording and for endurance cycling it is more appropriate to use a Rec.Rate of 10 sec. For this purpose I use following ‘safe and easy to use’ formula (ALT on; 10 sec rec rate):
o storable_beats = 100.000beats *(100% - 1%_per_hour_to_record)
o HRavg = storable_beats / hours_to_record / 60.
Example: If you want to cycle one trip in the tour de France and estimate the cycling time as of 9 hours. The number of storable beats are then 100.000*0,91=91.000; and HRavg = 91.000/9/60=169bpm (granted).
A more accurate equation is proposed in Section 3.3.3.
The t6 is capable to measure correctly the physical parameters for low HR too. This makes it possible to measure e.g. the daily energy consumption.
I recorded a Grand Day Out (including a 90 km low speed cycling tour with only HR recording). The following data were observed: SPDavg=15km/h; HRmax:=117; HRavg=67; HRmin=42; Total HR recording time before memory is filled: 30h42min10sec
Calories consumed (tuned STraM) 5572kCal or 4356kCal on 24 hours (including the ride and a heavy meal)
The following graph (Figure 7) shows another Grand Day Out, recording both the HR
and the bikePOD and footPOD (Rec.Rate 2 sec;
Figure 7 A Grand Day Out with HR and speed sensor activated.
The following data were observed: HRmax=152; HRavg=82; HRmin=46; total HR recording time before memory is filled (i.e. both HR and distance recorded well) : 14h57min36sec .
The memory of the t6 is 128kB in size. Member Paul (firstname.lastname@example.org ) found some elements on how the t6 memory is organized internally (see MSG 10 and 18), and showed 124kB s reserved for workout data logs.
Paul showed that this recording process uses 512 byte pages. For the header information and the recorded LAPdata (I assume this based on the raw data) there is always one page is allocated. And per sensor to be recorded, also one page at a time is allocated.
For a very small workout the following memory allocation holds:
· HRbelt connected: exactly 0,5 kB header + 0,5 kB
· HRbelt and footPOD connected: exactly 0,5 kB header + 1 kB
· HRbelt and footPOD connected, USE alti: exactly 0,5 kB header + 1,5 kB
Based on this Paul showed the following t6 memory usage algorithm:
pages per log =
1 (log header + space for 29 LAPS) +
1 (if the number of LAPs is above 29) +
1 (if the number of LAPs is above 63) +
1 (if the number of LAPs is above 97) +
1+int( (1+duration/interval) / 255) (if altitude) +
1+int( (1+duration/interval) / 255) (if distance) +
1+int( bits_per_beat*avgHR/60*duration/8 /510) (if HR)
5580s () log at 10s interval @ avgHR 165bpm using 9.2 bits/beat
3 (1 + (1+5580/10) / 255) +
3 (1 + (1+5580/10) / 255) +
35 (1 + 9.2*165/60*5580/8/510)
= 42 512byte pages = 21Kb
If the above were at 2s interval recording, that would be 1+10+10+35 = 56 pages
With this a more accurate (but also simple) formula can be found for cycling purposes:
When one takes the maximum laps, Rec Rate 10 sec and ALT on; and assuming 9 bits per beat on the average for the whole trip), the following expression for cycling purposes hold:
Taking the 9 hours trip of above (See Section 126.96.36.199, duration 32400sec), the HRavg = 180bpm
A more accurate formula for cycling would be:
o pct = (100% - 1%_per_hour_to_record * a) (a=1 if Rec.Rate=10; a=5 if Rec.Rate=2)
o HRavg = 110.000 * pct / hours_to_record / 60.
With this, the tour de France 9h cycling trip HRavg is 110.000*0,91/9/60=185bpm.
Basically the t6 has 3 numerical lines on the display.
There are 4 DISPLAY MODES in which the display of the t6 can be selected by pressing sequentially on either the UP/LAP button (upper right) or DOWN/LIGHT button (lower right):
· ALTItude/BAROmetric pressure
In each of these MODEs the SUUNTO button (or ENTER button middle right) can be pressed to select the menu of the corresponding MODE.
Remark that, once the Stopwatch is running, the UP/LAP button also records a LAP. Hence, to avoid unwanted LAPrecords in a workout, it is advised to use only the DOWN/LIGHT button to change the DISPLAY MODE.
Once the DISPLAY is in a certain MODE, the lower line can be changed by pressing the ALTernative/BACK button (lower left).
This makes it possible to set the instant speed on the top line, the instant running distance on the middle line, and the instant heart rate on the lower line:
1. From TIME MODE press two times on DOWN/LIGHT;
2. Then press a few times on ALTernative/BACK.
(With this button one can choose in this SPD/DST DISPLAY MODE among:
o Time of day;
o Instant heart rate;
o StopWatch time;
o Average speed since the start of the StopWatch;
o LAPdistance so far (=distance from the last LAPrecord)
If one chooses heart rate, the t6 display looks like Figure 8.
Figure 8 most wanted display for speed and distance running. This display can be preset prior to the workout, or can be adapted instantly. More information on the instant guidance is given in Section 7.2.
One observes that the speed is displayed up to 0,1km/h and the distance up to 10 meters.
There is one location in which one can observe the distance up to 1 meter; this is when an automatic calibration is done. After the calibration run (flying start and stop conditions are recommended by SUUNTO!), one has paused the StopWatch just at passing the finish marker. Then by selecting the TRAINING MODE (DOWN/LIGHT button) menu (SUUNTO button), one can select CALIBRATE. Here you can observe the running distance up to 1 meter.
The basic settings of the t6 can be modified in the menu UNITS of the TIME DISPLAY MODE. Here the user can choose to display speed values in km/h, mi/h, min/km or min/mi. This can be done as follows:
· select the UNITS submenu in the TIME DISPLAY MODE
· go to the SPD option where you can choose between speed and pace representation:
· if you have selected the pace representation, you also want to set the DST option:
o if you set the DST option to km, the pace will be in min/km
o if you set the DST option to mi, the pace will be in min/mi
Changing these settings can be done instantly, i.e. also during the recording process during a workout. However, I feel most users will stick to one preferred setting at a time.
The ‘Store interval’ (2 sec or 10 sec) of the samples sensors has to be set from the STraM. This is shown in Figure 9. Here you can also set the speed settings, and upload these to the t6.
Figure 9 t6 settings that can be changed with the STraM. After connecting the t6 Device (see the cursor in the window to the left), the t6 Settings appear in the bottom window. In the TAB ‘Settings’ the various settings of the t6 can observed, and eventually modified. The ‘Store interval’ can be changed here. The modifications are transferred to the t6 when the Update settings button is pressed.
The RAW DATA stored in the t6 is not visible directly. Once a workout is uploaded in the STraM, this software makes the data visible numerically in the different TABS below the graph:
1. the speed data is visible in the STraM up to 0,1km/h;
2. The distance data is visible in the STraM up to 10 meters.
The 4 TABS below the graph are organized as follows (see e.g. Figure 14):
1. TAB Details1: contains the major header information as the total running distance, running time, average speed,
2. TAB Details2: contains additional header information, personal motes can be entered;
3. TAB Marks: contains alls the LAPrecord data;
4. TAB Data: contains all the sampled sensor data
It is also possible to change your user settings in the STraM (e.g. speed displayed in km/h, mi/h, min/km or min/mi). Therefore you click on your name in the top left window ‘My Training’. Then your user settings can be changed in the lower window of the STraM. See Figure 10. Here you can also define user specific activities.
Figure 10 User settings in STraM. View on the first TAB ‘Settings’.
It is also possible to export a workout stored in the STraM. See Figure 11.
Figure 11 Exporting workout data from the ‘My training’ window. Right click on a workout, and select ‘Export to file’.
This stores a file with extension .sdf (Suunto Data File I guess) on your hard disk. After careful analysis of this file the following conclusions can be made:
1. the LAPdistances are stored with an accuracy of 10 meters;
2. The distances at the sample times are stored with an accuracy of 1 meter.
3. all the speed values have more significant digits than the sampled distances, indicating that the speed values are calculated ones form the distances at the sampled times.
A RAW DATA analysis done by Member Paul (many thanks Paul!) confirms this conclusion. More exactly, the RAW DATA show that the following basic speed and distance data is recorded in the t6:
· the distance increments at each sample time;
· The LAPdistances.
So (1) the distances at the sample times, (2) the SPLITdistances, and (3) the speed are all calculated within the STraM.
The following conclusions can be made also:
1. The exact workout distance is not known in the STraM, because the recorded LAPdistances only have an accuracy of 10 meters. Also the sampled distance is taken only at the sample times, which is almost never a stop time. However, when a 2 second recording rate is taken, the difference won’t be huge.
2. when a 2 sec recording rate is used, the most accurate workout distance available is probably the sample time distance, When a 10 sec recording rate is used, the most accurate workout distance is probably the SPLITdistance (=accumulated LAPdistance).
It is recommended that SUUNTO records the LAPdistance with the same accuracy as the sample time distances. As will be seen in the sequel, this will improve the data consistency.
3. From the recorded R-R heart rate data, average heart rates are stored at each time sample in the .sdf file, as well as average LAP HR and max/min/avg workout HR.
In order to find out which kind of speed values are calculated and displayed in the STraM, a .sdf file of a small test workout is exported and manipulated in excel. The result is shown in Figure 12.
Figure 12 Comparison of the speed profile provided in the STraM (blue) and the speed profile calculated with a backward trapezoidal integration rule (red)
From this figure one can conclude the following:
1. The first part of the graph shows that the STraM speed curves are hardly nervous (no jitter). It is able to show a speed trend instead of an instantaneous speed value at a certain timestamp.
2. The second part shows that sudden speed change is hard to reconstruct in the STraM. From second 14 there was no distance increment any more, hence the speed is 0km/h from that time on. The displayed speed in the STraM however becomes zero only 10 to 12 seconds later.
3. The experiment has been done with the bikePOD and a free rotating front wheel (off ground). All of a sudden I used the front brake to stop rotating the wheel. At the same time I took a LAPrecord. The Laptime was ,4. This indicates that, at least for the bikePOD, the t6 detects extremely fast when sudden speed changes occur. This is proven by the fact that already 0,6seconds after the wheel stopped there is not any distance increment any more.
Once one knows that the distance data is very accurate, and that the speed curve is a nice looking trend visualization of the cycling or running speed, these design decisions are very much appreciated: internally the basic data is accurate; externally the graphs show what they need to show.
1. It is shown in the previous sections that during a temporary disconnection of the HR belt during the recording process, the STraM will interpolate the missing heartbeats between the correctly recorded heartbeats.
2. Also, the HR graph in STraM shows an high resolution average of the beat-to-beat timings recorded: It is not the instant beat to beat change that is displayed as instant HR, but an average over a few beats.
According to my personal feeling, the HR report on the display also uses an high resolution averaging, but there more beats are taken to display the instant HR.
Both effects can result in a slightly different HR report between the STraM and the t6 logbook. However the difference in the HRmax report is almost never bigger than one beat. The difference in the HRavg report is also very small and depends mainly on the time the HRbelt was disconnected.
It is clear however, that the HR data processing from measurement to report is robust towards temporary transmission problems.
1. In the sequel it will be shown that the t6 does not record instant speed values, but distances only. STraM reconstructs the speed profile by creating a nice looking speed curve from these distance datapoints. Also here some averaging (smoothing) takes place to get a better indication of the normal speed changes.
2. It is shown in the previous sections that during a temporary disconnection of a footPOD or bikePOD during the recording process, the STraM will show a zero speed. One can show that these zero speeds originate from zero distance increment recordings at the sample times (at the sample times it is not the total distance that is recorded, but the difference with the previously recorded distance).
Feedback of SUUNTO on this topic:
“…When there's no connection to the speed sensor, distance increments are not recorded or they are zero values. If distance travelled during no-connection period does not exceed certain values there won't be errors in total travelled distance shown by watch or STraM. These distances are 256m for Foot POD and 4096m for Bike POD.”
So this means that as long as the temporary disconnection does not exceed 256 meters for running, there is not any measured meter lost in the global distance data: The t6 subtracts the old distance to produce the interval distance. That way, if a transmission is lost (‘---,-‘ on the display), it is automatically corrected when the next one is successfully received and the travelled distance is not exceeding 256 meter.
I like very much this way of catching-up not received distances and the way STraM deals with it. To show how it works, a steady running workout with SPD LIMITS 11,8km/h-12,2km/h is given as example in Figure 13.
Figure 13 Steady workout runned at target speed 12km/h.
At first sight the speed dip right before km 9 is weird. The workout datafile show the following values (the values in bold font are the recorded data, the other columns are derived data in STraM):
total distance (m)
Recorded distance increment (m)
STraM Speed profile (km/h) (lower red graph in Figure 13)
The colored rows are exceptional low (running 12km/h on the average would result in a distance increment of about 6,7 meters per 2 seconds). Because I did not stand stil, the zero distance increments must be due to a lost connection between the footPOD and t6. The low distance increment of sample 0:45:24 denote that the connection has been lost already between 0:45:22 and 0:45:24.
If one assumes an average speed of 12km/h, then the not received distance during these three samples is about 6,7*3-3= 17,1 m. This not received distance is catched-up when the connection is re-established, somewhere between 0:45:28 and 0:45:30, and added to the regular distance increment 6,7 meter for that sample period. As result, the first sample at which the connection is established again should have a distance increment of 17,1+6,7=23,8 m. This is perfectly in line with the true recorded distance increment at 0:45:30 (24 meter).
In the last column of the table, it is shown how STraM calculates the speed profile from these distance increments. STraM uses some low pass filter or some averaging to make the speed curve looks nice and meaningful. As can be seen here, at least two zero distance increments are needed to create a speed dip. The distance catch-up will not produce a speed spike since it lasts only one sample.
This method dealing lost distance records could be improved even more to solve as well with the speed dips that are introduced due to the zero distance increments, see item 10 (page 199) of Section 188.8.131.52 .
This is discussed after the software STraM is discussed in extension. If you can’t wait to understand why sometimes the t6 LOGBOOK values are different with the values reported in STraM, just jump to Section 7.3.5.
Several measurements show that, for cycling, the distance calculation process in the t6 is basically the counting of the number of times the magnet passes the bikePOD.
It is already shown in the previous chapter that the speed graph is optimized to keep track of the speed trend, rather than being an exact copy of the instantaneous speed.
On the display of the t6 another compromise is made for the displaying of the instant speed. In Figure 14 the experiment of Figure 12 is shown in the STraM application.
Figure 14 View on the speed graph and Data TAB of Figure 12 in STraM.
The second LAP (see the second MARKER) is taken at the time the speed on the display of the t6 becomes 0km/h. In this case the time between the two markers is about 18 seconds. This means that, although the front wheel does not rotate, the displayed cycling speed can take 15 to 20 seconds before it really displays 0km/h. This is due to the compromise made by SUUNTO: also very low cycling speeds can be shown for bikes with a big wheel diameter (1km/h is perfectly possible).
This benefit does not harm the data accuracy in the STraM, since the distance recording process is totally independent of this displayed speed.
One can conclude that the displayed cycling speed on the t6 is the speed the bicycle has during the last two passing of the magnet along the bikePOD. The displayed distance however (which is also used in the STraM) corresponds to the number of times the magnet has passed the bikePOD.
1. First put at least 4 bar air pressure is in my front wheel.
2. Then go to a 400m running track (ask first. If this is not allowed, you will have to look after a roller-skating or cycling track for which you also know exactly the distance.).
3. Then place the front wheel right on the start line, and place the magnet right before the bikePOD.
4. Then connect the bikePOD.
5. Then start the stopwatch and do 6 km.
6. Right at the start line stop the stopwatch,
7. Calibrate the bikePOD by the automatic procedure provided in the t6. (This procedure does not recalculate the recorded distances.)
I uploaded this workout data into the STraM and then into the Swiss Army Knife to catch correctly the distance up to 1 meter and to recalculate also the calibration factor (See Chapter 10). The conclusion was that the automatic calibration method of the bikePOD was fine.
8. Once the bikePOD is calibrated perfectly, look after a flat tour of about 1,2 to 1,5km in the neighborhood where you live. Do it on the same day, or at least be sure the tire pressure is exactly the same.
9. Do 4 times that tour and record the distance in the same way as is done in points above,
10. Extract the correct distance with the STraM and my Swiss Army Knife factor (See Chapter 10), and divide the accurate distance with 4. This will give you the exact distance of that tour
Be sure you always cycle on the same side of your tour (preferably inner side).
This distance is then very helpful for the rest of your life. When you want to recalibrate your bikePOD (other tires, other tire pressure) or your footPOD, that tour becomes very thandy...
Due to the distance measurement principle discussed above, the accuracy completely depends on the wheel pressure and weight on the bike. If I take 4 bars on my tires, I get accuracies on the distance of 99,9% and better. The inaccuracy involved is due to the slightly different cycling path and possible slightly different pressure due to the ambient temperature.
In Figure 15 it is also proven that the distance accuracy is totally independent of the number of LAPs taken.
Figure 15 Two identical cycling workouts. (red) only 1 LAP has been taken (StopWatch stop); (blue) autoLAP 0,2km used. The speed graphs on top are shown on the time axis. The speed graphs at the bottom show the same speed graphs as function of the cycling distance. During the workout with the autoLAP ON I ran deliberately faster than during the workout with autoLAP OFF.
From these graphs, it should be clear that the intermediate LAPs do not result in distance inaccuracies.
Because the distances
in the STraM are only visual up to 10 meters, and since the accurate distance
at the end of the workout is not recorded, I made a t6 Swiss Army Knife. See
Chapter 10. This excel sheet imports automatically an .sdf file and
reports the recorded and most accurate
1 LAP record
autoLAP 0,2km (52 LAPs)
estimated error per LAP
Table 2 Recorded and
The results are twofold:
1. if one takes 1big or many 0,2km laps for the same distance, there is no influence at all on the distance measurement;
2. The difference between the recorded and (Swiss Army Knife) calculated SPLITdistance is very small for big distances –less then 10 meters- (but it can be relatively big for small distances).
In Figure 16 the speed graph of the STraM is compared with a more nervous speed graph of the same recorded distance data.
Figure 16 (blue) STraM speed profile; (red) speed profile obtained with a backward trapezoidal integration rule.
From this one can conclude that the speed graph of the STraM well designed for speed trend detection.
Most of the results of the cycling analysis also hold for running workouts. E.g.:
1. Only the measured distances are recorded. Hence the displayed speed on the t6 differ with the displayed speed graph in the STraM (which is optimized for trend observation);
2. Extensive analysis also shows that the t6 also does not create measurement errors if one takes one big LAPrecord or a set of small LAPrecords.
The major difference with the cycling is the footPOD, which measures completely differently the running speed and distance.
SUUNTO uses the DYNASTREAM patented accelerometer measurement system to calculate the body speed by measuring the foot acceleration/deceleration profile of each stride. A white paper can be found in http://www.dynastream.com/datafiles/SpeedMax%20White%20Paper%20v4_1.pdf .
Some mathematics involved can be found in the thesis
In the sequel a summary of the other results of the footPOD is shown.
Extensive tests show that the speed on the display will be adapted to the body running speed already after around 2,5 seconds. In every case the instant speed is adapted within three footsteps of the foot on which the footPOD is attached.
The response to smaller body speed changes is even much smaller; one has the feeling that at every complete movement of the footPOD the instant speed is adjusted.
This has a major benefit for speed limits workouts. Because one knows the measured speed will always be adapted in less than 3 footsteps of the foot on which the POD is attached, overreactions on speed differences (desired-measured) are almost totally excluded.
This all increases the trust in the accurate instant speed feedback. As result, the over-all instant running comfort is really high.
Because the speed graph in the STraM is optimized for showing speed trends, it is impossible to detect correctly speed cracks in the speed graph of the STraM. This is shown in Figure 17.
Figure 17 Tiny experiment: 10 sec running, 10 sec stand still. Therefore the timers INT1 (interval time) and INT2 (recovery time) is used. (Top) is time representation of the speed, (bottom) is the distance representation.
One can observe the following:
1. First one notices the result of the different accuracy of the LAPdistances with regard to the distances at the sample times. It is possible that MARKER 1, 3, 5 is positioned before the vertical lines in the lower curve. But the 2nd, 4th and 6th MARKER should be positioned right on these vertical lines. If one is able to recording the LAPdistances also up to 1 meter instead of up to 10 meter, this ‘truncation’ inaccuracy can be solved. Another solution is to recalculate the LAPdistances by interpolating the most accurate sample time distances. This is done in the t6 Swiss Army Knife.
However this ‘truncation’ difference does not accumulate during the workout, and it also is in general always smaller than 10 meters
2. From this graphs one can easily observe the impossibility to detect the speed crack latencies in the STraM. Therefore one is forced to rely on the distance increments at the sample times.
In Figure 18 the result of the backward trapezoidal rule is used to show which time samples do not have distance increments.
Figure 18 (blue) STraM running speed profile; (red) speed profile obtained with a backward trapezoidal integration rule.
It is clear that the transient tine from running to stand still (deceleration to zero speed) is not zero. I guess this takes about 2 seconds. The acceleration however is a faster process. Then the red speed graph proves that the t6 always knows within 2 seconds if there is a change in distance or not. This means that, if a sampling time of 2 seconds is used, that the t6 records the proper distance increments within 2 seconds after a speed crack.
Walking-stand still experiments prove that these observations are correct.
Also it can be proven that acceleration from any speed or a deceleration to any speed shows the same results: the maximum latency in the data file of the t6 is at maximum 2 seconds for a sudden speed crack.
An example: Assume you calibrated on a 400m track, at average speed 12km/h, and the footPOD attached on you right foot. Then a high speed race at 15km/h on asphalt with the footPOD attached on the left foot, will in general result in inaccurate instant speed report and distance recording.
The Calibration factor depends on several factors:
1. the average speed you had during the calibration run;
2. the type of shoes you used during your calibration run;
3. the type of surface you ran on during your calibration run;
4. the position of the footPOD on your foot during your calibration run;
5. the left or right foot on which you attach the footPOD (for the same body speed they have in general a different acceleration profile)
6. additional conditions during your calibration run, like
a. heavy front wind / heavy back wind;
b. long uphill tracks / downhill tracks;
c. additional weight or ballast;
d. unstable running surfaces (slipping and sliding); …
The accelerometer based measurement principle imposes that, in order to get accurate distance recordings for your workouts, these workouts should be done with a calibration factor that is obtained in similar running conditions as you did the calibration. Apart from point 6 (which I consider as occasional disturbances), the influence of the other points can be displayed graphically.
The calibration chart shows graphically the influence on the CF from the first four points of the above Section. This graph offers a view of the most practical influencing factors: the speed, running shoe type and running surface dependency.
To obtain the calibration chart, a staircase speed protocol is proposed in Section 10.3.1. The t6 Swiss Army Knife automates the calculations and graphing (Section 10.3.2).
If one is only interested in a ‘single shot’ measurement of the calibration factor, then the automatic calibration method of SUUNTO is also the best calibration method. SUUNTO advises to calibrate with a steady speed and with flying start and stop conditions.
This automated method is very easy:
1. Do calibrations run with a steady speed and with a flying start and stop condition.
2. after the stopwatch is paused at passing the finish, you do two simple steps:
a. Navigate into the CALIBRATE submenu in the SPD/DST DISPLAY MODE, and choose footPOD. Then you will observe the used calibration factor.
b. If you press on SUUNTO, the t6 will display the actual measured distance up to 1 meter of accuracy. Just change this distance and press again on SUUNTO. The calibration factor is then updated instantly.
The recorded distance data will not be changed, and the recorded data can be uploaded to the STraM. Only future distance recordings will use this updated calibration factor.
Several tests show that this method will result in accurate calibration factors for that specific average speed and contact surface running shoe-running surface.
Since supplementary the data transmission from the footPOD is digital (so almost no wrong data reception will appear), this ‘blind’ calibration method can be trusted. There is no need to investigate the speed profile in the STraM to discover distance recording anomalies.
Assume a workout with the assumption it uses a perfectly calibrated footPOD. This means that:
1. the workout has the same an average running speed as the calibration run average speed;
2. the same combination running shoe - running surface as in the calibration run is used;
3. all other conditions are also similar (position of the footPOD on the shoe, same wind conditions, …)
Then one can ask oneself what would be the expected accuracy for such a workout?
Extensive tests show that the expected accuracy is 99,8% or better. This does not mean the real accuracy can be worse. The 99,8% is a statistical value. But most of the time a deviation from this value is due to effects that are not covered with the original calibration run.
One example of such a difference is the speed profile. E.g. interval workouts involve speed cracks. As already discussed, speed cracks involve latencies in the data recording process, and hence this can harm the accuracy. However, since the latencies for speed cracks are really small (less than 2 seconds in the distance record), it is to be expected the accuracy will still be relatively high.
In Figure 19 a summary is given of a whole set of workouts for which the distance accuracy is measured. In this graph, 100% interval means that a complete workout has been done solely by using interval (running – walking). Both symmetrical (same acceleration and deceleration profile) as asymmetrical (different acceleration and deceleration profile) intervals are taken into account in this graph. It should be clear that asymmetrical speed profiles pose more accuracy problems to the t6 than symmetrical ones.
Figure 19 Expected accuracy as function of the running speed profile. With ‘interval’ is meant both symmetric and asymmetric intervals.
Another influence on the accuracy is the speed dependency. If one does not invoke speed cracks, it is also possible to have a different accuracy as is expected from Figure 19, right side.
The t6 Swiss Army Knife also provides support to calculate such calibration charts from recorded t6 data files. See Chapter 10 on the proper protocol, and the support to obtain the chart.
In Figure 20 such a calibration chart is shown.
Figure 20 Calibration chart obtained with the t6, STraM and the t6 Swiss Army Knife.xls
From these graphs one easily can observe the speed dependency of the CF. E.g.:
1. When I run on common street with my AsicsGel (red line) with a target average speed of 13km/h, then the ideal CF is 967 (see bottom line);
2. However, when I use this CF but run at an average speed of 10km/h, the used CF isn’t accurate any more, it should be set to:
CF = -3,3 * 10 + 1010,1 = 977
This means that there is a difference of -3,3 per different average km/h, in this case this is a difference on the CF of about 10 (hence about 10 to 1000 or 1%).
This means that if one uses CF= 967, the expected accuracy will not be 99,8% but 99,8% - 1% = 98,8%.
It should be clear that the speed dependency of the CF totally depends on the combination running shoe – running surface and your personal change of stride.
Some people report there is a big variation in distance report between steady long uphill hiking (running) and steady long downhill hiking (running). From this they make the conclusion that the equipment is inaccurate on hilly terrain. To my opinion this is wrong:
· First it is dangerous to assume that the uphill or downhill itself is the origin the different (relative) accuracy. Mostly the uphill part is taken with a smaller speed than the downhill part. This difference of the steady speed makes that the distance difference and instant speed is measured with a different relative accuracy.
· This effect as such has nothing to do with the quality of the sensor or the accuracy of the t6. By taking different shoes, or by doing the same test on a different surface type (as one can see in the chart shown in Figure 20), the relative difference will change due to a different negative slope. Supplementary, since stride (also influencing the slope) is user dependent, it is dangerous to conclude that big differences of the uphill and downhill distances are due to inaccurate equipment.
· To be able to compare qualitatively the t6 with other equipment on this issue, one has to exclude the effects of the specific ‘running shoe’, ‘running surface’ and ‘individual foot stride’ (for many people even left or right foot have a different stride for the same speed!). So one has to sample (with both equipments):
o a lot of people;
o for each person, a lot of combinations running shoe – running surface;
o Then the results of both equipments are to be analyzed statistically.
Member Deleu proved he has a big slope (-14 per km/h !). The speed differences of each LAP can’t be the source of the problem since r is quite good (-0,99), and also SD (2,7). Most of the time such big slopes are due to loose footPOD. He confirmed that he had tied the footPOD very tight.
As result there can be only two sources for such a big slope:
· his ‘typical’ personal stride
· his used shoes (New Balance 753) in combination with the 400m running track surface on which he did the calibration.
To exclude one of these, he can:
· do another calibration run with different running shoes on the same 400m track;
· evaluate the slope of the New Balance 753 on common asphalt (if that is his common running surface of course).
The slope can turn out to be smaller, but also bigger.
This size of slope is really exceptional. This again is not due to the t6 or footPOD, but due the contact surface type and personal stride.
Deleu informed that he also had a negative slope (-10) for a calibration chart with the S625X. According to me this is also rather exceptional, since calibration charts of S625X show in general positive slopes. This makes me think of the fact that ‘probably’ excessive slopes for one equipment probably also produce excessive slopes for other equipment. This again points to the fact that not the equipment, but external factors play ground for this.
People who do have such big slopes should use for each workout type an accurate CF that is adapted for the target average speed. If this isn’t done, the accuracy for different speeds will be influenced a lot.
In Figure 21 it is shown that also for running workouts the STraM offers a bright view on the trend of the running speed.
Figure 21 (blue) STraM speed profile; (red) speed profile obtained with a backward trapezoidal integration rule.
As is shown in Figure 22, it is easy to record different speed PODs after each other in one data file. The only thing one needs to be sure of is the fact that right before a re-Connect the heart rate is measured well by the t6 (see Section 184.108.40.206).
Figure 22 Cycling and running in one data file. The first part shows a cycling speed. At the MARKER the bicycle is stopped, and the running is started (the footPOD is activated manually at the start of the run). After the bikePOD is at least 10 m separated from the running position, the t6 is re-Connected again (press-and-hold ALT/BACK).
Measuring heart rates under water is demanding:
1. every heart rate belt type has problems while swimming:
a. Pool water (with its high chlorine content) and seawater are very conductive. This may cause the electrodes on the transmitter to be temporarily short-circuited, preventing the ECG from being detected by the units.
b. Strenuous muscle movement during hard training may cause a strong water resistance that shifts the transmitter to a location on the body where the ECG signal cannot be detected.
c. The ECG signal strength varies depending upon the individual's tissue composition.
2. The killing problem with the t6 is that it uses ANT-technology (radio frequencies) for the HR data transmission. These radio signals do not penetrate well water. Other types using magnetic induction do not have that problem.
Since the ANT transmission does not function well under water, SUUNTO advises not to use the HR belt during swim workouts.
A delayed recording of an ANT device is e.g. the record in one file of the SPD/DST of a running part after having done a swimming part.
Tests show that it is necessary to follow correctly the procedure to record ANT devices as is explained in Section 3.2.2. and 3.2.3. E.g., after having done the swimming part one will notice that it is possible to connect to the footPOD(even the speed will be displayed after the connect), but StraM will not show the running speed of the running part from the recorded datafile because that data is not recorded. It would be nice if the t6 warns about this instantly. At this time however, remember the steps to be done prior to the recording of the data.
Especially when the recording time is evolving already for some minutes, it becomes impossible to record a device signal that was not yet connected at the beginning. I have no proof, but I feel it is the memory initialisation (and the specific page writings) that plays a role here.
Based on this and the discussion already done in Section 220.127.116.11, the following can be concluded:
1. First, in order to do a delayed record of a SPD/DST ANT device:
a. the ANT device has to be set ON, then the t6 has to connect to the device properly, and the StopWatch has to run for at least a few seconds in order to initialise correctly the t6 memory;
b. Then the Stopwatch can be paused (this is not mandatory), and the SPD/DSTdevice can be set OFF;
c. The StopWatch can be continued, and zero distance increments will be used as recording value for the missing SPD/DSTdevice; (With some earlier firmwares and calibration factors bigger than 1000, an artificial speed will pop up on the display -32,2km/h-. This is harmless, but you can upgrade if you want to get rid of this)
d. When it is time to record the SPD/DST device, put the ANT device on, and connect the t6 to the device. This can be done without pausing the StopWatch. Once the device is connected, the distance increments recorded are no more 0, but stemm from the connected ANT device.
e. When one wants to change to another SPD/DST device, one can redo step 1.d. as is discussed in Section 6.1.
2. Second, it is impossible to do a delayed recording of the HR.
The reason for this inability is discussed in Section 18.104.22.168 (time shifting effect). This is the reason why swimming and running can’t be recorded into one file. Maybe future firmware will do, but actually the re-connection to a HR belt is impossible once the connection with the HRbelt is lost for some time. And if the re-connection is possible, there will be a time shifting in the HR data.
In Figure 7 of Section 22.214.171.124 one can observe a seamless entire day recording where the HR-belt is kept on for the whole recording time (mandatory), and several times the bike POD is used and one time the footPOD is used in the same record.
The t6 contains only the very basic programming features of a Personal Trainer. I think this is influenced by the fact the t6 and STraM is EPOC oriented. However, all the basic functionalities are available and can be programmed without any learning curve.
The STraM software is also easy to handle and provides sharp, consistent and accurate data.
A few extra features and additions would make the whole complete.
The t6 has 4 DISPLAY MODES (TIME, ALT/BARO, TRAINING, and SPD/DST). These modes are selected by pushing sequentially on the DOWN/LIGHT button or UP/LAP button. It is advised not to use the UP/Lap button for that purpose, since this button also serves as LAPbutton during a data record (i.e. when the StopWatch is running).
Each of these DISPLAY MODEs has a mode menu. This mode menu can be selected by pressing on the SUUNTO button (middle right) when you are in a certain DISPLAY MODE.
So, when you are in SPD/DST DISPLAY MODE, and select the corresponding menu, you will find all speed and distance related settings and programming features: Here you can find the autoLAP, distance INTERVAL, SPEED LIMITS, CALIBRATE, …. Two examples:
a. In the distance INTERVAL (INTERVAL submenu) setting there are two programmable distances (INT1 as interval distance, and INT2 as recovery distance). They act as in a carrousel.
There is no possibility to have hybrid distance-time-heart rate interval programming.
After every INT1 and every INT2 a LAPrecord is taken automatically.
b. The speed limits (SPD LIM submenu) are not limited to a certain speed margin. The behavior of the t6 makes it possible to run within a very narrow speed margin without invoking too many limit beeps. I was successful to run within a margin of 0,3km/h on a tartan track.
Remark that the upper limit and lower limit function differently: the lower limit is within the margin (no beeps), the upper limit is out of the margin (beeps).
Also, the Speed limits aren’t stored in the workout header.
If you have set the UNITS of the t6 in pace units (e.g. min/mile) then the speed limits are to be entered in the same units.
Similarly, the HR limits, TIMERS, stored workouts (LOGBOOK), can be found in the TRAINING MODE menu. Two examples:
a. The HR limits can be programmed similar to the speed limits.
The HR limits are stored in the workout header.
b. The timers are more elaborated than the distance interval counterpart: There is INT1 time, INT2 time (also with the number of Intervals to be done), WARMUP time, and COUNTDOWN time.
INT1 time and INT2 time act as in a carrousel.
The COUNTDOWN timer is a timer prior to the start of the StopWatch. It is no cooldown timer after the interval workout.
After every INT1 and INT2, a LAPrecord is taken automatically.
At one time, the t6 can only store 1 set of all of these data at a time. This means that in practice only one program set is in the memory of the t6 available. It would be nice if more workout templates could be stored in the memory of the t6. Changing a setting however is very easy and done fast. Probably this is the price to be paid for an easy to use wriststop trainer.
There is also a training type indicator column on the right side of the display. See Figure 23.
Figure 23 t6 display and buttons
This indicator column is only active when the DISPLAY is in TRAINING MODE or in SPD/DST MODE (so the figure is somewhat confusing).
The displayed value is a code value and reflects the t6 programming state:
1 If HR limits is set ON
2 If SPD limits is set ON
3 If COUNTDOWN TIMER is set ON
4 If WARMUP TIMER is set ON
5 If TIME INTERVAL is set ON
6 If AUTOLAP is set ON
7 If DST INTERVAL is set ON
This indicator column is only useful if you have good eyes (small font is used), and if you know the code.
Some settings can not be set simultaneously, e.g.:
1. When the SPD limits are set ON, then automatically the t6 sets the HR limits OFF.
2. Assume the WARMUP TIMER and/or TIME INTERVAL is ON, then setting the DST INTERVAL ON will automatically set the WARMUP TIMER and/or TIME INTERVAL OFF.
3. Assume the WARMUP TIMER and/or TIME INTERVAL is ON, then setting the AUTOLAP ON will automatically set the WARMUP TIMER and/or TIME INTERVAL OFF.
4. Assume the DST INTERVAL is ON, and then setting the AUTOLAP ON will automatically set the DST INTERVAL OFF.
Note that the COUNTDOWN TIMER is a countdown timer prior to the start of the Stopwatch. So, as long as the countdown timer isn’t 0 sec, the measurement data will not be recorded.
So this timer is not an interval or workout CoolDown timer.
It is possible to change
instantly these settings during a workout, i.e. when one has programmed the t6
with TIME INTERVAL ON, and then the
menu allows changing this setting into
The only setting that is able to reflect immediately in the instant guidance are the LIMITS ON/OFF. This is useful if one wants to turn on/off the specific LIMITS alerting signals (See Section 126.96.36.199).
Only the HR LIMITS (even when they are set OFF) are stored within the workout record. It is suggested to SUUNTO to store also the SPEED LIMITS and to provide a similar support in the STraM as is done for the HR LIMITS.
When the StopWatch is
started, the t6 makes a copy of the program template, and it guides the workout
according to that copy. If one wants to
change the instant guidance (e.g. one wants to change the AUTOLAP ON 1km to
The units (e.g. DST Miles or kilometers) can be changed instantly during the workout. However, this has to be done in the corresponding UNITS submenu of the TIME DISPLAY mode.
The instant guidance of the t6 is supported by the two DISPLAY MODES ‘TRAINING’ and ‘SPD/DST’.
1. The TRAINING MODE shows three lines of which the third line is user selectable:
a. Upper line is instant heart rate
b. Middle line is StopWatch time
c. Lower line can be selected by pressing sequentially on the ALT/BACK button:
i. Instant timer value of one of the active timers:
1. instant WARMUP time (countdown behavior),
2. instant INTERVAL timers INT1 or INT2 (countdown behavior),
3. Instant Laptime.
When the instant LAPtime is displayed and a LAP is taken, then the
ii. Time of the day
iii. The average heart rate since the beginning of the recording or lap
iv. Instant altitude
v. (if a speed sensor is connected) the instant speed
vi. (if a speed sensor is connected) the instant distance
2. The SPD/DST also shows three lines of which the third line is user selectable:
a. Upper line is instant speed/pace (the units used are the t6 UNITS (instant selectable in the submenu SPD of the submenu UNITS of the TIME DISPLAY MODE)
b. Middle line is instant distance (the units used are the t6 UNITS (instant selectable in the submenu DST of the submenu UNITS of the TIME DISPLAY MODE)
c. Lower line can be selected by pressing sequentially on the ALT/BACK button::
i. The instant heart rate
ii. COUNTDOWN time / StopWatch time.
iii. Average speed since the start of the Stopwatch.
iv. instant distance since the last LAPrecord:
1. instant INTERVAL distances up to DST INT1 or up to DST INT2;
2. instant LAPdistance
v. Time of the day
When a LAP is taken, then the lower line shows a few seconds the
The most used display is shown in Figure 8.
These displays can be set prior to a workout. So by using the DOWN/LIGHT button and eventually the ALT/BACK button, one can easily put the preferred instant workout measurement data on the display.
Other instant guidance on the display isn’t supported. E.g.
1. If one is doing an interval, there is not any instant report on the display which number of interval one is actually running, there is also no instant feedback of the programming state one is actually working in (WARMUP, INTERVAL, RECOVERY)
2. There is no feedback on the display which LAP one is actually running.
So you need to keep track of the instant programming state of the t6 in your mind.
The TONES setting (in the GENERAL submenu of the TIMES DISPLAY MODE) is about the tones produced if one pushes a button of the t6. When this setting is ON, and one presses a button, a beep is produced. This is a useful feedback when one pauses the StopWatch or want to take a LAP.
If by accident a sensor is disconnected, then the t6 will not produce any button tone during a short period. After a while (less than one minute) the button tones are produced again automatically. That short period of silence is annoying since there is no immediate feedback any more of selecting the StopWatch pause or LAP during a workout.
There is a shortcut to solve this problem faster (re-selecting the TONES ON setting in the GENERAL submenu of the TIMES DISPLAY MODE), but this is not a straightforward one during a workout.
Feedback of SUUNTO on this topic:
“…When the t6 is already connected to a sensor, it knows exactly the time slots when the radio communication takes place. When connection to the sensor is lost for some reason, the t6 starts to seek it again similarly like when connection is made manually before workout. When there is no connection, t6 cannot know the time slots when transmission takes place and thus it needs to keep the radio open until finds all the sensors. During this time there's is no power left for any other power consuming events, like sound or backlight. That's why they are prohibited during connection establishment. If this isn’t prohibted, the voltage level of the battery could get too low for short time and it may cause t6 to reset itself. Actually it's bug if it tones can be forced on from the menu…”
Every instant change of state of a programmed workout also produces typical alerts. These alerts can’t be set off.
E.g. When a COUNTDOWN, WARMUP, TIME INT1 and TIME INT2 is ON, then
· 3 short beeps warns the user that the COUNTDOWN is finished (there is something with that first of the three short beeps)
· 2 short beeps warns the user that the WARMUP is finished
· 1 long beep warns the user every time a TIME INT1 is finished
· 2 beeps warns the user every time a TIME INT2 is finished
E.g. When a COUNTDOWN, DST INT1 and DST INT2 is ON, then
· 3 short beeps warns the user that the COUNTDOWN is finished (here the first short beep sounds well)
· 1 long beep warns the user every time a TIME INT1 is finished
· 2 beeps warns the user every time a TIME INT2 is finished
Another type pf warning signals are the LIMIT beeps:
· long beeps are produced as long as the measured signal is below the LIMIT low;
· short beeps are produced as long as the measured signal is above the LIMIT high;
There is no other feedback that warns the user that he/she is training beyond the limits set.
The alerting signals of the active limits can be set OFF/ON instantly by press-and-hold on the button SUUNTO. In fact, the t6 toggles the LIMITS ON / OFF of the limits that were set ON last:
· If SPD LIMITS are set ON at the start of your workout, the long-press on the SUUNTO button will toggle SPD LIMITS ON / OFF.
· If both the SPD LIMITS and HR LIMITS were set OFF at the start of the workout, then the long-press on the SUUNTO button will toggle the LIMITS ON / OFF of the type that was set ON last.
When you then start the Stopwatch, the only programming setting that can be changed instantly during your run is the LIMITS ON/OFF (NOT its value –e.g. the speed limits values itself-, NOT its type –speed or heart rate type-. Therefore you can use the shortcut long-press-SUUNTO.
You can change all the other program settings instantly during running, but they become only active after a complete restart of the stopwatch.
Even with the HR LIMITS OFF, the t6 measures the time the heart rate is below/in/above the HR LIMITS. These values and the HR LIMITS itself are stored in the header information of the workout.
At this time the t6 does not manage the SPD LIMITS in a similar way. The only effect SPD LIMITS ON has, are the alerting signals. There aren’t any other values stored in the header.
In December 2004, a firmware update will solve this bug that produces INT 2 beeps that are not distinct enough with LIMIT low beeps. See Section 12.1 for additional comments.
Although the programming and instant training guidance hasn’t numerous options at this time, it is perfectly possible to do all the basic workouts. There aren’t much features, but all the features that are implemented do function correctly and accurately.
To show an example a distance based interval workout with speed limits is chosen.
The workout goals are: 16 times a 400 meter run at a speed between 15km/h and 16km/h; 100 meter recovery walking.
With the calibration chart (See Section 10.3) the most optimal CF is selected to cover the fast run accurately. For a speed of 15km/h and the specific contact surface, the optimal CF is 960
Figure 24 The calibration chart discussed in Section 10.3 calculates automatically the most optimal CF for the target running speed (see lower line: 15km/h) and the running contact surface.
In the SPD/DST DISPLAY MODE submenus the following settings are entered:
· INTERVAL ON; Int1 0,4km; Int2 0,1km.
As result the figures ‘2’ (limits ON) and ‘7’ (Interval ON) is displayed at the right side of the t6
1. Start the Stopwatch, and, to get rid of the speed limits alerting beeps by press-and-hold SUUNTO. I do my warming up (this can take 1 or 2 or 3 … complete intervals.
During that warming up, I set the t6 display on SPD/DST DISPLAY mode, and select the instant LAPdistance (with the lower left button) at the bottom line. Then I have on the display the speed on top, the instant total distance in the middle, and the instant LAPdistance at the bottom.
2. When I am warmed up, I start my first real distance interval at e.g. at the start of the third interval. Right after starting my first real interval I press-and-hold SUUNTO again to hear the limits alerting signal for my interval guidance.
3. When I finished my 16th recovery, I do my cooling down, and don’t bother the interval beeps any more.
4. After my cooling down I stop the stopwatch.
When a workout is recorded and closed, one can navigate to the LOGBOOK menu in the TRAINING DISPLAY MODE. In this logbook one can display the summary data of a workout (VIEW option) , or one can delete an entire workout as well (ERASE option). The LOGBOOK also shows the available free memory.
In the view mode, the workouts are selectable from a list showing the log date and log start time. By selecting one workout, the following summary data is displayed on several pages by pressing on the UP/LAP button successively:
· total workout time, number of laps
· average HR
· min/max HR
· time within HR range and the HR range at the time of the workout (also when the limits are OFF or set on SPD)
· time over LIMITS HIGH HR ( “ “ )
· time below LIMITS LOW HR ( “ “ )
· if a SPD/DST data is recorded: workout distance/average workout speed
· if ALT was set ON:
o low point
· Then for each lap:
o HR at LAPtime/avg HR
o if a SPD/DST data is recorded: LAPdistance/average LAPspeed
o if ALT was set ON: ascent/descent
The graphs in the STraM looks like is shown in Figure 25. From this figure one can observe that the EPOC isn’t much influenced by the warming-up or the cool-down. So if one don’t bother the total runned distance for interval workouts, one can split the workout in three pieces (pull-down menu ‘Actions’; then select ‘split log’ option), and keep the part that is displayed form 1 km to 9 km.
By splitting the log, the warming-up distance and cooldown distance, time, … is kept in the database. This will keep all the data in the running totals (weekly totals, monthly totals, …)
Figure 25 Distance based interval with speed limits report in StraM. By default, the HR and EPOC graph is shown as function of time. Here the graphs are shown with respect to running distance. The instant VO2 and running speed is also displayed.
Splitting the interval part (or recording only the interval part in the t6) has the benefit of comparing successive similar workouts in