Measuring bare headphone drivers?

Been buying various sets of Aliexpress headphone drivers to experiment with building my own headphones as the 50mm Peerles ones that a lot of DIYers use are out of stock at the moment. Figured if I have a pile of drivers, I might as well get some measurements to give other DIYers an idea of how they perform.

Is there a standard to testing just the bare driver as I imagine puting it in a set of headphones and testing it with something like the MIniDSP Ears would introduce a change in sound signature from the headphones and pads? In the same way I worry about just pointing a UMIK-1 at them as they'd be a drop in bass respose.

 

That’s true. Speaker drivers tend to get tested on an infinite baffle. Headphone drivers don’t quite have a standard way of comparison independent of the headphone cups.

I would think the best way would be as open as possible with a consistent distance from the mic or EARS rig. Maybe one of those stand that can hold a single lens for an optical table

 

Should be easy enough to 3D print a jig to hold the driver a set distance from a mic. Would probably do 2 sets of measurements. One of the bare driver and one in the Aurora Borealis knockoffs I designed and 3D printed as a "real world" audio test of how they fare in a set of headphones. Wouldn't be them performing at their best as I'd want to tweak the headphone design for each driver but would make a good enough baseline and allow me to do some subjective performance reviewing too.

 

IME / IMO -

Using a cylindrical driver housing just wide enough for the drivers to fit flush, with a lip to slip open cell foam pads over is the best way to A/B electrodynamic drivers while removing the most confounding variables. Basically like a Grado without the aesthetic features, or a Mini (one of my own designs). @spoony has used something like this too. If you have a 3d printer you could easily design and make something like this. There are probably free thingiverse files too!

edit - word of warning, after getting used to the sound of a close-up driver in an enclosure with very few acoustic features, many other enclosure designs may begin to feel somewhat comparatively comissive (meaning adding some qualities that aren't there when trying to isolate the driver performance in and of itself). It'll be a dance of steps forward and backward after that point. I have grown to prefer the unadulterated sound of a KISS design like this, even though it could be considered 'cheating' vs a more robust enclosure design. YMMV!

edit 2 - you wouldn't want to A/B drivers open-air because they will all shit the bed in terms of bandwidth. A light but even seal like open cell foam is the simplest solution I've been able to imagine. Unless of course you had a beefy driver capable of good performance in an open baffle. This applies to very few )(if any) of the OEM options out there.

 

A big thumbs up to the idea of doing measurements of drivers, and another to the idea of sharing them.