and while we are at it I can jump down a second rabbit hole: so assume we know when it left the factory it was calibrated and traceable. So any time thereafter it may no longer be so. To deal with that you could have a way to send it back to the company (or somewhere) and have it re-certified, as often as you see fit. I doubt that there will ever be a working (user) solution to this dilemma but for certain situations, at least if a standard existed then we have a place to start. Right now nada!

This was actually part of my next question: you anticipated it. Even if a left-side only PM was calibrated before it left the factory, how would you know it was still accurate six months later, after use on bumpy roads, in freezing cold and burning hot conditions, and after the bike had been dumped in a couple of crits?

Personally, I'd want a way to check that. I've said before that the raison d'etre for a power meter is that its data will answer questions or solve riddles. What you don't want is to have a riddle where your first question is, "I wonder if my PM is wrong?" To answer riddles, we need both accuracy and precision but not always at the factory -- we need it for when we ride.

Probably the only way to do that would be with key-life tests, which I doubt any PM manufacturer would do.

You mean any PM, not just left-side?

That's kind of what I was referring to in my post w.r.t. temperature, shock, and vibration tests. You can add in endurance tests to that. In my world (military-related), we'll come up with a usage profile where the product is tested under a "typical usage profile" for months on end, involving regular applications of shock, temperature, etc, etc, then testing under a range of expected usage conditions. And then on top of that, we'll have random samples come back from "the field" and the first thing we'll do is run them through the same test again.

The most nerve-wracking for me, personally, as design engineer, is the simulated direct lightning strike. (and people wonder why military electronics are crazy expensive)

But this increases the cost of the product significantly. I think, we, the cycling consumer, would rather have a $250 PM that's reasonably ballpark and correlates OK most of the time with assumed-ground-truth-SRM than a, say $500 PM that's science-grade.

The only challenge here is that these days, there's very little - if any - tie-in between power meter cost and accuracy. In fact, one could argue that on a scale, the most expensive power meters available today are actually the least accurate compared to both mid-range and even least expensive models. One only need to look at recent entrants to see that.

(I know you're not saying that, but still point it out)

Without repeating what others have said in the thread, I will note there is actually interest within the power meter industry in having standards (mainly because it benefits the more established brands). It's been oft talked about by both power meter companies and some in this very thread, including some initial draft documents. Getting that into some sort of formal standards body and/or organization, is where things have stalled a bit.

---

-
My tiny little slice of the internets: dcrainmaker.com

As I've said before, we the accuracy and precision you need depends on what you need to do. Different applications require different levels of data fidelity. It turns out that training is one of the least demanding things you can do with a power meter; drag measurement is one of the more demanding but, if you're a track sprinter you may have different interests in data fidelity than if you're a road TT'er or a XC downhill racer. But I think a '9' key is pretty important in any event.

[If you're wondering what the '9' key refers to, I described it back here: https://forum.slowtwitch.com/forum/?post=7249085#p7249085 ]

Of course - that's what I intended to communicate by saying we collectively chose the $250 PM with at least ballpark accuracy. I consider myself on the upper end of consumer-grade PM usage, and I don't care that much.

There's a small cohort who care about more than that. And those are the ones who'll have a hard time.

Don't blame 'em. I've been involved in several of those efforts at NIST and IEEE and it's hour-after-hour of telecons, data calls, draft documents. And inevitably the pace of innovation beats the standards-definition body, so it seems silly at times. It's not any fun for anyone before there's a payoff in X years. And each company participating has to volunteer to set aside X hours of labor by their engineers in what's an increasingly competitive market.

I think the Zwift-and-other competitions might require it at some point, though. Without naming people or companies, there are sketchy things going on that I'd hesitate to call full-up cheating. But are scraping the gray areas for all they're worth.

God I love this forum, ask a question and someone is a person with real knowledge in the field and can add real value to a discussion... I can name a few in this thread already! As I said early on this may be an intellectual exercise but from where I come from standards are all over the place and I am sure electrical engineers (or most engineers for that matter) are used to these standards and how they come to be! I sure hope someday we get to the point where true calibration standards will exist and we can have a new discussion! Again thanks for all of the input so far this has been fun and informative.