audio-talk

vinylnvalves wrote: ↑Sun Dec 04, 2022 4:51 pm CLD will only work if there is relative motion to shear the green glue between the stiff membranes. IHMO, mass loaded vinyl will work best with stiff cabinets.

Just to avoid any misunderstanding, it doesn’t need to be shear. In the usual AV application I think the relative motion will be compression and tension (panels moving towards and away from each other) which I think should carry over into a speaker cabinet. I have no idea if it will sound good, but it is at least a compound which has been developed or selected to absorb energy at audio frequencies at room temperatures. One tube would be enough to experiment so surely not too expensive?

In the context of these little monitors, £13 is loose change

Ah Mark... no expense spared when it’s other people’s money . The glue in tape format looks the easiest to use 50mmx 30m - so 1.5m2 for £20 Doesn’t sound that bad. A metallic foam stuck to the back wall would be up there.. absorption and damping.

This thread has been the first to prick my ears up for a long time. Liking the green glue tape idea. I know quite a few people in building trade who if handed the mastic gun version would clart up the outside also with it stuck to their hands. Although that would probably improve sound further …. i dint think you’d like that.

Trousers flapping slightly. Just picked up the Denon amp that Jon has been trying to buy for a few weeks. He got it in the psot yesterday.

Denon amp's good, especially for the money, but not up to Nick's work !

Meanwhile, Troels has nicked my idea :
http://www.troelsgravesen.dk/Purifi-Be.htm

...however I will look at his crossover. i had it in my head that 2nd/2nd wouldn't work because I'd need to reverse the polarity of the drivers, but if the acoustic roll-off adds to it, and you don't need to reverse, then maybe it's worth quickly trying. it would make it easier to tame the 4kHz lump on the Purifi.

I’ll certainly be interested in what you find out, so do please keep reporting. We are soon to ‘make over’ our living room and floor standing speakers (my WAD KLS3’s) are apparently not in her plan so I may need to find a much smaller alternative. The Troels Gravensen kit looks a good compromise……….I think

That driver did something special from the listen at Owsten.

Nick wrote: ↑Thu Dec 29, 2022 12:24 am That driver did something special from the listen at Owsten.

That’s good to know. I wish I had been there to experience it. Maybe I could sort something out with Mark for a listen sometime.

Yes, for sure, Greg. Just bear in mind that Jonny will be taking them over in a couple of months once Dave has built the final cabinets. However my estimates of timescales are usually wildly optimistic.

For whatever it's worth, based on the driver responses & doing a basic minimum-phase extraction, I suspect Troels's filter is something like a primary inductor on the low pass of about 0.82mH - 1mH, with a shunt cap of about 33uF & resistor in the 1ohm - 1.5ohm region, the LCR roughly 1.5mH, 22uF & 1ohm or thereabouts. For the high-pass, resistors of about 4.7ohm - 5.6ohms apiece, probably something around 8.2uF on the series cap & about 0.25mH for the inductive shunt.

I'm not saying these are exactly what is used (they definitely won't be -I suspect the shunt cap & resistor in the low pass in particular will have the largest variation to the actual design) but with the Y axis offset indicated & an approximate for the Z axis, the axial response doesn't appear dissimilar, though not quite identical as you'd expect given that it's just a rough sim off converted (partial) data. As far as I can tell, it's not really LR4, but it works, albeit the phase alignment isn't spectacular. That being said, as noted this is just an approximation from partial data, so it's not going to be spot on & the real speaker may well be better.

That's interesting, Scott, I'll read that more carefully with LTSpice open, later. It suggests he's pulling the woofer down earlier than i thought. i was sketching-up a revised low-pass with 0.39 or 0.47mH yesterday to get a -3dB at about 1.5kHz.
Are you sure that the resistor in the LCR is only 1 ohm ( ish ) ?
Only he says in the text that it's quite a high value resistor and only takes 1 ot 1.5dB out at 800. This is a shunt LCR , I could see the R being low in a series LCR.

It's crossing about 1.8KHz at roughly the usual -6dB for flat-summing even-order slopes.

Sorry, good catch, that's what happens when I have two different screens open at once; wrote out the values from a floorstander I'm fooling with. Should have read 2.2mH, 33uF & 22ohm. Definitely not exact, same as the rest, but should be leaning in the general direction, if you like that approach to filtering (I'm ambivalent, but that's just me).

0.39mH series, nothing else? Assuming a starting baseline of about 86dB effective, then that should put the driver at about 85dB around 2KHz, and remaining lurking around there, notwithstanding the usual variations from diffraction &c. to about 4KHz when it finally starts to roll away. The rising coil impedance with frequency cancels out the effect of the series inductor, so it essentially just provides a bit of EQ and that's about all. This is why LTSpice &c. isn't much use for crossover design. Not that I'm saying the values above are what Troels will have used (they won't be) but they give a generalised direction based on the on-baffle FR and impedance curves of the drivers.

Thanks, more good advice. I was not taking into account the driver response ( yet ) so 0.39 is unlikely to be enough. I have a 0.56mH handy so will try that.
I have a Zobel on the Purifi already, but i am finding that you can do things to the Zobel to make it look like the cap part of a 2nd-order filter, just haven't got my head around it yet.

Gotcha on the LCR bit, I'm not sure if I want to do that myself, will just try the 2nd-order low pass first.

Spinning a few plates as usual at the moment, will post more in due course.

Adding a resistor to the shunt cap in a low pass damps the low pass characteristic.