Yes, LT1693 has ~40uV noise, comparable to 7805, not a good choice for MC pre and phono. For line level application however, John Walton's panel did not find noise performance at that level to be an issue. It is not fair to condemn the part and other line level devices using it.
There is a good chance many members of the New Jersey Audio Society would not agree.
Last edited:
I believe it was actually Gfab in Scotland, which was sold to Diodes Diodes buys TI's Scottish fab
TI also transfers processes quite often, if the product revenue justifies it.
About measurements 'datas' you have to 'believe' that measurements had be done correctly on a technical point of view, and ... honestly.
Anyway, they will say little about the way it will 'sound'. Do-you agree ?
So, on my point of view, it is 'anecdotal' in a way that measurements will just ensure the performance will be better than what you think thresholds of audibility. Result, you'll have to carefully listen to decide if it sounds as you expect. And what you expect will depend on your taste and the application.
Now; if you consider "plural of anecdotes" with the fact that they are concordant, you have first to consider correctness and honestly the same way. But you will have an idea more precise about the "way it sounds". Saving time in eliminating gears that don't goes in the direction you want to go, while concentrating with your personal listening tests.
Boths approaches are complementary and of same value, not opposite on my point of view.
About J.C. i will not follow-you on the bashing game that seems the favorite sport here.
First, I worked with his Ampex's tape recorders. And, if the Studer's of the same period were better at measurements, I always found that his electronics sounded better, and, in any case, were still irreproachable at measurements and fully satisfied the qualification of 'professional' material.
On the rest of his positions, of course, I will not follow him on a few points that I find more funny than detrimental, apart for his own image, (silver cable 'breaked' and directionalized, Bybee magic detox filters, integrated circuits is bad ;-). But, for the rest, I do not think that his "legendary" reputation is undeserved. The adorable Papa Nelson surely has a lot more charm to attract sympathy. They two share the same real passion, thousands of hours of design, tests and comparative listening. The two are looking for something that visibly escapes objectivists and manufacturers of industrial unregulated circuits. Even if they are high level engineers and have produced circuits with breathtaking performance. "Character".
That deserve respect.
Even if I'm not sure what I'm looking for goes in the same direction than him for the moment. Blondes & brunettes ;-)
Indeed. It does not look honest at all (random noise ?), but, as you said obvious "copy and paste" of the same print.
Last edited:
PMA can hear the difference using DBX. So, you must have bad equipment or your hearing ability is bad, or maybe both.
Good in engineering it does not mean good in listening ability. Those have different training. Many EE not have good listening ability and people who have good listening ability almost do not understand about engineering. Because of these reason, there is many funny myth in audio.
Or he considers Pavel's preamp design as very poor, AD797 and AD744 as completely rubbish or both. 😀
Problem with EEs with high listening discrimination ability often result in neurotic behavior.
Problem with EEs with high listening discrimination ability often result in neurotic behavior.
Last edited:
Usually in several minute of one song, only have less than 1 second sound difference in subtle sound different. It is hard to concentrate at several minute to find the different. But if you found it, you just need concentrate at the event.
You must run trials, until you find where the significant different. If you found it and understand the different, it is easy to find the different using ABX test.
Ohh, up-to-date facts from an active professional. I'm not sure this thread can handle it ;-D
//
Hehe. I'm a little pissed off too with the discontinuation of the LME49830, I have four amps based on it driving old Hitachi lateral FETs. I hope they outlast me, don't plan on building more amps, if they do give up I'd probably just use some of Bruno's class D's and be done with it 🙂
+1
I would add that it was strange to read that the audio market was not economically interesting. There is 'audio' in the four of the best-selling objects on this planet. Smartphones, cars, televisions and personal computers. And there are many manufacturers who communicate with the sound quality of their products as an argument.
Last edited:
Audio quality seems to come way down on the list of priorities for any of the things on that list T, and where music is used in documentaries, for example, as backing music, it isn't backing at all, but often loud and inappropriate to add drama and excitement in David Attenborough's nature programmes where there is already plenty of drama and excitement, I can't bare it and turn the volume down and watch with subtitles.
The reality is that the volume markets are using highly integrated solutions because they prioritize PCBA real-estate, low number of external components, and cost.
ESS may have had some design wins in China and with mediocre LG phones, but it's a fad that is already dying if not dead. They were only looking for a differentiating feature to set them apart - the customers do not actually care and it is doubtful their codecs are any better than the top parts from Qualcomm and Cirrus.
No sane phone manufacturer is going to buy discrete op-amps when they can buy something like this:
WCD9340 Aqstic Audio Codec | Qualcomm
Even if you dispute that the electronics in these devices are good enough to be audibly transparent, the real problem with sound quality in all the devices you mentioned are the speakers. Phone speakers - good for the size... but bad, Car speakers - mostly bad, TV speakers - awful.
Exactly. Although how they manage to dream up the names for these is another matter!! "Aqstic"!!
The '028' moniker IIRC came from the PMI series opamps. The '07' was one of the earliest low noise precision opamps and I recall seeing them used in audio way back in the early 80's (anyone remember the Ben Duncan preamp?), and later on the OP027/8 (trying to remember here . . . this is waay back).
We used their REF01 in some of our industrial stuff, but never the OPA's because they were expensive and settled instead on the LM308A for precision DC work like T/C conditioners. For SG amps, we later used the Intersil choppers.
You seem to be forgetting that the opamp PSRR divides most of that rail noise down, so the equivalent input referred noise is decades dB lower. That said, I often use a simple 10 to 15 Ohm resistor in series with the opamp supply rails and a 100uF cap to ground, so HF is nicely attenuated and the rails are heavily damped.
The margins for semi suppliers in some of those are not good at all. Auto almost always commands better margins because of the quality and qualifications required since a field failure is an expensive proposition and its not good for the auto brand. Auto has done more than any industry to push semi quality over the last 25 years.
If you do something new and exciting in mobile (which we did - it was active speaker protection), it takes just 4-6 quarters for your margin to be whittled away as competitors pile in. Cost reduction programs kick in and down the plug-hole you go. Get out, and do something new, and the whole process starts again. And on discretes, we regularly did 'portfolio pruning'. Customers hated it, but would not pay sustainable prices.
As for the computer market, the only guys that made money in that were Intel. Its a mugs game for suppliers. In China, OEM's are paying <$10 for a 250 Watt silver box supply and mobile computer bricks aren't far behind.
I remember trying to explain to Charlie Hansen (RIP) why things were the way they were in semi's but I may as well have been talking to someone from another planet. The investment costs are huge and the demand for good quarterly results and profitability relentless - I spent 10 yrs as a Product Line GM - 5 in auto and 5 in computing/consumer before moving to Asia to head up Marketing. Been there, done it. Got the T shirt now semi-retired and having fun . . . 😀
Last edited:
No I think when a manufacturer does a process tweak and it is also good for audio parts they will use it for that.
John,
Thanks for stepping in. On the parts finished after you moved, so far I am using a bit more than 3,000 of them and way more should be required in the next year or two. It is nice to see such good parts at unreasonably low prices!
BTY I do not expect any sane manufacturer to change parts or processes based on listening tests. I expect listening tests are more of an educational effort to correlate that experience with which measurements or other processes seem to be important.
I can only imagine the corporate reaction to "We have a problem with that item as the midrange definition is lacking!"
I did enjoy the visit from your colleagues and owe then a thank-you email and follow up on the issue of DC blocking capacitor values and circuit topography discussed a bit earlier. Only open issue was Vadim seemed not to like my use of electrolytic capacitors on the PCB but held off saying anything. Still wondering!
Vadim also reminded me of my grandfather, same accent.
No one here is interested of the relationship between listening tests and measurement.
They think, if they do the best in designing something, the imperfectness must be below human threshold. So, they do not have design philosophy. No need to trade-off of anything.
- Status
- Not open for further replies.