That capacitor is part of the input filter of that stage. Reducing it will reduce loss at say 20Khz but will also let in more HF trash coming from the DAC. Increasing it will roll off the HF significantly . You need to simulate it to see what the 20 Khz drop is. I doubt it is very much. Lifting by 1 dB isn't going to make any audible difference.
I'll run some tests on my amp later. Maybe something might make it sound brighter but it isn't that cap you are playing with ! My amp has clean HF and very good transient bite . It is slightly warmer than some other amps when compared side by side but on it's own it sounds fine.
I actually bought it after they had reviewed it and said that the 667 had very good sound ! I am not disappointed !
I might add that straight out of the box it did not sound as it does now. It was more rolled off ! It is used every day at least 8 hours since I bought it in 2012.
Even the TV is run through the amp. So maybe it's run in well ?
I'll run some tests on my amp later. Maybe something might make it sound brighter but it isn't that cap you are playing with ! My amp has clean HF and very good transient bite . It is slightly warmer than some other amps when compared side by side but on it's own it sounds fine.
I actually bought it after they had reviewed it and said that the 667 had very good sound ! I am not disappointed !
I might add that straight out of the box it did not sound as it does now. It was more rolled off ! It is used every day at least 8 hours since I bought it in 2012.
Even the TV is run through the amp. So maybe it's run in well ?
No, and No.
See the message above. Also not (C1+C2)/2 but C/2 as you correctly write in a sentence but incorrectly in the formula.
That filter is -0.03dB down at 10Khz and -0.23 dB down at 20 Khz.
You will gain nothing by changing any passive parts.
However changing the opamp might make a difference though you cannot be sure what exactly will happen.
I've replaced the NE5532 in a similar position in another DAC. The treble was sharper but it lost the bass weight and impact and transients did not have slam anymore. I used the LME49720 and a few others. In spite of the slight reduction in 'brightness' the 5532 was preferable overall. I didn't find the piano sounding dull or things like that.
Change your speaker wire first ! You might get surprised. Try 21 swg enameled wire. For a 2 meter length one pair of wires in parallel for each of the +ve and _ve lines works well. For longer runs use three or four strands per +/- line. Cheap and easy to try out. You might get back the 'edge' you feel is missing in your system. Use cotton sleeves to insulate the +/- lines. Available in the local electronic parts market.I didn't use anything as I had new wire without damage to the enamel insulation.
You will gain nothing by changing any passive parts.
However changing the opamp might make a difference though you cannot be sure what exactly will happen.
I've replaced the NE5532 in a similar position in another DAC. The treble was sharper but it lost the bass weight and impact and transients did not have slam anymore. I used the LME49720 and a few others. In spite of the slight reduction in 'brightness' the 5532 was preferable overall. I didn't find the piano sounding dull or things like that.
Change your speaker wire first ! You might get surprised. Try 21 swg enameled wire. For a 2 meter length one pair of wires in parallel for each of the +ve and _ve lines works well. For longer runs use three or four strands per +/- line. Cheap and easy to try out. You might get back the 'edge' you feel is missing in your system. Use cotton sleeves to insulate the +/- lines. Available in the local electronic parts market.I didn't use anything as I had new wire without damage to the enamel insulation.
Thanks ilimzn and ashok for writing.I have asked nothing wrong but wrote on earlier post which had a formula that was giving final figure which I never learnt in science.Secondly my doubt on using on 2 electrolyte caps in opposite also was partially correct.These corrections may be needed for other readers and not try it blindly elsewhere.
Yes,I do read a lot and have a little experience in diy other than few kits.So I did ask to share your experiences which will be helpful.
regards.
Yes,I do read a lot and have a little experience in diy other than few kits.So I did ask to share your experiences which will be helpful.
regards.
Hi,
A bipolar capacitor IS essentially two back to back polar capacitors,
except that the foil is back to back series dual layers in one device.
I suggest you haven't tried your "myth busting" exercise, it will be fine.
Each polar layer or capacitor protects the other from reverse bias.
You don't seem to understand what a "REAL" bipolar actually is.
rgds, sreten.
Last edited:
yes I'm testy, short on patience on this issue that I have corrected dozens of times
nonpolar/bipolar electrolytic caps have full thickness oxide grown on both foils
you apparently haven't read Bateman's articles, even Electro Cap literature
http://www.nichicon.co.jp/english/products/pdf/aluminum.pdf
please put your eyes on the refs, link them here instead of pulling the "corrections" out of some dark orifice, repeating old wives tales or even correctly remembered but Wrong "audio guru" tripe
nonpolar/bipolar electrolytic caps have full thickness oxide grown on both foils
you apparently haven't read Bateman's articles, even Electro Cap literature
http://www.nichicon.co.jp/english/products/pdf/aluminum.pdf
please put your eyes on the refs, link them here instead of pulling the "corrections" out of some dark orifice, repeating old wives tales or even correctly remembered but Wrong "audio guru" tripe
Last edited:
THANK YOU Jcx!
This is the fundamental difference between real bipolar caps and anti-series connected polars. In the latter case there is a 'middle' that has NO oxide hence the oxide on the other side of this will NOT protect the oxide on the forst side when it is reverse polarized.
@Sreten:
How about YOU try my myth busting exercise - say, in a speaker. Make sure to wear protective glasses. In fact, do connect scope probes in the middle of the connection and see for yourself. Needles to say I _HAVE_ tried it. And I have done some research when and why you can occasionally use this technique, this goes across literally hundreds of cases.
People all too often ASSUME that this anti-series connection will be used as a coupling cap where the DC component across the combo will be largely constant. And we all know where assumptions tend to lead...
Ashok,
In a good system (namely no weal links downstream, starting with ears, room, speakers) 1dB drop at 20k can SUBTLY change the sonic signature of the system. It can sound sligthtly softer, subdued.
I've done a lot of CD player tweaks which often indeed do have to do with optimizing filters (usually because high tolerance caps and resistors get exchanged with low tolerance, making the channels better balanced but sometimes not all values are available!) and it does have an effect (I usually listen to the CPD through headphones first) but as you say it is indeed subtle.
Oh, one thing which isn't subtle is changing cheap ceramic caps or cheap inductive poliethilenes with decent NP0/C0G ceramics, silver mica or polystyrene. Thisng so indeed clear up. It can also be measured (gets down to capacitor distortion) but not easily. HF IMD tests show it (say 15+16k 1:1). Still - this will not be easily audible if the rest of the system isn't up to the task.
I also have to agree that despite it's age and new OPamps, the venerable NE5532 is not at all easily replaceable, definitely not as a 'sure thing', cookbook style 'exchange 5532 for xxxxx and youa re guaranteed a greatg result'.
Well I have tried some of good brands from Supra clasic 4, Audioquest 14/2, Monoprice 16awg,DAC cable 14 awg and few others(I have all with me),but nothing improved more than 7%.I also use headphones to check sound quality and Yamaha is fine ,but warm in pure direct.I have tried Yamaha in treated demo room and other speakers too.It sounded thin to me and my hifi friends compared to other brands.After reading Denon 1909 make over thread,I thought of improving Yamaha too.
Denon AVR-1909 'Extreme Makeover' | Audioholics Home Theater Forums
He says he can listen to Denon in pure direct too and doesn't sound dull.
You are the perfect guy to try out the 'inexpensive' enamel wire 'cable' .
For 15 feet you will need slightly more as you will loose some length while twisting them lightly. Maybe 17 feet ? You will need to buy about 200gms or 1/4 kg. You can parallel 4 lengths for each +ve and -ve cable. The weight is for a stereo pair. ( 21swg). Must cost under Rs200/- ! Use solderable Banana plugs available in the electronics market.
About parts change , it depends on what the passives are like. Better to identify them all before considering changes. I did look at that page you linked. Yes it does make it sound quite attractive but you must know what went into your AVR. I might actually open it and have a look ! In any case I need to vacuum it as this place is quite dusty.
Pity I have very little time as I'm going out of town for three weeks.
Pity I have very little time as I'm going out of town for three weeks.
Some sense at last.I just typed in a big long rant about getting a grip on yoursleves over this stuff. Then I deleted it. So I will post a new rant.
I dread to think how many speakers, amplifiers, preamps, crossovers and whatnot I have designed and built. So does my wife
Some of these are / were top of the line, some PA, some middle of the road and some archaic. The strange thing that I have learnt from this is that gross aberrations aside, (which almost always are loudspeaker related):
- Just about any configuration can sound pretty decent.
- Loudspeakers and the room are the #1 issue
- Crossover configuration is equally important, in fact I would say it comes in #0
- Cruddy source material or matching / earthing comes next
- If you are looking at op amps, resistors and capacitors, you are definitely losing the plot
Here is a secret from an old bugger:
- Get a decent bottle of red
- Get a comfy chair
- Get your favourite piece of music
- Stop "listening to the equipment", sit back, have a red and enjoy some music.
If after that, or earlier, you can't stand it then go looking for the gross error in the system.
All the smoke and mirrprs that surround audio upgrades are even more pointless on a PA mixer like the Yamaha.
I would like to update on this.I just compared 667 with 671 AVR and found that it uses same DAC chip as 667,but opamps are LM833MX. Circuit is almost similar,but value of C4240(near input)is polypropolene 820pF. Voltage used is 8-0-8V in stead of 12-0-12V in 667.I have heard 671 U.S. model and it sounded powerful and more dynamic with Pioneer AJ speakers.
Yamaha 671 & Pioneer SP-PK51FS ! The journey starts
So can opamp 5532 be replaced by LM883MX keeping 12V supply?Will there be issue? Looking at data sheet,it can be run on 5 to 15V-
http://www.ti.com/lit/ds/symlink/lm833-n.pdf
I compared the 667 to the Creek 5350-II power amp. As expected the 5350 had more transient attack and tighter control on the bass. Did make it sound more dynamic. But the difference was noticeable though not huge .....my opinion ! Depends on the music also I guess ( I used Special EXF, Dave Grusin etc.) . However the treble doesn't sound too rolled off or dull ! My source is an OPPO BDP93 . Note it isn't just the capacitor, the resistors around it also matter. I can see you are not familiar with circuit topology. Makes it much harder to achieve something. Good to learn it , or brush up what you learnt .
This difference we hear doesn't come from just parts change ! Especially the LM833. I am quite familiar with that chip from a long time ago. It ranks below the 5532 ! I know, I have that chip somewhere . I used it several years ago ( say 15 years or more !). But you never know, it might just be the electrolytic coupling caps ! I have not opened it yet so I don't know what brand / type they are.
If you really want a drastic change I'm guessing you might need a new amp ! Meddling with your amp will reduce its resale value if you don't end up liking the changes. I'd really think the best thing is to take your amp to a system you like and compare it there. You probably did this already. If you can live with the difference it will be great. If not sell it and get a new one. The monetary loss might be less than what you might eventually spend ( time and money ) doing mods.
But spend enough time comparing your amp with others with only the amp changed in the system.
If you can borrow the amp you feel sounds much better , take it home and put it in place of your amp and see how it sounds there. If your amp is close enough then let things stay as they are and enjoy the music !
Next time when you upgrade you can pick your preferred amp more carefully.
I'd keep my 'stereo' system separate from the HT system ! I might add that a decent class D will get you what you want very easily !
Bottom line. My suggestion would be to build a kit stereo power amp, chip amp, or discrete or classD from an Ebay kit. Many sound superb . If you don't listen to stereo too loud , you can get a SMSL 50 watt per channel classD amp. Sounds great and fantastic value for money !
This difference we hear doesn't come from just parts change ! Especially the LM833. I am quite familiar with that chip from a long time ago. It ranks below the 5532 ! I know, I have that chip somewhere . I used it several years ago ( say 15 years or more !). But you never know, it might just be the electrolytic coupling caps ! I have not opened it yet so I don't know what brand / type they are.
If you really want a drastic change I'm guessing you might need a new amp ! Meddling with your amp will reduce its resale value if you don't end up liking the changes. I'd really think the best thing is to take your amp to a system you like and compare it there. You probably did this already. If you can live with the difference it will be great. If not sell it and get a new one. The monetary loss might be less than what you might eventually spend ( time and money ) doing mods.
But spend enough time comparing your amp with others with only the amp changed in the system.
If you can borrow the amp you feel sounds much better , take it home and put it in place of your amp and see how it sounds there. If your amp is close enough then let things stay as they are and enjoy the music !
Next time when you upgrade you can pick your preferred amp more carefully.
I'd keep my 'stereo' system separate from the HT system ! I might add that a decent class D will get you what you want very easily !
Bottom line. My suggestion would be to build a kit stereo power amp, chip amp, or discrete or classD from an Ebay kit. Many sound superb . If you don't listen to stereo too loud , you can get a SMSL 50 watt per channel classD amp. Sounds great and fantastic value for money !
Last edited:
Thanks Ashok for advice and help.Yes I am from Medical field and my curiosity attracts me towards electronics and sound. As you know that I have already done all experiments(getting friends amps home,marantz,Denon),I concluded Yamaha has better sound for movies,but lack that extra weight. That's why I looked into service manual and asked in the forum.I also have a stereo setup with 2 DIY pre(DCB1 and SSP)paired to rotel power amp.Source is asus xonar D2X> digitail out to Subbu DAC.After trying so many combinations,I found Yamaha is little behind with sound quality which I liked.
As in my earlier post,when I connected pioneer DVD player's analog out to Yamaha analog in,the sound was 20% superior to the sound connected same player via HDMI(same song).So I questioned DAC section.Did you compare power amp separately or connected to Yamaha preout?
If Oppo is connected to Yamaha with analog out and HDMI,you may know what I am trying to express.
Yes oppo has better DAC,but my example was pioneer DVD player.
Anyway I will prefer learning to brush up so far I learnt.
As in my earlier post,when I connected pioneer DVD player's analog out to Yamaha analog in,the sound was 20% superior to the sound connected same player via HDMI(same song).So I questioned DAC section.Did you compare power amp separately or connected to Yamaha preout?
If Oppo is connected to Yamaha with analog out and HDMI,you may know what I am trying to express.
Yes oppo has better DAC,but my example was pioneer DVD player.
Anyway I will prefer learning to brush up so far I learnt.
Last edited:
In what kind of circuit did you come to this conclusion? IMHO it makes no sense to talk about the sonic properties of opamps in isolation.
Looking at Samuel Groner's measurements, the datasheet and other information floating around on the web, I can only think of a few things that could make the part grossly misbehave:
1. Capacitive loading - the datasheet advises an isolation resistor from 50 pF (!) up but neglects going into more detail, though I guess 220 ohms would get the job done (reportedly the old TL07x is about the only kind of common audio opamp that can use up to 470). RIAA networks with "Neumann pole" preferred.
2. Severe output loading - unlike a 5532, it's not really a load driving champ, clipping at about 6 Vrms into 600 ohms with +/-15 V supplies. 2k up preferred.
3. Sloppy layout and rail decoupling - a 9 MHz fT would have been reasonably fast for an audio part at the time. (1.) would not make matters any better.
Input bias current also is a bit high, and input impedance distortion about 3-4 times worse than for a '5532 (if still better than typical JFET input parts), so this may not be the ideal part to use after a 100k volume pot. The latter may also suggest highish input capacitance (not totally unexpected in a low-noise part), mandating either lowish feedback network impedance or appropriate compensation. Common-mode distortion is so-so but acceptable, lacking the conspicuous degradation towards HF (in CMRR as well, I suppose) found in '5532s.
Overall the LM833 shares some typical traits of older low-noise opamps, particularly the "dirty little secret" that you can't fully exploit noise levels at low gain when high signal levels at lowest distortion are desired - it's far from the worst offender though, with an estimated noise penalty of about 3-5 dB at a gain of 2. (A TI MC33078 is something I suggest to be best used with some Class A bias and a buffer stage.)
Looking at Samuel Groner's measurements, the datasheet and other information floating around on the web, I can only think of a few things that could make the part grossly misbehave:
1. Capacitive loading - the datasheet advises an isolation resistor from 50 pF (!) up but neglects going into more detail, though I guess 220 ohms would get the job done (reportedly the old TL07x is about the only kind of common audio opamp that can use up to 470). RIAA networks with "Neumann pole" preferred.
2. Severe output loading - unlike a 5532, it's not really a load driving champ, clipping at about 6 Vrms into 600 ohms with +/-15 V supplies. 2k up preferred.
3. Sloppy layout and rail decoupling - a 9 MHz fT would have been reasonably fast for an audio part at the time. (1.) would not make matters any better.
Input bias current also is a bit high, and input impedance distortion about 3-4 times worse than for a '5532 (if still better than typical JFET input parts), so this may not be the ideal part to use after a 100k volume pot. The latter may also suggest highish input capacitance (not totally unexpected in a low-noise part), mandating either lowish feedback network impedance or appropriate compensation. Common-mode distortion is so-so but acceptable, lacking the conspicuous degradation towards HF (in CMRR as well, I suppose) found in '5532s.
Overall the LM833 shares some typical traits of older low-noise opamps, particularly the "dirty little secret" that you can't fully exploit noise levels at low gain when high signal levels at lowest distortion are desired - it's far from the worst offender though, with an estimated noise penalty of about 3-5 dB at a gain of 2. (A TI MC33078 is something I suggest to be best used with some Class A bias and a buffer stage.)
Last edited:
My favorite test is to take 10 op amps and use sockets to hear what they sound like with 10 in a row operated at different gains and different impedances for the feedback and input. Just a half dozen boards with different set ups for these will really tell a lot about how the op amp sounds. Most SLS with a few being okay and a very few sounding far better than they have any right to sound. This is also a handy way to test for IM and transient distortions. The 833 was a big winner at making transient distortion no matter the setup. The 833 was discontinued but see it is now back to be the plague of bad sound it was then. More TI junk. No TI allowed here.
Your mention of "dirty little secret" made me laugh. Of course the input noise come around the feedback loop and makes the device noisy. Known by engineers but probably not amateurs who have no idea how to calculate a bandwidth noise product. Funny stuff!
As for speaking of "op amps in isolation" there are some op amps that no matter what circuit they are put into sound like ----, the 5532 is one and the 833 is another. These are just bad op amps. By what I get from your statements possibly the 741 should do the trick always. Maybe I read this wrong. I would never embarrass myself suggesting anything TI. TI is widely known for making junk parts and will not allow their use here in any product or design. There is always another way without the garbage of TI assuring premature failure and poor performance.
Your mention of "dirty little secret" made me laugh. Of course the input noise come around the feedback loop and makes the device noisy. Known by engineers but probably not amateurs who have no idea how to calculate a bandwidth noise product. Funny stuff!
As for speaking of "op amps in isolation" there are some op amps that no matter what circuit they are put into sound like ----, the 5532 is one and the 833 is another. These are just bad op amps. By what I get from your statements possibly the 741 should do the trick always. Maybe I read this wrong. I would never embarrass myself suggesting anything TI. TI is widely known for making junk parts and will not allow their use here in any product or design. There is always another way without the garbage of TI assuring premature failure and poor performance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.