Some driver tests

[Scottmoose]

Since I sometimes get drivers through here for testing, for the sake of interest, I thought I'd add a handful here when I get chance.

First up: my friend Dave has very kindly sent me a package of interesting units from his stock that he (rightly) thought I'd be interested in playing with. The 2 pairs of brand-new Silver Flute W14RC25-4 in the box in particular grabbed my attention as it's a unit I've wanted to get my evil hands on for a while. Silver Flute isn't particularly know outside the North Americas as they're basically Madisound's 'house brand', with a range of 3 drivers offered in 4ohm & 8ohm nominal impedances. These are the 5 1/2in model with the 4ohm coil -which is more or less the one that I've had a particular eye on for a while. Paper cone, with wool fibres incorporated into the pulp for some additional internal damping, rubber roll surround and a cast alloy basket (!) for $28.30US. The basket is rather nice -no venting behind the spider, but the coil is vented through the back-plate & it's remarkable you can even get a cast basket for this price. The drivers are stock, with no pair-matching: the only minor modification is that Dave has applied a single layer of thinned Mod-Podge to the cone to seal it. Additional mass from this is roughly 0.1g or slightly less.

I don't have my test baffle set up at the moment (lack of space) so I can't do any IB frequency response or HD measurements, but I've put them through my DATS setup & you can tell a fair amount about the drivers just from that (next post). Not everything or anything like: you can't design a multiway with it, but it gives some guidance.

[Scottmoose]

I put all four units through my DATS setup. Drivers are from Dave's stock & he's confirmed they're 'as received' (other than his sealing the cones) without any kind of pair-matching or other running. I've broken them in with varying LF sines to rated 1-way Xmax, and prior to testing they were run for 30mins on mild loading, then left to cool for 10mins before measuring. Drivers were measured with DATS on its highest 11.4dBu setting; I subsequently ran them on a linearity test with varying load sweeps and they remained extremely consistent, confirming QC. Vas was derived via the known mass method rather than delta mass or delta compliance -a method Jeff Bagby was promoting before his untimely passing, since manufacturers are typically very tight & consistent on Mmd (actual mass of the moving components minus air-load), and as a result, for most home testing it ends up being as good or better than the usual approach.

The key fundamental, large & small signal values are shown below. As you can see, they're remarkably consistent for 'off-the-shelf' units without any particular matching, let alone drivers retailing for < $30. Fs is within a couple of Hz, Qt is just about within a 0.02 spread, & Vas within 1 litre. For LF load design purposes, these are functionally matched & 'zero meaningful difference'. Very impressive.

Going more deeply, we've got relatively low mechanical damping, as some prefer, confirmed by decent compliance & a mean conversion efficiency of 0.373% isn't bad for a driver of this size & price. As a 4ohm unit, sensitivity is good: over 91dB 1m/2.83v. I haven't shown the Wright parameters as we've got the actual impedance to eyeball, and that's quite instructive.

Inductance is relatively low, though the steadily rising curve indicates a fairly simple, if well-designed motor, without any additional copper Faraday shielding. You can't reasonably expect much more for the price. Looking in detail at the curve, you can see a minor deviation at about 525Hz, indicating a small amount of energy storage; this is followed by a more significant resonance at about 1.7KHz, and another at what I know from the published FR graph & other 3rd party measures is the main mode centred on about 6.8KHz. This is broadband & a bit of a swine to deal with as it will amplify HD at sub-multiples of that frequency, and fairly broadly. It may not be a large driver, but it's not one that will be best used much > 2KHz, and don't even think about 1st order if you do (unless you like listening to the results of that HF hash rippling into the passband ). With a decent tweeter though & a little filtering, a really driver for the money with lots of performance potential. It should do a reasonable job on the LF end too -no sub, but as good as most, Scan Revelators notwithstanding. Very impressed -I've an MTM already planned for these, with my favourite 'affordable' tweeter.

More to come when time permits.

[Scottmoose]

Next up, this little beast:

DT-25N.jpg (129.62 KiB) Viewed 4824 times

This is the Monacor DT-25N, a very compact, relatively inexpensive 1in horn-loaded dome tweeter with a neodymium slug magnet. Claimed resonant frequency is 1.6KHz, I measured two random samples as 1.4KHz - 1.45KHz. Not bad. The connection tabs are so miniscule it's ridiculous, but otherwise, it's a neatly made little thing with a nice finish to the dome. As noted above, at the moment I don't have time or space to get my test baffle up so I can't do any FR or HD measurements, but consistency between the two is reasonable, with sensitivity coming in at about 94dB 1m/2.83v for a nominal 6ohm coil. Eyeballing the impedance curve, there's a fractional swelling around 2.3KHz, but you're really having to look for it with a very inflated scaling as shown here; likewise another at about 7.9KHz. Nothing significant. The graph scale here makes the rising impedance with frequency look greater than it is: it's about a 1.5ohm increase from ~4KHz - 20KHz, so it's actually quite clean. Given the size & lack of a large decompression chamber I wouldn't want to cross this < about 3KHz, although other measures indicate the response itself is basically flat to Fs, as distortion will be increasing quickly. Still, a compact, very efficient, neatly made & relatively affordable little tweeter with (on current evidence) low linear distortion. Depending on baffle effects, and where you're crossing you might even (just) get away with a 1st order high pass, although you'd need a damping [shunt] resistor at the very least to prevent it spiking back up at Fs.

[Scottmoose]

And another. One I've known well for a long time: Fostex FF85wk.

ff85wk.jpg (22.18 KiB) Viewed 4822 times

A nice little relatively high efficiency (for size / type) 3in wideband / fullrange with a modern (non-rotting) foam roll surround and a multi-layer paper cone, using a long-fibre base with charcoal-impregnated short-fibre kenaf (cellulose) upper layer, with a ridged central alloy dustcap that's direct-bonded to the end of the coil. In effect, like all wideband drivers with direct-bonded central domes these are mechanical coax drivers as the cone is partially decoupled from the dome as frequency rises; the crossover simply happens to be done mechanically rather than electrically. This is one of a matched pair I got from Dave, which also have the top surface sealed with a thin watered glue, final generation of EnABL blocks applied and some basket damping applied also. None of these have any significant effect on overall mass, general electrical behaviour etc. but they can help control the TL modes as the cone moves out of oscillation & into controlled flex before it decouples.

T/S parameters are impressively close to the factory spec. (calibrated DATS, conditions as described above for the Silver Flute & Monacor)
Fs = 115.4Hz
Sd = 28cm^2
Mmd = 1.915g
Mms = 2g
Re = 7.06ohm
Le = 0.169mH @ 10KHz
Cms = 0.91 mm/N
Vas = 1.04 litres using known mass derivation
Qes = 0.6863
Qms = 4.866
Qts 0.6015

The impedance curve is revealing: nice clean high Q spike at Fs confirming the high Qm & flat out to beyond 1.6KHz; you can see around there (strictly about 1.83KHz) the bump in the curve characteristic of the edge-termination & the upper piston corner of the cone; another mode appears at about 4.6KHz, where it's starting to decouple from the coil / central cap, which has a -3dB point source frequency of roughly 6.9KHz by its 16mm diameter (you can't see that on an impedance plot, but it can be calculated & is shown in the factory off-axis plots; given the trending between this & the published is good from that alone I'd expect reasonable correlation (other 3rd party measures tend to suggest this also). Pretty good, and a good little mid-tweet.

[Scottmoose]

...at times like this, I could weep.

Peerless ['by Tymphany', give me strength] HDS 830869.

What a nice, quite affordable 8in driver it was. 'Was', because of course Tymphany have more or less pulled out of the DIY market, so it's a case of 'get them if you can'. Yes, there is an alternative that leans in its direction, in the shape of Scan Speak's snappily-named Discovery 22w/8534G00, and that's an excellent driver, but it doesn't alter the fact that this Peerless was a very decent unit for the money.

830869.jpg (33.79 KiB) Viewed 4799 times

830869 02.jpg (58.17 KiB) Viewed 4799 times

A nomex (aramid) cone with polymer central inverse cap, vented around the throat to reduce compression. Trunkated cast-alloy basket that can be surface-mounted, with lots of venting below the spider. Nice flat spider too. Well-designed terminals. Decent leadouts. Nice build quality. None of which would mean much if the performance wasn't up to scratch. On available evidence: it is. Measured the pair I have in my store waiting for the evil plan I have for them to come to pass (you can but hope). Within a couple of Hz of each other on Fs, & the rest line up similarly tightly. Same conditions as previously specified, DATS calibrated, 11.4dBu maximum drive level. Major parameters:

Sd = 227cm^2
Mmd = 22.02g
Mms = 24.43g
Re = 5.8ohm
Le = 0.6964mH @ 1KHz
Fs = 32.3Hz
Cms = 0.996mm/N
B*L = 8.61T/m
Qes = 0.386
Qms = 5.47
Qts = 0.36 [nearly]
SPL = 91.28dB 1m/2.83v

Good balance, easy to get some LF from them in a mid-size box.

And from the impedance curve? As above, sharp, high Q peak at Fs confirming the high Qm figure. There's a minor amount of energy storage indicated in the small deviation from the curve trend centred on about 700Hz; you can expect HD2 to spike a little there, but this is fairly typical of 'soft' cone types when its edge resonates & goes into antiphase, and the suspension struggles to completely damp it out. It's pretty innocuous here & the rest of the range is well-behaved; the driver has some Faraday shielding but not full symmetric drive, and the coil former is aluminium so impedance is rising with frequency modestly; not bad for a 32.4mm [inner] diameter coil though. Factory & most 3rd party FR measures are quite consistent & nicely linear up to the final, reasonably narrow breakup mode before rolling off, so I'm fairly confident these will behave as-expected (basically like that!) when I'm able to get the mic in front of them, and addressing that & improving probable HD into the bargain should be straightforward enough.

[IslandPink]

Nice work !

[Scottmoose]

More to come as time permits.

[steve s]

Interesting to read scott.. thanks

[Scottmoose]

Pleasure. Hopefully will be hauling a few more out of my store over the next few days.

I have to admit, even running just T/S and impedance (former largely derived from latter) I'm very impressed with the little Silver Flutes and the Peerless 8in in particular. You can certainly beat them, but they are / were very good value & the consistency is top-notch. Which is more than I can say for a few others.

[Ant]

Id be interested to see what the big fanes were actually like now the 15 has been discontinued