Yamaha YH-2 modding: a noob’s progress (part 2 of 3) Measuring modded vs (almost) empty pairs The graph below shows the effects of removing much of the damping from the two modded pairs. Taking everything but the ~14mm patch of foam in the central metal insert from pair1 results in removal of the 7k peak, introduction of a trough-peak pair at 2-3k, and a high peak at 160 Hz with a sharp dropoff below (green vs black curves). Taking everything but the damplifier & butyl from pair2 – so the cup-edge vents remain sealed – results in a drop through the treble and around 1k and a peak at 130 Hz, again with a sharp drop below (blue vs grey). The major difference between the two pairs after disemboweling is that the cup-edge vents on pair1 are now completely open, whereas on pair2 they remain completely sealed. That pair2’s 2k peak remains (though narrowed somewhat because of a new trough to the left of it) suggests that the lack of cup-edge venting might be responsible. This idea matches @Ishcabible ‘s results from drilling holes in the backs of the cups of a pair of otherwise lightly-modded HP-1s. The introduction of a trough in the same place to the emptied pair1 further suggests that at least some degree of baffling of these vents is desirable. It’s tempting to infer from these observations that ~2k is the resonant frequency of the rear chamber, even though the cup volumes are somewhat different (there’s quite a bit of butyl/damplifier in pair2, and HP-1 cups are bigger). The mid-bass peak introduced to both pairs is reminiscent of what happens when leaving rear central cup vents more open vs baffling or closing them in other headphone mods (e.g. T50RP, HD250), so is likely due to the pivot-hole-connected vents in the central insert being uncovered by damping removal. The poorly-damped stock pair2 configuration had a similar peak in the @Philimon measurements. The difference in peak position between the two pairs measured here could be driver-related, maybe different stiffness of or tension on the supposedly tensionless diaphragms; or related to other differences between the two (pair2 has the conical magnet holes, pair1’s are cylindrical). Differences in the treble could relate to the pads. Pair1’s pads are very soft and compress markedly more than pair2’s on my coupler (and my head). Pair1 also retains its small central foam dot, which is just big enough to cover the 4 innermost magnet holes – but in theory, this should give more treble, not less. It’s interesting to compare the impulse responses and decays of the modded vs undamped versions if only to reinforce that yes, damping of these drivers is really necessary. Here are impulses and CSDs corresponding with the curves in the FR graph above (left channels only for brevity). Pair1 first, then Pair2: Compared to their modded counterparts, the impulse responses of the undamped pairs look pretty ragged and have low-amplitude oscillations that persist out to beyond 1 ms. The modded pair1 impulse is interesting, the second peak being almost as high as the first. The CSDs show obvious differences, the undamped versions having much longer decays below about 3k. For pair2 the difference in the 800-2000 Hz area is more than is apparent because the dip introduced to the FR of the undamped version there results in a lower ‘start line’ (SPL at t=0) for the decay surface. Changing the graphs’ vertical ranges shows that the SPL drops in the highs of the damped vs undamped versions mask slightly longer decays there too, but the contrasts are nowhere near as much as in the mids. In both versions of both pairs, there are no obvious ‘ortho walls’ anywhere. I listened only briefly to pair1 without damping, and FR changes aside, it had lost the speed and clarity which were appealing in the modded configuration. Here are more thorough summaries of the sorts of changes brought by damping, from wualta in the HF ortho thread (scroll down to his last couple of headings) and @Philimon later in the same. Effects of different damping elements Having made some inferences on how different parts of the cup influence the frequency response, it’d be interesting to try to establish which damping components are responsible for which changes. A series of experiments on pair1 allows for some speculations. On the graph below (lots of curves, enlarge it): Black is the dBel84 config (microfiber, black foam, arctic fleece; photo in the previous post) Blue is with the microfiber removed Green is with microfiber removed and black foam switched for a 40 mm diameter, 8 mm thick cosmetic foam pad with ¼” holes over 2 of the centre-insert vents Purple is as green but with a hole over 1 centre vent Tan is as purple but with a solid cosmetic foam (no holes) Red is as green but with the microfiber returned The small patch of black foam in the centre-insert bowl and the 'Arctic fleece' cup fill remain in place for all. The red measurement (first in the series) was made with SPL = 90 dB @ 1 kHz and the volume knob was not changed for the rest, so the absolute changes of each configuration are illustrated. What I infer from these results is: Blue vs black shows the microfiber mainly suppresses the lows & lower mids, the effect diminishing to zero around 3k. Blue vs green and vs the undamped (‘empty’) curve in the first graph in this post suggests it’s the black foam that’s largely responsible for the upper mid dip and 7k peak. Red vs the others suggests the microfiber contributes here too. The differences among green, purple and tan (same foam with different #s of holes over the central vents) suggest that variable damping of this part of the driver (the second set of driver holes counting out from the centre) and/or variable venting to the pivot exit influences the lower treble (less with less damping) as well as the bass & lower mids (more with more venting). I was surprised by the magnitude of the effect of the thin microfiber. I didn't at first believe it would reduce the lows so much (red vs green) so I tested the same config on the other side, being careful to ensure I'd got the holes in the foam pad lined up properly with the cup vents. The result was the same. While these results show the effects of different materials and seem consistent with variable central venting influencing bass, it’s not as clear how they bear on @dBel84 ‘s ‘differential damping’ strategy. Damping more holes in the inner – but not innermost – part of the driver does seem to boost the lower treble, but the results aren’t consistent: green (2 holes in the foam) is lowest, but purple & tan (1 & no holes, respectively) aren’t significantly different. More about that in the next post.