well, on carlosfm's suggestion, i decided to try an OPA2132 in place of an AD8620 in my lightly modified Sony C222ES SACD player... while the 8620 sounded good in many respects, i felt the player sounded a little thin and "hifi" at times. most likely a better clock would be needed to really improve sound, but i wanted to do the best i could with the opamp first.
so far i am not so impressed with the results. the sound seems very recessed and distant now, with a stronger "opamp signature" - by that i mean it lacks air and harmonic detail, as if resolution drops with increasing frequency. some have attributed this quality to negative feedback; i don't know if that's an accurate characterization but it's what i've tended to hear from many opam circuits in the past. i also find the bass response to be less powerful, and it's not really any warmer sounding as i had hoped.
i've only had 20 hours of break-in so far, so i'll give it a few more days before coming to any final conclusions, but i think the 8620 did sound better even when freshly installed. it is certainly more detailed, and i think its more forward presentation was more accurate. perhaps the 2132 is more in need of some external biasing to get its juices flowing?
so far i am not so impressed with the results. the sound seems very recessed and distant now, with a stronger "opamp signature" - by that i mean it lacks air and harmonic detail, as if resolution drops with increasing frequency. some have attributed this quality to negative feedback; i don't know if that's an accurate characterization but it's what i've tended to hear from many opam circuits in the past. i also find the bass response to be less powerful, and it's not really any warmer sounding as i had hoped.
i've only had 20 hours of break-in so far, so i'll give it a few more days before coming to any final conclusions, but i think the 8620 did sound better even when freshly installed. it is certainly more detailed, and i think its more forward presentation was more accurate. perhaps the 2132 is more in need of some external biasing to get its juices flowing?
the original part was one of those yucky JRC chips... 4560 or whatever. it was garbage, and by virtue of high offset necessitated a blocking cap. since my replacements are always FET input parts, the blocking cap is gone, as are the muting transistors, and i have a 100 ohm Holco wired directly to the output jacks.
i just remembered i left some silly .047uF Panasonic PP film caps in there as supply bypasses. i installed them a while back in a fit of audiophile foolishness... need to take those out, i'm sure they're mucking up the sound. i'll replace them with some small-value Panasonic HFQ or FC right by the chips. also, i'm not convinced the Holcos are good resistors... in every project in the past that turned out on the bright side, they were a common denominator. one time i prototyped a circuit with some cheap carbon film resistors from radio shack, and in some ways i thought it sounded nicer.
i just remembered i left some silly .047uF Panasonic PP film caps in there as supply bypasses. i installed them a while back in a fit of audiophile foolishness... need to take those out, i'm sure they're mucking up the sound. i'll replace them with some small-value Panasonic HFQ or FC right by the chips. also, i'm not convinced the Holcos are good resistors... in every project in the past that turned out on the bright side, they were a common denominator. one time i prototyped a circuit with some cheap carbon film resistors from radio shack, and in some ways i thought it sounded nicer.
The player I was referring to was sold as SCD-555ES in Europe (I think that model number was something entirely different on the US market). Anyway, I had a brief look at the schematics again, and it is quite surprising for being a Sony. The I/V conversion uses both halves of a 2132. The mix stage uses one half of a 2132 and leaves the other hald unused!! Then comes two 6172s for the filter, and finally there is once again half a 2132 with the other half unused. They must really have had a reason to use the 2132, since they took the expense to use it just a single op amp in 2 cases. For both channels that means there are four duals used only as singles. Quite surprising for most commercial equipment, I think.
Christer said:
actually, it was probably cheaper for them to buy lots of 2132's in bulk, and use it all over the place, rather than have to stock two different chips in different quantities. not only in terms of bulk purchasing, but also simplifies the assembly line (fewer types of parts). same reason they'll use 47uF electrolytics as blocking caps when a 10uF would certainly do, because they're already using the 47uF elsewhere. i've never seen a single opamp IC used in mass-fi gear. in this case, using more chips was probably easier for layout as well.
i don't doubt Sony engineers have their reasons for designing the way they do, but a lot of their stuff is really overdesigned to the detriment of sound quality. your description of the complexity of the analog filter gives me the willies.
Yes, I was also thinking about the possibility that it may be cheaper to 2132s only, but still, in most other models they use cheap dual op amps and still don't waste half of them like that. However, my main point from the start was that they had choosen the 2132, rather than any of the more common duals in the same or lower price range, so they must have thought it was good in some respect.
Of course it may be overdesigned. My friend who has that player threw out the analog part and installed a Zapfilter instead.
Of course it may be overdesigned. My friend who has that player threw out the analog part and installed a Zapfilter instead.
ah, much better.
i got rid of the film cap bypasses by the opamp, replaced with a pair of Panasonic 10uF 50V FC i had in my parts bin. soldered one lead directly to the chip SMD pads, the other lead to an end of a ground bus bar Sony thoughtfully included on the board. (see, i'm not saying everything they do is terrible.) as i suspected (and carlos, jocko, or anyone else would have muttered
), the film caps were messing things up, and things are sounding much better now. it's smoother and less irritating without any loss of detail, as if some HF ringing has been eliminated... makes me wonder if the inductance of the film cap was indeed causing some ringing on the supply lines. woodwinds and female voice in particular are much nicer, and everything has better body and presence.
i have to wonder now if the issues i had before with the AD8620 were really just the film caps, and maybe i shoulda stuck to that chip.
oh well, i don't think the PCB can take much more soldering (i've already had to reconstruct one trace) so i'm probably stuck with the OPA2132 for now, at least until i get around to purchasing a nice temperature-controlled iron with a delicate tip. the 2132 doesn't sound *bad* at all, just not as good as the 8620, albeit when they were both messed up by the film cap. i'm sure after some break-in (plus the FC caps settling in) it'll open up some more.
i got rid of the film cap bypasses by the opamp, replaced with a pair of Panasonic 10uF 50V FC i had in my parts bin. soldered one lead directly to the chip SMD pads, the other lead to an end of a ground bus bar Sony thoughtfully included on the board. (see, i'm not saying everything they do is terrible.) as i suspected (and carlos, jocko, or anyone else would have muttered
), the film caps were messing things up, and things are sounding much better now. it's smoother and less irritating without any loss of detail, as if some HF ringing has been eliminated... makes me wonder if the inductance of the film cap was indeed causing some ringing on the supply lines. woodwinds and female voice in particular are much nicer, and everything has better body and presence. i have to wonder now if the issues i had before with the AD8620 were really just the film caps, and maybe i shoulda stuck to that chip.
oh well, i don't think the PCB can take much more soldering (i've already had to reconstruct one trace) so i'm probably stuck with the OPA2132 for now, at least until i get around to purchasing a nice temperature-controlled iron with a delicate tip. the 2132 doesn't sound *bad* at all, just not as good as the 8620, albeit when they were both messed up by the film cap. i'm sure after some break-in (plus the FC caps settling in) it'll open up some more.
Panasonic FM series
hm, i'll have to check out those FM caps. from the Digikey page:
sounds pretty good to me.
my parts bin was stocked back when FC caps were the "latest and greatest." i think i still have Panasonic HF and HFQ in there too...
as far as SMD caps go, is there more uniformity vs. through-hole parts? i.e. do they all pretty much suck, or are there "good" SMD caps? i'm thinking if i ever build some SMD circuits from scratch (don't hold your breath), i may just have to suck it up and put some small chip ceramics/tantalums right by the pins.
hm, i'll have to check out those FM caps. from the Digikey page:
sounds pretty good to me.
my parts bin was stocked back when FC caps were the "latest and greatest." i think i still have Panasonic HF and HFQ in there too...
as far as SMD caps go, is there more uniformity vs. through-hole parts? i.e. do they all pretty much suck, or are there "good" SMD caps? i'm thinking if i ever build some SMD circuits from scratch (don't hold your breath), i may just have to suck it up and put some small chip ceramics/tantalums right by the pins.
Dorkus, please give the OPA2132s what they really like: 47~100uF on each supply pin to ground, and a small (63~100V rating) 100~330nF MKT cap across V+ to V- pins.
But I must warn you: that Sony also needs a good clock, the great dynamics of the OPA2132 makes no favours to the untight bass of the player if you don't solve the problems behind it.
PS: please don't talk about Sony engineers, ok?
But I must warn you: that Sony also needs a good clock, the great dynamics of the OPA2132 makes no favours to the untight bass of the player if you don't solve the problems behind it.
PS: please don't talk about Sony engineers, ok?
Your parts bin sounds just like mine. You buy a ton of something, and then they end up not being used. You have to find stuff to put them in, only to find out something works better in its place.
I would avoid tantalums, unless you know how to derate them. Besides, they probably sound awful.
The SMD ones probably are nasty also. But that is why you get a good 'scope to see if you really need them. With luck you can get by using some stacked foil types. The wound ones with long leads will not cut it here.
Jocko
I would avoid tantalums, unless you know how to derate them. Besides, they probably sound awful.
The SMD ones probably are nasty also. But that is why you get a good 'scope to see if you really need them. With luck you can get by using some stacked foil types. The wound ones with long leads will not cut it here.
Jocko
well, that was silly.
i think i figured out why the OPA2132 isn't sounding so hot in my player.
it's running off +/-7V rails. i completely forgot.
i'm guessing the AD8620 probably performs better with such a low voltage supply, but obviously it still leaves a lot to be desired. now i remember why i went with the 8620 in the first place...
oh well. i suppose if i want to make any serious improvements, i'll need a whole new output section and power supply, with a new clock to boot. i'd go with an outboard DAC, but i'd like to keep the ability to play SACDs, dying format that it is.
i think i figured out why the OPA2132 isn't sounding so hot in my player.
it's running off +/-7V rails. i completely forgot.
i'm guessing the AD8620 probably performs better with such a low voltage supply, but obviously it still leaves a lot to be desired. now i remember why i went with the 8620 in the first place...
oh well. i suppose if i want to make any serious improvements, i'll need a whole new output section and power supply, with a new clock to boot. i'd go with an outboard DAC, but i'd like to keep the ability to play SACDs, dying format that it is.
Christer said:
The SCD-333ES too, and it also uses OPA2132 (SMD) opamps all over.
I didn't change the OPA2132s there, but I used an awful amount of caps to bypass all that properly.
IMO this is a very, very bad player. Looks nice on a pic, many good parts, but when you really start looking at the circuit...
carlosfm said:
The SCD-333ES too, and it also uses OPA2132 (SMD) opamps all over.
I didn't change the OPA2132s there, but I used an awful amount of caps to bypass all that properly.
IMO this is a very, very bad player. Looks nice on a pic, many good parts, but when you really start looking at the circuit...
Well, as I said previously, it's a friend of mine who has the 555 and he bypassed the whole analog section with a LCAudio Zapfilter and thinks it sounds much better that way. However, my original point was that whether Sony succeeded or not, at least they did take a lot of extra costs in this player compared to most of their other ones, and it is interesting that they did choose the not so common 2132. It also has dual R-core transformers etc.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.