The criteria for loudspeaker choices.

@rpeluso My specific bad experience was with a PSA PerfectWave Transport ($4000 retail). I had three drive failures in 6 years. I played the unit around 4 to 6 hours per week.

I learned that PSA used the cheapest grade of computer drive (around $10) they could buy. These drives are not designed to spin a disc hour after hour. They are made to spin up a data disc, read the data, and then spin down. The choice of this drive for a $4000 transport is unforgivable in my opinion. They knew better and they simply cheaped out. By the time the third drive failed PSA no longer supported the unit and it became a boat anchor.

Second example is the PSA Power Plant 12. The review by ASR showed that the unit actually makes things worse for the downstream component but the biggest problem was that the "High Current" outlet actually reduces the wattage to the amplifier! But even more troubling was Paul's response, accusing Amir of using the wrong outlet. Paul even demanded an apology. Well, his employees (I won't call them engineers) corrected him and he had to admit that ASR did the test correctly. Did he offer Amir an apology? You guess the answer.

Third example is the PSA DirectStream DAC that uses an FPGA processor instead of a typical DAC chipset. This unit is quite expensive and has tested very poorly by Stereophile and ASR. It is actually one of the noisiest DACs available with very poor bit resolution. I won't go into the reasons for this but it is a perfect example of how a company tells a hyped story and charges a lot of money for a poorly engineered product.

For me, the above is enough to decide that I will avoid the PSA brand with prejudice. For the record, I don't trust Audioquest and several other brands that have put out products that have been demonstrated to actually make the signal worse than much cheaper alternatives.

Lastly, I don't want to start a war here. If owners of PSA products love them and hear wonderous sound quality, more power to them. My only advice is to be careful about the engineering design and build quality.

Thanks for the answer!

@8t-note

Have you actually listened to the DirectStream DAC? There are a lot of people who have including myself. I’ve compared it to ASR’s excellent measuring DACs. I was very obvious to me that there is a lot more to sound than numbers. I don’t have experience with other PSAudio products so I can’t comment.

Well, as the speaker designer at PS Audio, I thought I’d chime in.

I’ve never met Bruce Thigpen at any trade shows but, by all accounts he’s a genuis inventor and did a number of useful planar magnetic patents over the years, some of which are licensed by planar magnetic headphone companies among others.

However, the Eminent Tech speakers make me scratch my head. The crossover used is a total mess with broad overlaps and interference between the drivers (almost like there is no filter on the woofer and/or midrange). Cone breakup from the woofer is left unfiltered out and there is a bunch of interference from the resonances.

Also, sensitivty is almost 5 dB lower than specified (about 77-78 dB) and so it’s among the lowest on the market (even lower than magnapan.

stereophile measurements of LFT-VII showing driver responses with crossover

LFT-16 driver measurements

While the smaller LFT-16 looks intially better, the mid and woofer done integrate currently, even when switching driver polarity.

Also, when you have such a high crossover point to the tweeter, the midrange panel is beaming a lot before handing over to the tweeter and the off axis behavior is poor.

As far as crossover parts go, you guys selling crossover part upgrades as a side business have a biased viewpoint as to what is appropriate and perceptually significant. That being said, we use nice components in the FR5, appropriate for it’s price point, I think.

@cbrunhaver: You raise some justified issues, and make some excellent points.

The shallow crossover filters employed in the ET LFT-8 (1st-order) are also used by Magnepan in their planar-magnetic loudspeakers (PS Audio’s Paul McGowan’s reference speaker for quite a few years was the MG3.6, which he was very happy with in spite of it’s low-order filters). When Danny Richie of GR Research measured the MG3.7, he found a lot of frequency response irregularities caused by phase cancellations between drivers, a result of the large expanses of driver overlap (inevitable with low-order filters).

The ET LFT-8 (both b and c iterations) employs a 1st-order high-pass crossover frequency of 180Hz for the LFT-8 planar-magnetic midrange driver, so phase cancellation between woofer and midrange driver is not a problem (due to the very long wavelength of 180Hz). What IS a problem is the out-of-band woofer resonance (seen at around 1800Hz in the Stereophile frequency response graph), but that may be eliminated in the LFT-8b by installing a cap onto the woofer, shunting the high end response of the woofer to ground (as detailed in a few posts in the Planar Speaker Asylum forums). That problem is eliminated in the LFT-8c, which has a different (dipole) woofer system. I eliminated the problem by using the Rythmik/GR Research OB/Dipole woofer system in place of the LFT-8b’s stock woofer (easy to do, as the 8b has separate woofer and planar-magnetic driver binding posts), which transforms the LFT-8b into a 100% OB/dipole design.

The criticism of the 1st-order high-pass and low-pass filters located at 10kHz IS quite valid. A 10kHz wavelength is only approximately 1.36" long, so small movements of the listener’s head can result in frequency response changes resulting from phase cancellations between the midrange and tweeter drivers. I’d love the hear the LFT-8 with a lower midrange-to-tweeter x/o frequency, and/or steeper filter values (perhaps 4th-order, as I believe Siegfried Linkwitz employs in his OB/Dipole designs). The phase relationship between tweeter and midrange drivers in the LFT-8 may be adjusted by small changes in loudspeaker toe-in.

I don’t know where the 77-78dB sensitivity figure came from, but it is not accurate (83-84 is more like it). The LFT-8 infact requires significantly less power that do Maggies (I have both). Steve Guttenberg found 50 watts to work pretty well with both the 8b and 8c. And if you bi-amp the speaker, the planar-magnetic driver itself presents an 11 ohm load to the amp, great for tubes

Regardless of all the above, the fact remains that the ET LFT-8 reproduces 180Hz up to 10kHz WITH A SINGLE PUSH-PULL PLANAR-MAGNETIC DRIVER (without a crossover in that frequency band!), resulting in high coherency and transparency, low coloration sound. I better wrap it up here, before I’m once again accused of shilling for Eminent Technology and Bruce Thigpen.