Searching \ for '[EE] Polar protocol' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=polar+protocol
Search entire site for: 'Polar protocol'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Polar protocol'
2009\04\28@124425 by Dr Skip

picon face
Does anyone have any info that would help understand the wireless protocol or
signal Polar uses for their heart monitors? I've got a frustrating situation...

Some background:
I have an elliptical exercise machine that claims to receive the signal from
Polar chest strap transmitters. It can use this to vary the difficulty to keep
your pulse within a range, or just display it with other data on the screen.
This seemed to be an important feature to the retailer selling it. ;) The unit
is not new enough to return, and manufacturer tech support is via phone, with
employees that believe all users are idiots and incapable of understanding
technical issues, while not having a clue as to what their product does...

The problem:
Using the chest strap with the watch it came with, the pulse reads accurately.
When using it with the elliptical, it reads about 30% higher than actual. I
could see missed beats, but this is odd. It _seems_ to read about 3, then get
an in-between pulse signal based on the heartbeat indicator. It isn't 100%
consistent, but close. I've turned off all other electronic devices within 25
ft of it, and tried without the watch on (just the elliptical display) and no
difference - it doesn't seem to be an interference issue.

Needless to say, the elliptical manufacturer points to Polar and the strap as
the potential problem, Polar says the strap is OK as it works with their watch
and that's all they care about, and the retailer now seems to think that
feature is 'unimportant' and not worth worrying about.... HA!

It's not worth a lawsuit as I can use my watch and manually play with
intensities and such, but it's also an real-world engineering problem of sorts,
and as such, solving it would add to one's body of knowledge (as well as reduce
some of the rest of this body ;) The documentation is non-existent in this
case. The overage in pulse reading is consistent enough that it does follow the
real pulse on average. However, the number is high enough that I can't use it
for the system's functions such as varying difficulty because it's higher than
the range can handle.

So, has anyone played with these before? Any ideas?

TIA,
Skip

2009\04\28@125521 by M. Adam Davis

face picon face
You might find some useful information here:
http://rick.mollprojects.com/hrm/index.html

And you can buy a polar HRM receiver as an electronic module here:
http://www.sparkfun.com/commerce/product_info.php?products_id=8660

-Adam

On Tue, Apr 28, 2009 at 12:44 PM, Dr Skip <spam_OUTdrskipTakeThisOuTspamgmail.com> wrote:
{Quote hidden}

> -

2009\04\28@142547 by Tony Smith

flavicon
face
> Does anyone have any info that would help understand the wireless protocol
or
> signal Polar uses for their heart monitors? I've got a frustrating
situation...
>
> Some background:
> I have an elliptical exercise machine that claims to receive the signal
from
> Polar chest strap transmitters. It can use this to vary the difficulty to
keep
> your pulse within a range, or just display it with other data on the
screen.
> This seemed to be an important feature to the retailer selling it. ;) The
unit


There's at least two protocols.

The newer devices have a serial number that's matched to the receiver, eg
the watch.  Your chest band will only work with your watch.  The older ones
don't.

That said, the newer ones are supposed to be ignored by the older equipment
(wrong protocol, presumably).

Olin has done some work with this sort of stuff.

Tony

2009\04\28@171346 by M. Adam Davis

face picon face
On Tue, Apr 28, 2009 at 2:23 PM, Tony Smith <.....ajsmithKILLspamspam@spam@beagle.com.au> wrote:
> There's at least two protocols.
>
> The newer devices have a serial number that's matched to the receiver, eg
> the watch.  Your chest band will only work with your watch.  The older ones
> don't.
>
> That said, the newer ones are supposed to be ignored by the older equipment
> (wrong protocol, presumably).

I thought they both used the same 5kHz signal burst, but the older
straps were simply dumb bursts with no data, while the newer bands
encoded data (ASK?).

This allowed the old receivers to still 'see' the bursts from the
newer straps, they would just ignore the data.  The newer receivers
should still see the old bursts and report them as older style with no
serial numbers.  So by and large the systems are stil compatible, but
I don't recall where I got this information from and I could be
mistaken...

-Adam

2009\04\29@041800 by Tony Smith

flavicon
face
> > That said, the newer ones are supposed to be ignored by the older
equipment
> > (wrong protocol, presumably).
>
> I thought they both used the same 5kHz signal burst, but the older
> straps were simply dumb bursts with no data, while the newer bands
> encoded data (ASK?).
>
> This allowed the old receivers to still 'see' the bursts from the
> newer straps, they would just ignore the data.  The newer receivers
> should still see the old bursts and report them as older style with no
> serial numbers.  So by and large the systems are stil compatible, but
> I don't recall where I got this information from and I could be
> mistaken...


I figured they were deliberately made incompatible so the newer coded ones
wouldn't interfere with the non-coded stuff.

>From my limited observations that seems to hold up, a lot of people make
equipment to pick up the signal, but they state it's for non-coded chest
straps only.  I've got a couple of coded straps, and they don't work.

Tony

2009\04\29@081236 by olin piclist

face picon face
Dr Skip wrote:
> Does anyone have any info that would help understand the wireless
> protocol or signal Polar uses for their heart monitors? I've got a
> frustrating situation...

I did the firmware for a bicycle ride computer a long time ago that had to
interpret this signal.  The product was the PowerTap by Tune, which was sold
to Graber a years ago.

The signal was just a carrier pulse every heartbeat.  They used a very low
frequency carrier, like 5-8KHz I think, so it's really more magnetic
coupling.  I think they used a receiver unit from Polar or maybe Sports
Instruments, which produced a digital signal that indicated when carrier was
detected.  The PIC code had to take this signal and determine heart rate
from it.

The problem is there is a *lot* of noise on this signal.  There are missing
pulses, spurious pulses, real pulses split in two, etc.  The trick was some
clever non-linear filtering and algorithms to take into account what a human
heart can and can't do.  After probably a week of tweaking I eventually got
it to respond a little faster than the Polar watch while not giving
erroneous readings.  It was a long time ago and I don't remember the
details.  One trick I remember though was to keep a small history of recent
readings, like maybe 5-8 of them.  The reported value always had to be
within the min/max range of the stored readings, despite what the low pass
filter said.  This is how I got fast response to true changes.  Much of the
code was about detecting spurious and missing pulses and dealing with them
accordingly.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.


'[EE] Polar protocol'
2009\05\14@180157 by Dr Skip
picon face
Thanks for all of the information in replies. I had some other issues, so
haven't had a chance to mention it until now.


-Skip

2009\05\14@190423 by Chris McSweeny

picon face
On Wed, Apr 29, 2009 at 1:13 PM, Olin Lathrop <olin_piclistspamKILLspamembedinc.com>wrote:

{Quote hidden}

I suspect these issues are fairly generic to HRMs. Certainly much the same
problems with analysing the R-R data recorded by my Suunto T6c HRM,
requiring similar solutions (determining min/max values is pretty important
- simply limiting these to sensible values seems sufficient to get rid of a
lot of spurious data). Presumably the HRM itself has protocols much like
these internally.

Chris

2009\05\16@230527 by Dr Skip

picon face
Progress!

It seems my attempts were with a coded transmitter. I suspect, after discussion
here and careful observation in use, that somehow the coding causes an
alteration of every third pulse - maybe enough data that it could appear as a
separate HR pulse. It was pretty consistent about getting a double pulse
reception every third. What I've found so far (and this may not include any
high end models with fancy authentication) is that a 'coded' receiver will sync
with a particular transmitter, then ignore the others it may receive. It
doesn't seem to be locked in. I may be wrong, as this was just gleaned from
researching my particular problem, but that's what I turned up.

I was able to get hold of a T34 Polar chest strap (which Polar has no info on -
it's probably old) which is not coded, and it reads properly. The bad news is
that the battery is not user replaceable on it. The cost of the unit is much
more than a simple battery, and their return form asks for a $75 authorization
or check, and they'll just take what they need to do what they think best...
ARG! There are a lot of folks it seems who have been charged for 'other
transmitter problems' when they send it in for a battery job and suspect
they've been had in the process.

Thanks for all the pointers on this.

A related question is; has anyone tried to open one of these transmitters up
and do it themselves (when the time comes)? It's quite sealed and molded. A few
search items refer to others who may have done it, but I've not been able to
find first hand definitive admissions (with the procedure described as well ;) .

Thanks,
Skip

More... (looser matching)
- Last day of these posts
- In 2009 , 2010 only
- Today
- New search...