Polar Power Kit - Install Help Please

RidwarePhil said:

Hi Guys,

A few weeks ago i unvested in a polar S725 HRM and Power kit. I carried out the install following the guide and pictures on polar USA and from some of the links suggested by other members of this forum.

Generally the system is working but I am seeing large variability accross the gears. So a couple of questions for those of you that have systems installed correctly to try and debug the system.

I'll see if I can help. To be honest, IMHO that "guide" from Polar is somewhat helpful, but there's a good bit of misinformation as well...but first, how are you observing this variation across gears?

RidwarePhil said:

1) In a small rear cog and a small front cog is the chain parallel to the power meter when viewed from the side.

Parallel doesn't matter. What matters when looking from the side is that when the chain is on the large-large gear combo, it is never further than 25-30mm from the part of the sensor that's marked "middle". That's the centerpoint of an ~1 in. square inductive sensor inside of the module. In fact, in most installations, the small-small combo will end up rubbing on the top of the case...but you shouldn't be using that combo anyway due to cross-chaining 😉

I can't emphasize ENOUGH how important it is to not violate this vertical spacing requirement in getting repeatable and reliable power readings.

RidwarePhil said:

2) In a large rear cog and a small front ring is the chain parallel to the inside edge of the power sensor when viewed from the top.

Again...parallel doesn't matter. In this case, what matters is that the chain passes over that ~1 in. square area described above when in this combo.

RidwarePhil said:

3) In a small rear cog and a large front cog is the chain parallel to the outide of the power sensor when viewed from the top.

Once again (thanks A LOT Polar!) parallel makes absolutely NO difference. See the answer to #2 above.

RidwarePhil said:

4) What range of gears are you using I am running 49:53 and 12:23 on a 9 speed Ultegra setup.

I've used mine mostly on my road bike which is 53:39 in the front with either a 12:25 or 12:27 in the back. I've even used mine mounted to the UNDERSIDE of a raised chainstay MTB with no issues either...aside from certain gear combos rubbing, which I didn't need to use anyway.

RidwarePhil said:

5) My bike is fairly small and has a traditional steel geometry. I cant get the centre of the power meter exactly midway between the BB and rear skewer as it clashes with the frant mech. What type of error will this introduce.
I assume this will introduce a DC power error

No problem. That's a ridiculous instruction by Polar. The position of the sensor is going to be driven by the cadence magnet location on the crank. Having the sensor slightly forward or rearward of the center of the chainspan won't affect the power calculation one bit since the frequency of vibration that the sensor is detecting doesn't change, just the magnitude of the vibration amplitude. The frequency signal will still be the same, it just may not be as strong. But, if you hold to the vertical spacing requirement I emphasized above, this will be a non-issue.

RidwarePhil said:

6) I mounted the power sensor on the chainstay using a moulding compound that sets hard to enable me to get a flat surface on the chainstay. Is anyone else using a hard mount or are most people using the rubber pads supplied.

That's a clever solution. On my 2 current setups, one chainstay just requires a very thin piece of bar tape under the sensor, and on the other I just use a short length of 3/8" diameter automotive fuel line as the "pad". Works great.

RidwarePhil said:

If anyone thinks that they can help I would appreciate any advice just PM and I can discuss this directly. I would also appreciate other any pictures of known working installations.

Cheers Phil

No pics (I really should document some of this stuff, huh?)...but I hope my input above helps you out. My tips above are based on almost 3 years of experience using the Polar power module and from personal correspondence with one of the original inventors of the device.

Tom

Gee Tom -I like using the small ring small cog combo because it's better than an equivalent big ring big cog combo which puts the chain farther away.

All I can add is that a neodynium magnet glued to the pedal shaft is the best cadence magnet solution. These magnets are so powerful you can get a thin one that won't stick out and will not get ripped off when you're in the big ring small cog combo. Then mount your rear speed sensor mount down; if mounted the other way if the sensor is nudged then it'll break into your spokes costing you a $25 repair at Polar which happened to me.

JTE83 said:

Gee Tom -I like using the small ring small cog combo because it's better than an equivalent big ring big cog combo which puts the chain farther away.

"Better" in what way? For typical gearing, the equivalent (or near) gear ratio using the large chainring will be more centered on the cluster and has less frictional losses and lessens chain wear.

Besides that, on my bikes anyway, it's pretty tough to prevent the chain from "tickling" the big chainring in the small-small combo. That's annoying 🙂

I have just completed another test to analyse the polar power sensor on my turbo trainer and noticed the following strange effect.

I completed the test in the small ring at 16mph and in the large ar 20mph.

power starts high in 2nd gear, then decreases for 3, decreases again for 4, decreases again for 5th, increases for 6th to a power level near for 2nd, then reduces for 7th, reduces again of 8th and reduces again for 9th.

This pattern was near identical for the big and small rings.

Could this be a speed sensor / wiring effect rather than a power sensor positioning error. I think if it was a power sensor issue I wouldnt expect it to ramp down then back up again.

Any more advice please?

Tom Anhalt said:

"Better" in what way? For typical gearing, the equivalent (or near) gear ratio using the large chainring will be more centered on the cluster and has less frictional losses and lessens chain wear.

Besides that, on my bikes anyway, it's pretty tough to prevent the chain from "tickling" the big chainring in the small-small combo. That's annoying 🙂

Do you mean the chain ticking on the front derailleur? I discovered that you can partially click the left STI Lever (both Shimano and Campy) to slightly adjust the front derailleur position to avoid chain rub. This will help avoid the front derailleur from ripping off due to fatigue failure which happened on my Raleigh Prestige.

I have 4 bikes with the old power kit. My new CF Soloist will have the new Polar kit!

RidwarePhil said:

I have just completed another test to analyse the polar power sensor on my turbo trainer and noticed the following strange effect.

I completed the test in the small ring at 16mph and in the large ar 20mph.

power starts high in 2nd gear, then decreases for 3, decreases again for 4, decreases again for 5th, increases for 6th to a power level near for 2nd, then reduces for 7th, reduces again of 8th and reduces again for 9th.

This pattern was near identical for the big and small rings.

Could this be a speed sensor / wiring effect rather than a power sensor positioning error. I think if it was a power sensor issue I wouldnt expect it to ramp down then back up again.

Any more advice please?

Aaahh...I see. That's why I asked my first question ("How have you observed this variaion?"😉 You've discovered something that was observed shortly after the original Polar power modules came out. In a "constant speed" power test on a typical trainer, you'll be very likely to observe such gear-to-gear variation. Some have speculated that it's a vibrational issue (resonant vibrations of the trainer structure) but in my opinion, it's more of an issue with the type of test.

Bascally, in order to do this test, you're going to end up with combinations of low force and high cadence. With the lack of "random" road inputs to excite the chain at it's natural frequency, it will become more likely that the chain vibration sensor will mistakenly "lock" onto the signal of chain pins passing the sensor. This will give you power "errors".

I've done this test and seen the same thing. However, I've also done this test on a Computrainer set to "erg" mode (constant power) and the gear-to-gear variation was not present. In other words, with "normal" chain tensions and cadences, this wasn't an issue.

So...what you observe is not unexpected, and *could* be somewhat affected by the sensor position in the extreme positions. I highly recommend you check your setup to the tips I posted above.

BTW, "head-to-head" tests of the Polar system against SRMs and PTs "on the road" do not show any gear-to-gear variation and all 3 PMs track very well, so don't worry that this is something you'll see on the road...with a PROPERLY positioned sensor, that is 😉

JTE83 said:

Do you mean the chain ticking on the front derailleur? I discovered that you can partially click the left STI Lever (both Shimano and Campy) to slightly adjust the front derailleur position to avoid chain rub. This will help avoid the front derailleur from ripping off due to fatigue failure which happened on my Raleigh Prestige.

I have 4 bikes with the old power kit. My new CF Soloist will have the new Polar kit!

Nope. I meant the chainring. I'm aware of the "trim" function on the STI levers 🙂

I think it's a combination of 53/39 chainrings and short chainstay lengths on my bikes that conspires for this to happen.

In any case, even if it didn't touch the chainring, I'd still be annoyed by just the extra chain noise in the small-small combo due to the chain having to shift laterall so much. Again, this is more prevalent on short chainstay frames. If I can hear it, that can't be good from a power loss standpoint 😉