Showing 1 of 1264 conversations about:
View Full Discussion
Jude getting some measurements to us...will try to post here as soon as I can!
Jude was able to sneak in a measurement while working on the SoCal CanJam video:
AAA 789 THD at -132.3 dB = 0.0000243%
Loopback THD = -148.4 dB = 0.0000038%
The amp is barely above the noise floor of the analyzer itself.
A single measurement, the result of which was already addressed in the specs, that was not worth the wait.
No notch filter, but as both the noise floor and distortion peaks are above those of the analyzer, there is information to dissect.
5.5 Vrms is 100mw into 300 ohms
The specs say -140dB THD, the measurement says -132dB.
The good news is that it is low order with a rapid decay (would be interesting to see if that is still the case with lower loads) :
2nd = -132dB
3rd = -142dB
4rd = -150dB
5th = -157dB
and the rest <-160dB
Main hum seems to be lower than the one of the AP, <-140dB
Sorry if I am a bit salty here, but I hopped to see real measurements (either from THX or from third parties) before the end of the drop...
I am still in though! ;)
It does look impressive, but then, these measurements are the best case scenario. I would like to see how this amp performs when pushing a watt, or better yet, several watts, into a low-impedance load. Also, give us frequency sweeps: showing only the values from 1kHz makes me suspect that the amp does not perform nearly as well if you measure the peak THD+N for a 20Hz-20kHz sweep. I know marketing says you need the biggest numbers, but if this amp could manage, say, <-100dB THD+N 20Hz-20kHz or even <-90dB at 4W into 16 ohms, I would have bought one. In a perfect world amp manufacturers would provide a graph of Vrms into 16 Ohms vs peak THD+N 20Hz-20kHz.
Are the THD measurements 1kHz or 20Hz-20kHz?
FR 16R BAL.PNG and FR 32R BAL.PNG look basically the same and have the same timestamp.
FFT 1mW 16R.png and FFT 1mW 32R.png seem swapped to me (32R should be lower than 16R, at least according to all the other measurements)
Re item 1 (FR): Good catch - I've replaced the FR 16 BAL.PNG with the correct graph.
Re item 2 (FFTs seeming swapped): I re-ran them and they're correct as-is. The change in noise floor at 32 Ohms appears to be an artifact of the AP changing range (and injecting more AP gen/ana noise) with the higher drive level required for 1 mW into the higher impedance.
Fascinating! Thank you for double-checking (and measuring in the first place). Your engagement here is exemplary and way above any other drop I have considered joining.
Was this a ”production unit” or some kind of pre-series?
Thank you so much! Your transparency is what everyone in the audiophile gear industry should strive for, although I am sure that there are a few companies that have strong incentives not to be so transparent. It is wonderful to see a product that seems amazing at first glance, and then remains amazing under close scrutiny. Being able to stand next to, and even best in some cases, the $1300 Neurochrome HP-1 with this $350 amp is an incredible feat and the team behind this should be very proud. Is there any chance that a version of the 888 is in the works? ;-)
This is a pre-production unit from the recent pilot build. It uses same design and PCB as production. There's no reason to expect any changes to the performance.
You need to test this pilot build on realworld European 220-240V power, right!? My hand is raised! ;)
There are actually problems sometimes when US HIFI reaches Europe the first time. I heard a couple of stories.
Will this one have the European CE marking?
Just out of curiosity, what do you think would likeley change/improve in these measurements if the dc-blocking caps on the input were shorted? Lower distortion? Lower noise? Something else?
Does this product have a burn-in time before giving 100% performance?
Caps forming etc, is it mumbo-jumbo?
Hi @rajaprukIt wouldn't change the noise. Perhaps the 20-40 Hz distortion might get even lower.
No burn in time is required for the 789. I have a more detailed post on this in this forum.
Question for Andrew Mason,
Now that production of the first batch has completed and they're shipping to us, have you had a chance to examine the final product and make sure that that the finished product meets THX's design specifications?
Hi euphxenos. We've set up the factory with an automated Audio Precision routine designed by me to run every unit through the gamut of tests required for proper functionality and THX quality per our design specs. So you could say we examine every final product.
I'm glad to hear that you're doing QA on these at the factory. If you trust the factory, that's a good place for QA to happen. I've already seen a couple of reports on head-fi from people who received units that were DOA. Given that and past experience with Massdrop QA, I'm not sure how much *I* can trust the factory, so I'd still be interested in hearing the results of your testing.
burn-in? burn-in? research...