'Just'. Hasn't this been one of the complaints against SD DACs since the 80s? DSD even more so. I'm not sure what the requirements are for EMC for audio gear any more, but there is certainly an 'allowance' of wideband noise that equipment has to handle. How that is handled is of course up to the designer.
Actually, you're quite right, but it was not an ultrasonic signl, but a a hypersubsonic in form of a female finger traing to dust the titanium twtter less that 2 minutes afte she was told to stay away from my speakers at any cost. The tweeter was brutally murdered by a female finger.
Fortunately, I anticipated just this sort of thing, and have purchased 3 spare tweeters, a wise decision especially then, since the manufacturer, Audax (Son Audax) has been shut down by Harman Intrnational since then (in 2006).
Quite so, but the paradox is still there. It appears the easiest overall solution is to use an input filter at 50 kHz or lower, if you are unwilling to bet that your CD player is perefctly clean up there.
In fact, the burnout will likely happen way below 50W and 50 kHz, a friend and I experimented with a few tweeters (admittedly, low cost, low quality), and some gave up the ghost at 30 kHz and about 25W.
I do not see a big problem there. DSD HF noise or noise shaping HF content is far below audio band signals. Pls see attached.
Attachments
Music is not constant level sinususoidal signals...
Yes, in the actual state of the technology, the active devices we are using are not fast enough, in amplifiers using feedback loops, to provide FLAT (including phase) bandwidh in the audio range.
On my side, I will be satisfied when i will be able to get an amp with high enough and constant feedback ratio up to 40KHz. We are far from this.
I don't think they were shut down, it is just that the were no longer selling to the DIY market. Loudspeaker manufacturers want to keep the best drivers to themselves! But recently I have seen new Audax drivers being marketed again to our crowd. Don't know about your tweeter though.
Never said it was. We were discussing HF grot getting in. Or have a missed a direction change again?
Not sure it's a paradox, but it is an area where design decisions are required. handling common mode crud and stuff above 1MHz is I would hope, known to all and solutions accepted. Its the 40-200KHz area where there appears to be most difference in opinion. (thinks captain Rum on blackadder)
I am reminded of J. Marshall Leach' 250W power amp. He used an interesting solutio there, he incoeporated a 12 dB/Oct, Bessel input filter into the input stage topology.If memory serves, the -3 dB point is at around 50 kHz. However, that's -15 dB at 100 kHz, which is a reasonable attenuation rate.
Andrew, please correct me if I'm wrong, your first filter in your 250W amp kicks in with -3 dB at around 300 kHz?
Andrew, please correct me if I'm wrong, your first filter in your 250W amp kicks in with -3 dB at around 300 kHz?
That tweeter was their second best tweeter, right after the Eye of the Tiger tweeter, which was a commercial faliure, not many people used it, and having hear it, I think I understand why, its price/performance ratio was not good, it simply cost too much for what it offered. In my view, my tweeter, T92, sounded better for about 1/3 of the asking price. Although I should in all fairness mention tha the edito of a British audio magazine, Noel (not Cowards, but I forgot his last name) did produce and publish a loudspeaker kit based on Audax drivers, including Eye of the Tiger tweeter which did sound damn good.
I was told by Harman Interational that their Audax operation was shut down for good. But times change, perhaps they were bought by someone else, who sam reason to reactivate them. It would certainly make me very happy. Anyway, given my stock of tweeters, and the damage rate of one per 13 years, I should be good for another 26 years. 😀 Let's hope I'm still around in 26 years, at a ripe old age of 88. And anyway, there's always the stand-by pair of AR94 speakers. Not the same thing by a long shot, but still a very nice pair of speakers.
Last edited:
Really?! If that is 10dB/div, I see a problem. That will get further amplified by a line stage and/or power amp gain. Will that gain be highly linear in that freq range so as to not cause any audio freq artifacts. This is an area where CFA could have an advantage.
The freq of an input filter should start where this noise freqs begin...... no higher than 100KHz from other measurements on gear. Fortunately, that freq is high enough so as to not cause (assumption) audio affects of its own. Bessel would be my choice.
BTW -- last night I looked up the issue on the internet and there has been a lot of discussion on DIYaudio already.
However, there are many other ways unwanted HF can enter the system.... cable pickup... direct - espec grounding - and field coupled. Notice I said input filter.... not output filter, unless that output filter has low Zo. If not, the extra impedance will raise the whole cable sensitivity and make it more susceptible to HF pickup. Then there is the smps noise that gets into everything if great care is not exercised in design and..... cable, grounds, ac power lines et al.
Sorry JC, I know this is about all analog/LP systems.... but, environmental HF influences on amplifiers are there as well. Especially low level signal circuits Such susceptibility is not usually a part of basic testing. But this 'noise' from digital (CD) is greater problem IMO. It is always there and doesnt go down with audio signal level or environmental changes.
THx-RNMarsh
Last edited:
Balanced interconnects, attention to pin1 and a decent box will stop a lot of that. If you live next to WXYZ you might need a full EMC pack!
RCRC on the input also works fine. 47 Ohm 1nF 47 Ohm 1nF is what I used in my last thingy and it is 1 mdB down @ 20kHz with just 1 degree phase shift. Not that I think phase shift matters much and certainly not this high.
An input transformer solves this and other problems, but nobody dare speak the T word.
All good fortune,
Chris
All good fortune,
Chris
It has been my experience that ferrite in audio is 'almost' universally bad. Almost always a misapplication, where a some other change should have been made instead an attempt to patch a spot with ferrite.
In the case of, lets say, a transient pulsed DC rail for a class D amplifier (a-la hypex), it obviously begins to make some sense.
- Status
- Not open for further replies.