carleton

Yeah, I've wondered about how back pressure affects them. I'm not exactly sure.

Dalai

With a freehub, there is negligible back pressure on the pedals. Back pressure = back pedaling. Therefore road PM's don't need to worry about any negative force on the pedals like than encountered on a track bike. If the PM isn't developed to ignore this back pressure, it will record this as watts as it just recognises force being applied to the pedals.

Quinn8it

I've got an older Wired SRM that's back in for service for the 3rd time since October. I'm having issues related to back pressure and zero offset. Basically my zero offset number jumps erratically between efforts.
SRM thinks it's "opporator error" and sent me a complicated Zero'ing technique to follow. Not only can I not find anyone else with one of these that has to do this Zero'ing process, but when I followed it to the letter it still didn't work..
I'm currently back on the Stages - which has its own set of issues...
IMHO- there is no perfect Track PowerMeter...

rapwithtom

What are the issues with Stages? I'll be in the market for a track power meter in the next month or two...Stages seems the best compromise right now?

Dalai

@Quinn8it is your SRM a road unit being used on the track? As I've heard using roads units fixed can be a bit hit or miss. My wired Science Track unit still works flawlessly as do my wireless and wired units running non fixed gear...

queerpunk

there's always a 110bcd Quarq + a Bergstrom Technologies chainring adaptor.

Dalai

If I was shopping for another track PM and couldn't find a used Track SRM, I'd consider a Track P2Max unit.

Voodoo76

I switched to Power2Max on the road (from Quarq and PT prior to that) about a month ago. Seems like a very solid unit, zero issues setting it up. I was concerned that the accelerometer based cadence would be flakey compared to the Quarq but so far so good. Was working at 140rpm+ today w/o issue.

wens

No 167.5 cranks

I'm a bit concerned about the accelerometer on my short home track too

Quinn8it

the SRM I'm having issues with is a 2008 wired track version.
i also have an even older wired road SRM that is on my ergo, set up with a freewheel and zero issues.
the SRM thats having issues is likely due to inevitable back-pressure forces from slowing and getting off the bike not resetting after a forward pressure and zeroing is done. we'll see, SRM has had it back for a couple weeks and no word yet..

all in I've used:
Power Tap wireless Road/ Power Tap wired Track/Stages Track Dura Ace (on my 3rd)/ Stages Road Sram/ Quarq Sram Red Road/ SRM Wired Track/ SRM Wired Road

the Stages is decent. I've had 3 of them. the first was stolen, it actually worked the best. the second had huge issues with massive (25,000watt) spikes on jumps and standing starts. Stages sent me the 3rd, and refunded my money (for the 2nd) due to all the issues i had with power spikes.. the 3rd unit will still spike- but typically in a "roll-up Standing Start" where you slightly back-pedal and jump... that creates a hammer effect that spikes power..same thing is common on SRMs, sometimes even just getting on or off the bike.. its very easy to fix in Golden Cheetah.

other issues:
its recently started dropping data while on the rollers.. presumably due to low pressure on the cranks. this is always in ANT+ mode and happened on 2 different garmin 500s, yesterday i tested it in Bluetooth mode on a LYZENE computer and had no issues..

cadence is spotty and prone to dropping out especially at high RPMs.. its gotten better over the years with FirmWare updates, but the option to use a standard cadence sensor and override the accelerometer would be great.

its annoying that there isn't a great computer available. the Bluetooth platform is capable of reading many times per second, but there is no readily available computer to match that.. 1" reads is pretty slow..

carleton

For the SRM and other power meters, cadence is derived from the power meter.

Basically, the power meter turns on and starts transmitting when you put a certain amount of torque on the cranks. It "wakes up" automatically. But, then goes to sleep and saves the battery when you aren't pressing hard enough after x minutes.

When the cranks come around and the magnet trips the reed switch it delivers the power data to the head unit.

Notice how there is no mention of cadence yet.

The head unit looks at how often power data is arriving and says, "Hmmm power data is arriving 2 times per second. That's a rate of 120RPM. I'll record the cadence as 120RPM." This is why the SRM software had a switch for you to indicate whether you had 1 or 2 reed switches in your cranks so it would know whether the readings were coming once or twice per crank rotation. This is why if you had 2 reed switches but had he software set on 1 reed switch, your cadence and power would be doubled (power is a factor of cadence, too).

So, guess what happens when you are riding rollers and aren't pressing hard enough to keep the power meter awake? Not only does your power data stop transmitting, your calculated cadence data doesn't get calculated either.

carleton

You are right. There is no perfect track power meter. They are all made for road first and adapted for track. Road and TT is where the money is.

I wish I knew how to design and manufacture them. I know what I'd like out of one. But, I can't make it.



hmmmm...that same thinking is what inspired me to make my apps...

carleton

Nerd Correction: I phrased it that way to tell the story easily.

What really happens is that it looks at all of the recorded data from the power meter and looks at the time BETWEEN each reading and calculates the cadence.

If the records are separated by 1.0s, then that's 60rpm
If the records are separated by 0.75s, then that's 90rpm
If the records are separated by 0.5s, then that's 120rpm

So, it's the time between each power meter reading that determines what the cadence is at that time.



I know this because a while ago I made some prototype Bluetooth head units...then realized that the world didn't need another one

skysurfer

We need to get together one of these days... I'm an EE good with hardware but a terrible programmer. As I type this I'm at my shop resting my arm from shearing core material for an eddy current brake experiment.

Due back at DLV in a couple of weeks for the cert course as a refresher and hopefully no embarrass myself and leave feeling alot better about it this time.

Voodoo76

One of the things I liked about the Quarq design was that it has 5 reeds. You should be able to calculate rpm 5 times within a crank revolution. I have no idea if Quarq uses them this way or not. An external sensor (or 1 internal reed switch or an accelerometer), is measuring the cadence at the end of each revolution as you point out. So at low rpm you are averaging part of the power equation over a pretty large chunk of time.

I will say that looking at P2M raw files so far doing accelerations (keep in mind I'm not a pure sprinter) the rate of change in rpm doesn't look much different than my Quarq data.

carleton

The fundamental shift will have to come with something not Bluetooth.

The limitations of how often the head units read is with the Bluetooth spec. Basically, it's capped at 1 or 2Hz (I can't recall). I think ANT+ is similar. It's been a year and a half since I was knee-deep in that stuff.

So, you'll have to make a dedicated head unit, too, if you want 2Hz or faster recording.

Quinn8it

You sure?
The Stages iPhone app has a "high speed" mode that captures an insane number of samples per second- and that runs on Bluetooth.

carleton

I can't recall the details of the limitation. I'll look back in my notes.

The specifications can be found here: https://www.bluetooth.com/specificat...specifications

Voodoo76

I always assumed that 1sec data recording was driven by .fit or .tcx file formats (and all of the analysis software & websites that use data in 1 sec intervals) rather than hardware limitations.

taras0000

Is it possible to maybe make a unit that sends the data in packets at 1-2Hz? Something like the crank takes its input over the course of a pedal stroke, with something like 20 readings, and only sends these packets to the head unit once or twice a second. The head unit then averages out the data on screen, but the raw info can be downloaded and lots of information made available.

carleton

That's a great idea!

I need to brush up on my knowledge of the spec so that I can answer these questions. It was Summer/Fall 2014 when I was playing with the idea.

taras0000

I mean, it means you would have to have more than just strain gauges and a transmitter in the crank, but with something that is mounted like a Quark, there should easily be enough room for strain gauges, a small microprocessor, mem card, transmitter, battery.

Quinn8it

You guys are a trip! I've missed you!

I'll be lifting weights and riding kilos- let me know when you've got a PowerMeter for me to test!

carleton

Yeah, exactly. So instead of transmitting instantaneous readings, it transmits a payload of readings.

Dude, this is entirely possible. Plus, this would enable a new level of analysis. If you add angle to the payload, one would be able to see how much force is being applied at that angle. But, it would be limited to 180 degrees being that the other crank takes over at that point.

But, from my limited knowledge, what you are suggesting seems to be possible.

It may not be feasible for road use with really long saddle-time on really long rides. But, on the track, where saddle-time is a fraction of that of the road (especially for time trials) it may be feasible.

Maybe have it on a switch from the head unit: "Detailed Logging" where it activates the mode only when you want it, to save battery and whatnot. Maybe a red light that blinks for "Recording".

I really don't think trackies need the "always on" recording that roadies like. I guess it goes into the TSS bucket. But, for detailed analysis, every trackie knows when they are about to do an effort. If so, press the big red "Record" button.

OK...everyone who reads this has agreed to Taras and I's non-disclosure agreement

carleton

We've missed you, too, man.

This place is like an old neighborhood bar. We'll always be here and you'll always be a regular, even if it's been a while