Quote:
Originally Posted by DDF /img/forum/go_quote.gif
Must be talking about a different part? Pulling down the 8620 spec sheet from the AD site:
The spec sheet rates it for 30ma with 2V o/p, which is a min 67 ohm load. Unfortunately no distortion spec at this load, but it looks more than fine for HD (it'll push 60 mA into a short).
Damping factor is more than fine. At gains as high as ten its under an ohm to 100 kHz, and at the gain of the V5s (2) its under an ohm out to 500 kHz.
The large signal response is unbelieveably clean from the spec sheet, and the small signal overshoot blows away the 6121, so its recovery also looks to be exemplary.
All in all, it specs out great for headphones.
But these are just specs. I notice the amp gets warm, which is strange. Mine may be oscillating (just a guess), which would explain some of the reservations I hear, and explain the heat.
|
Hmm, I remembered a different practical result from the 8620 when I tested that, but maybe I am thinking of a different part. I don't think the impedance comes out so low when trying to drive something like headphones (especially since the load is variable, which brings into account the overall stability of the system), but maybe I'm wrong on that. I wouldn't use the 6171/2 as the benchmark, though, as it suffers in several areas in terms of performance. I can't find anything in the SSO specs that gives sufficient data to extrapolate recovery on headphones; my experience with the chip gave the impression that it wasn't that stellar at low impedance, but it could be something else was responsible for this.
Maybe I was thinking of a different part for some of the criticisms, although I didn't think so...odd. I've tried to use the 8610 several times in audio, though, and have never gotten particularly exemplary results, either in listening tests or running it on some test equipment. Subjectively, I think it sounds inferior to much of AD's suitable offerings, but I guess it isn't the worst I've ever heard, although it was one of the worst I'd tested in the PINT. In terms of measured performance, I didn't manage to get out of it what I expected by the hype of its datasheet, although I think some of the criticisms of the 627 that it points out are fair, and it did perform better than something like the 6172. After using about 50 or 60 op-amps now in various duties, I still don't think the 8610/20 is really a good choice, especially for this kind of application. I didn't get around to testing it but maybe its problem is something like TIMD or something related rather than simple HD problems...or something else that isn't occurring to me at the moment. Unfortunately, AD doesn't provide a plot from DC of the open loop response, though the very high gain at DC and relatively low slew rate suggests that it's probably not very wide. That isn't always fatal but it does seem to be detrimental to performance with some designs. The 8610 seems to perform considerably better with a buffer at its output, which is one thing that made me look to recovery amongst other things, although still not what I'd consider a high performance choice for audio.