I don't have the resources, neither mental nor hardware, to measure these ultra low distortion.
I have tried these for driving 8ohms speakers and for driving 32ohms and 300ohms headphones.
The headphones seemed to me to be completely unaffected by the added electrolytics in the signal path.
I am not so sure about the effect with the 8ohms speaker.
I used series/parallel quad:
one pair of polar -ve to -ve in series,
one pair of polar +ve to +ve in series,
then parallel these two series strings.
The effective capacitance of the quad is roughly the average of a single capacitor.
I thought distortion in electrolytics rises very quickly and in proportion to the AC voltage across the cap. So, the trick is to really oversize the device so the -3dB is at least an order of magnitude below the fo of interest. So if its 20 Hz, size the cap for a -3 dB of 2Hz or lower. Of course, if you want to use low values of cap, then bipolar or film is the way to go without having to worry about any distortion.
Let me make it clar here, this is just my understanding - Richard is the expert on these matters and my point of reference on the subject
Let me make it clar here, this is just my understanding - Richard is the expert on these matters and my point of reference on the subject
I have read similar in many sites.
My experience tends to confirm that this really is the case.
The science behind the over-sizing seems to also fit with listening experience.
Aim for <2Hz and the 20Hz to 20kHz audio band gets through.
So do I - now. Until then, I had zero experience with twin drivers, not counting my AR94s, which also use two 8" drivers, however in a 2.5 way setup. They have he same general diameter, but the lower one is obviously different from the top one by having an easily seen much larger dust cap. The bottom one is cut off at around 150 Hz, leaving it all to the top one. And it wroks well indeed, but that's different. In the JBL Ti600, both drivers are "on line" from top to bottom.
I admit I was a bit sceptical intially, but hearing is believing, and it does work nicely, not boomy, rather pretty clear, relatively neutral and well controlled. Not too mcuh low end extension, but I'd gues around 50 Hz or so. In return, clean and clear. All in all, I think it was a very good buy, and my wife certainly likes them.
Actually, even I (since I don't do speakers, did it once and that was enough) know that a smallish driver can be made to have good low end extension, but usually at the cost of efficiency. These JBLs still manage a nominal 91 dB/1W/1m, which is quite reasonable. With heir H/K 680 integrated amp (nominally 85/130W into 8/4 Ohms) some 110+ dB SPL is quite possible, perhaps 1-2 dB above that before clipping. Quite enough for a smallish room and home listening.
Kindhorman, yes, unfortunately Harman Interntional put the lock on Son Audax in 2006. That's a bummer for me too, as replacement drivers are no longer available except with freak blind luck.
In my experience, it's the tweeters which tend to go first, so I have a stock of 3 spare ones, brand new meaning never used. Which is good, because one of the ones in the speakers seems to have given up the ghost, I have yet to investigate.
Crime of the century if you ask me, Son Audax was, along with Britain's Celestion, one of the two oldest loudspeaker driver manufacturers in Europe. I read that somewhere. I suppose economic reasons are omnipotent, but we are losing too many once very good companies. We lost Infinity, since what is now sold under that trade mark doesn't even resemble what they once were. Now it's a boom 'n fizz commodity based on percieved value, damn big cabinets with damn big bass drivers, looks like a good deal, but sounds like a bummer.
In my experience, it's the tweeters which tend to go first, so I have a stock of 3 spare ones, brand new meaning never used. Which is good, because one of the ones in the speakers seems to have given up the ghost, I have yet to investigate.
Crime of the century if you ask me, Son Audax was, along with Britain's Celestion, one of the two oldest loudspeaker driver manufacturers in Europe. I read that somewhere. I suppose economic reasons are omnipotent, but we are losing too many once very good companies. We lost Infinity, since what is now sold under that trade mark doesn't even resemble what they once were. Now it's a boom 'n fizz commodity based on percieved value, damn big cabinets with damn big bass drivers, looks like a good deal, but sounds like a bummer.
I thought distortion in electrolytics rises very quickly and in proportion to the AC voltage across the cap. So, the trick is to really oversize the device so the -3dB is at least an order of magnitude below the fo of interest. So if its 20 Hz, size the cap for a -3 dB of 2Hz or lower. Of course, if you want to use low values of cap, then bipolar or film is the way to go without having to worry about any distortion.
Let me make it clar here, this is just my understanding - Richard is the expert on these matters and my point of reference on the subject
Andrew, as I understood Bob's comment, the way to go is to use two bipolar types back to back, probably with a bypass film cap. That's supposed to reduce distorion, and specifically, Nichicon NP caps were mentioned.
As for oversizing, I agree completely. NEVER be stingy on caps.
Pono measurements available online now Pono PonoPlayer portable music player Measurements | Stereophile.com Not the most linear of devices, but of course open loop outputs sound so much better. For us objectivists there are better devices out there. Compare the -90dB plot with 16bit data against the xonar STX sound card. I note he doesn't even show that with 24 bit data. I wonder if it was equally bad?
Shall I say that the test graphs shown are very carefully "selected"?
How are you detecting it ? Sounds more like radiated hash from a switching PSU or something along those lines.
Analog To Digital Converter Files Identified, Converters Profiled | Analog Planet Now I admire Fremer for trying, but am not sure if the results can be taken as statistically accurate. If they can then those of us who want to rip vinyl can breathe a sigh of relief. Some of the greybeards will also feel smug that they have been telling us for a while that all ADs are effectively transparent.
All I have to do is wave my scope probe in the air. Every time I try to measure something I find this 30KHz wave and am wondering what went terribly wrong.
That definitely sounds like a radiated field from a switching PSU or a CRT monitor flyback transformer. It can be picked up at quite a distance... several feet from the source.
Is it also there if you clip the scope ground to the tip, and make the loop as small as possible?
Jan
That most probably means that your 'regular' measurements ar not affected.
Thee's a ton of junk floating around that you can pick up with a probe as antenna; you'll never get that down to zero.
In a circuit the impedances are generally too low to pick it up.
Jan
- Status
- Not open for further replies.