Electronics Gurus - Stages Power Crank

carpetrunner

Likes Dirt
Try asking Dr Google "stages power teardown" image search
There appear to be several different PCB versions and bridge configs - which may have different colours :(
The OP looks like a SMP1.

how about this one;
352300
 

Nambra

Definitely should have gone to specsavers
Nice find @carpetrunner! Those images are also here, which looks to be related to some sort of FCC compliance check (probably making sure that the Bluetooth radio emissions meet the FCC regulations).

That certainly looks like @Ackland's circuit board and even the 4 wires are the same colour - definitely a good starting point.
 

Nambra

Definitely should have gone to specsavers
Another test document reference showing similar wire colours, although the strain gauge setup is 4 individual sensors. However, you can just make out some labels on the little circuit board in the middle at each of the wire connections. looks like Red is VCC (ie. + supply), Green is GND (ie. 0V), and White and Green are 'B+' and 'B-' respectively, so the sensing wires across the strain gauge bridge.


352301
 

Dales Cannon

lightbrain about 4pm
Staff member
Another test document reference showing similar wire colours, although the strain gauge setup is 4 individual sensors. However, you can just make out some labels on the little circuit board in the middle at each of the wire connections. looks like Red is VCC (ie. + supply), Green is GND (ie. 0V), and White and Green are 'B+' and 'B-' respectively, so the sensing wires across the strain gauge bridge.


View attachment 352301
That is what I would expect. Have had some experience with similar bridges for weighing many many tonnes of black stuff. Surprisingly robust and surprisingly fragile at the same time.
 

Nambra

Definitely should have gone to specsavers
So a quick update on this. @Ackland and I took this offline a couple of weeks back and he sent the Stages SPM1 crank to me as I offered to take a look at it for him.

Long story short, it's working again now although I took a bit of a roundabout way getting there. @Ackland was given the correct information from the Australian distributor as to which wire went where on the PCB, but I suspect that during the original repair, a small solder bridge between two of the connections stopped the unit from working. As I'd more or less disconnected it all to check the strain gauge resistances while I was working on it and then re-soldered the connections, it must have resolved the issue. My inexperience in using a power meter also threw me a bit - you definitely need the crank on a bike with a fair amount of pedal force to actually register readings on the phone app. It was probably working sooner than I'd thought, but my on-bike testing wasn't good enough.

For the record, correct wiring sequence as shown below; red, green, black and white from the top.
353157


If anyone wants the nerdier details I'm happy to share - I know a lot more about power meters now.

Hopefully @Ackland manages to find a robust way of reattaching the electronics & housing back to the crank and be back to cranking out the Watts again soon.
 

Nambra

Definitely should have gone to specsavers
I'm still laughing at the response I got from Stages, after telling them how much I hate throwing stuff out when it can easily be fixed... basically they said something to the effect "oh don't worry, all damaged power meters get sent back to us for refurbishment, so it won't get junked..." ie. you go buy a new one and we'll fix the old one...
 

slider_phil

Likes Bikes and Dirt
Does it need a factory calibration after it's reattached? Not the usual slope Cal in the app but the one where you hang a calibrated 50kg weight off the crank.

It's probably within 5/10% once reattached anyway, but might be worth a Google.
 

Ackland

chats d'élevage
I'll see what numbers its putting out when I get it back and decide whether I need it calibrated.

Realistically, the value in power meter data isn't the absolute numbers as long as its consistent.
 

cokeonspecialtwodollars

Fartes of Portingale
As you've stated for this application exact calibration isn't necessary, consistency is key. It would be interesting to know how Stages calibrate them though as strain guages typically work on forces applied along a single vector.
 

Nambra

Definitely should have gone to specsavers
Now that I’ve basically reverse engineered the device, I’d offer that calibration is probably not necessary. It does work on a full bridge strain gauge with temperature compensation. Each of the 4 resistance elements on the strain gauge are 1000 Ohms, a much higher resistance than the wires connecting it to the circuit board (so soldering it back together is unlikely to have affected factory calibration at all). There’s a precision instrumentation amplifier that conditions the strain gauge signal for the microcontroller that combines it with cadence (via an accelerometer), does all the smarts and determines the power output.

The Stages phone app has some diagnostic info such as firmware version, serial number etc., as well as what appeared to be two factory calibration values, one for temperature the other for power from memory. Then there’s the “zero calibration” function that sets the rest state signal of the strain gauge - you have to have the crank in the 6 o’clock position for the zero function to work as it checks crank position using the accelerometer. Unless the bond between the strain gauge element and the crank arm somehow failed, it should remain fairly accurate and certainly consistent over time which is really all that @Ackland wants.

Of course if he tells me that he’s now putting out 4.7kW on the flat after getting it back on his bike, I’ll have to eat all of the above.
 

nzhumpy

Googlemeister who likes bikes and scandal
Whooooshh, I know bugger all about electrickery but I love the fact that there's people out there that do get it.

That is a great outcome.
 

born-again-biker

Is looking for a 16" bar
@Nambra - Legend status acquired.

Stages - Eat a bag...!

Just to ad some utterly useless, but mildly interesting info.
We use strain guages at work on a lot of our fuel and water tanks. Anything from 10 to 200m3.
From memory the raw signal is converted to 4-20mamp and then displayed as /where needed.
They do randomly fail from time to time. Definitely delicate and robust at the same time.
Calibration sometimes requires pumping an exact /known quantity into the tank and then setting baseline from there.
I think they have some correction adjustment at the signal converter for fluid density, temp etc.... but it's been a while since I was near one.
20 year old ship, so no Bluetooth options


Sent from my LG-H870DS using Tapatalk
 
Top