the Impulse H2 measurement webpage

May 08 - This page was generously contributed by Rob Holt and Simon Ashton

'I spent Saturday [29 March 08] with Simon doing some detailed measurements of all things H2 with some very interesting results. In no particular order:

Distortion:

A big surprise this. Down to about 100hz distortion averages 0.2-0.3% THD @ 90db. That focal 120TD5 is very clean - <0.1% at the top end.



Next some driver responses: Tweeter (no filter) Horn (filtered) Bass (filtered).
Note the tweeter response is superb. Also note that the bass driver runs very high. More on that later:







Next, the H2 as it was originally intended with Kevlar TD120K tweeter.
Points of note are the prominent response of the horn and depressed tweeter level except for a peak at 15khz:



Next, the H2 with a 120TD5 TiOxid substituted and no other changes.
Better tweeter level and less peak:



Same but with the 120TD5 tweeter phase corrector removed.
Definitely flatter at the top except for a rise at 18-20khz (and no off axis deterioration that we could measure):



A response at the listening position: One speaker, then both (with my room bass problem in all it's glory):





Now the interesting stuff
Ignoring the low end where the room is screwing things a bit the main features are peak at the top of the horn response followed by a gentle broad suck-out. That's audible as a slightly hollow 'cupped hands' quality if you compare the H2 to some decent direct radiators as I've done recently. Very good for horn design but quite noticeable.
Going back to the output from the bass driver I mentioned this runs quite high and maintains good output right up to 5khz. Crossover on the bass driver is first order while it's second on the tweeter which sits immediately above it. Do they cancel? - you bet - so our first mod was to reverse the tweeter polarity. In addition we removed the parallel tweeter resistor to gain another 1db or so at the low end of the tweeter.
We got this:



The mid excess is modestly reduced and the tweeter level is both raised and flatter overall. Net result seems to be better tonality and reduced colouration.

Have a go! "

( I concur, having tried this - it works! Many thanks Rob & Simon - mc

JULY 2011: The only bit I'd demur from is removing the phase plates from TC120 tweeters. Don't; do it: I don't care what the measurements above say, every time I try this and live with it a while, there's something which goes missing in the mid-high treble energy range above 'presence' ; c. 8-10Khz at a guess. A certain energy / vitality that comes back once the parts are re-fitted. And yes, I have also tried this at least single-blind, on more than 4 occasions over 5 years, across various source and amp changes - MC)


www.acoustica.org.uk (opens new window)

  • 17.07.11: MC comments updated after yet another blind trial
  • 03.0.08: Page added.