Hi, I would like upgrade this old preamp amplifier, the actual opamps are NJM4580E smd, I can get NE5532 smd and OPA2134 smd very cheap. Is possible a direct replacement without addind decopling caps or heavy mods? Do you think is worth this upgrade? Forget the filter circuit part, it will be bypassed
Thanks
Thanks
It's a SONY, eh? Mid-'80s to mid/late-'90s or thereabouts, or that's what I get from the schematic style.
4580Es are in EMP-8 package, which however appears to be almost the same as SOIC-8.
First off, you most probably won't be seeing any dramatic improvements.
That said, let's go through the various stages:
IC801, 807: Balanced input impedances (11k each), no common-mode signal. Output loading can get pretty severe if gain is cranked up. So you need good transfer linearity and good load-driving ability, with mid-low voltage and current noise. OPA2134 is inferior at load driving from about 1 kHz up, so the 5532 it is then. Potential upgrades with yet better performance: LT1469, LT1630.
IC802, 808: An inverting amplifier with potentially low impedances, but receiving noise from preceding stage. Output loading not particularly severe. Pretty much a toss-up, use whichever you like. Alternative: LM4562 (LME49720/49860).
As far as the tone controls are concerned, hmm. Not an expert on these. It seems like the impedances seen by inputs are mid-high and not necessarily matched here, mandating well-controlled input impedance distortion. You could really do a lot worse than with the 5532 here (the 2134 is not even in the ballpark). Some people also like other bipolar parts with moderate input stage current in such applications, like the MC33178 in spite of its rather modest specs (now if that one works well, something like a 4556 also should).
4580Es are in EMP-8 package, which however appears to be almost the same as SOIC-8.
First off, you most probably won't be seeing any dramatic improvements.
That said, let's go through the various stages:
IC801, 807: Balanced input impedances (11k each), no common-mode signal. Output loading can get pretty severe if gain is cranked up. So you need good transfer linearity and good load-driving ability, with mid-low voltage and current noise. OPA2134 is inferior at load driving from about 1 kHz up, so the 5532 it is then. Potential upgrades with yet better performance: LT1469, LT1630.
IC802, 808: An inverting amplifier with potentially low impedances, but receiving noise from preceding stage. Output loading not particularly severe. Pretty much a toss-up, use whichever you like. Alternative: LM4562 (LME49720/49860).
As far as the tone controls are concerned, hmm. Not an expert on these. It seems like the impedances seen by inputs are mid-high and not necessarily matched here, mandating well-controlled input impedance distortion. You could really do a lot worse than with the 5532 here (the 2134 is not even in the ballpark). Some people also like other bipolar parts with moderate input stage current in such applications, like the MC33178 in spite of its rather modest specs (now if that one works well, something like a 4556 also should).
Hallo ! Yes, you are right, Is a old Sony XM 7547 amp (purchased from Germany). I Like the build quality of the old Sony Stuff 
Thanks for your clean explication, I'm impressed for your level of knowledge, I will follow your recomendations. The tone controls can be bypassed with a external switch so It will be controlled with the head unit active filters.
I have some small wima's Mks2 4.7uf 50v. and I will replace the input electrolitics caps, maybe is a little upgrade too.
Thaks again for you kind support !
Gruss
Bernat
Thanks for your clean explication, I'm impressed for your level of knowledge, I will follow your recomendations. The tone controls can be bypassed with a external switch so It will be controlled with the head unit active filters.
I have some small wima's Mks2 4.7uf 50v. and I will replace the input electrolitics caps, maybe is a little upgrade too.
Thaks again for you kind support !
Gruss
Bernat
Attachments
Not an expert per se, just been poking my nose into Samuel Groner's op-amp measurements and interpreting them.
Incidentally, from this page it would seem that NJM4580 has better current drive capability than NE5532 (or at least that's what I get from the results - unfortunately the test setup used gets multiple distortion mechanisms involved, with output loading often being the deciding factor), so the latter may be performing worse at high gain settings.
Along with the "upgrade" types that I suggested for IC801/807, you could also try a NJM4556 - its noise still is negligible there, and being a headphone driver it's not easily bothered by low-impedance loads. I think that one comes in DMP-8 though, which is a hair bigger than SO-8.
Incidentally, from this page it would seem that NJM4580 has better current drive capability than NE5532 (or at least that's what I get from the results - unfortunately the test setup used gets multiple distortion mechanisms involved, with output loading often being the deciding factor), so the latter may be performing worse at high gain settings.
Along with the "upgrade" types that I suggested for IC801/807, you could also try a NJM4556 - its noise still is negligible there, and being a headphone driver it's not easily bothered by low-impedance loads. I think that one comes in DMP-8 though, which is a hair bigger than SO-8.
thanks again for your time !
the amp arrived yesterday and after log time of audition I can say I dont't like his sound, is flat and boring, the treble is velled and the soundstage dissapeared. respect to my reference car amp (In this amp changed de opa2604 for opa2134 with very good improvement, I tried too the OPA627AU but was impossible to listen , no bass, no highs just a very distorsioned midrange, probably oscillation problems ).
¿Do you think I will get oscillation problems with the OPAMs you recommend, is already compensated the circuit in the the scheme?
Kind REgards
Bernat
the amp arrived yesterday and after log time of audition I can say I dont't like his sound, is flat and boring, the treble is velled and the soundstage dissapeared. respect to my reference car amp (In this amp changed de opa2604 for opa2134 with very good improvement, I tried too the OPA627AU but was impossible to listen , no bass, no highs just a very distorsioned midrange, probably oscillation problems ).
¿Do you think I will get oscillation problems with the OPAMs you recommend, is already compensated the circuit in the the scheme?
Kind REgards
Bernat
Well, due the difficults to find the recommended parts I replaced all the opamps with LM4562 just for experiment, I like overall result , the upper midrange seems more relaxed and easy to listen but I think I loosed some bass and dynamics. Maybe I need to replace (or remove) some resistors of the opamp circuit? Thanks
It appears there are two 220 µF electrolytics and two 100 nF ceramics for all the opamps in the input section, with no further local bypass caps. Silly of me not to pay attention. This layout is probably not suited for anything remotely critical WRT decoupling, so it's not a surprise that the OPA627 wasn't amused.
Speaking of the OPA627, I hope an adapter was used for two each?! That would have been the ideal place for some local decoupling caps.
I wish people would just forget about the OPA2604 for run of the mill opamp duties. The strength of this part is operation at higher supply voltages than usual. Only then is it remotely attractive. A bit finicky about decoupling, too.
LM4562s should be reasonably tolerant but still aren't totally uncritical. You can still get them to oscillate.
4556s for IC801/807 should work fine though.
Speaking of the OPA627, I hope an adapter was used for two each?! That would have been the ideal place for some local decoupling caps.
I wish people would just forget about the OPA2604 for run of the mill opamp duties. The strength of this part is operation at higher supply voltages than usual. Only then is it remotely attractive. A bit finicky about decoupling, too.
LM4562s should be reasonably tolerant but still aren't totally uncritical. You can still get them to oscillate.
4556s for IC801/807 should work fine though.
Hi Sgrossklass, now is running the LM4562 in the input IC and the inverting IC, I decoupled V+ and V- directly to the opamp Pins with 47uf elna silmic . Sounds better than before, much better, I already ordered a pair of NJM4556 for IC801/807 I'm sure its much better for this circuit as you say.
Have a nice weekend
I notice the pots are not AC coupled in this design. Using opamps with high input bias currents could lead to audible noise as pots are turned.
Personally I would try something like the FET OPA2134 which eliminate the bias issue. From what I can see the stages are heavily compensated so I would not forsee any obvious stability issues but as always a check with a scope is advisable.
http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
A less obvious choice might be the TLE2072 Excaliber series which have very high current drive capability (-/+80ma).
Personally I would try something like the FET OPA2134 which eliminate the bias issue. From what I can see the stages are heavily compensated so I would not forsee any obvious stability issues but as always a check with a scope is advisable.
http://www.diyaudio.com/forums/anal...u-have-checked-see-its-stable-havent-you.html
A less obvious choice might be the TLE2072 Excaliber series which have very high current drive capability (-/+80ma).
Thanks Mooly, I've been reading the Opamp thread and I'm thinking to buy a cheap used oscilloscope. Next step I will do is to buy come Smd to DIP socket adapters, much more easy to swap ic than soldering every time and the plus of adding local decoupling in the socket.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.