"But I couldn't hear any change sound quality."
There are many, many caps in the signal path.
Going back to the dirty window senario, only cleaning one may not really help that much.
I generally use 0.033µF 50V polypropylene for bypass duty for the signal path, they are very small. Mylar is OK too. Digi-Key has 0.033µF 100V parts in stock for about $0.35 each in one-lot quantity. I think they specal ordered the 0.033µF 50V polypropylene for me (large quantity).
There are many, many caps in the signal path.
Going back to the dirty window senario, only cleaning one may not really help that much.
I generally use 0.033µF 50V polypropylene for bypass duty for the signal path, they are very small. Mylar is OK too. Digi-Key has 0.033µF 100V parts in stock for about $0.35 each in one-lot quantity. I think they specal ordered the 0.033µF 50V polypropylene for me (large quantity).
Well,I even tried a wire to bypass that coupling cap,but no change in sound.If I triy bypassing caps in my DIY pre and yes I can hear a lot improvement even via in headphone.I got those caps from this one-
An externally hosted image should be here but it was not working when we last tested it.
[/QUOTE]
Even with my DAC,change in C22 (near DAC chip), C17(near clock) with different caps,make a lot difference.
I used Panasonic cap in C22 and sound became sharp.When I used Nichicon,sound becomes warm with better soundstage to my liking.
So do I have to change all the coupling caps at a time and then check in Yamaha667?
You need an oscilloscope to do this. Find the supply rail +ve or - ve rail and monitor it. The changes that you can very easily see on an oscilloscope will not be easily visible on a DMM.
However my 667 sounds great even at elevated volume and maybe you needn't meddle with it. Actually I'll check this for you later this week as I have the same amp.
It isn't safe to test the powered up amp by connecting a dense board with a loose capacitor . If you accidentally touch a high voltage dc track and then connect to your test point, it 'might' blow something. Depends on what you are touching. You could solder leads to the point and bring it out and add the part .However the leads will add inductance to that point. If you blow your amp, you had it !
Last edited:
Thanks for kind advice.
Actually DAC board is at extreme right and separate from main board.
I used headphone jack to test and not to speakers.Board uses not more than 12V near opamp.Fortunately I have steady hands.
Still got surprised why there was not any change in sound quality when other devices do respond?Like DAC.
Last edited:
I suspect that if many knew how many rack mount boxes are in the production chain, and what is inside those boxes and understood the principle of the weakest link, they would pass out on the spot.
Long story short, if the active devices in those production boxes are 553x and they are coupled by NP electrolytics it is a very, very good day, indeed.
What they think they hear, they almost certainly hear through the magic of sighted evaluations. The cosmic dice of technology roll against them.
Yes, I went through a big desk years ago and found 10 polarized electrolytics in the signal path of each channel strip, and I can't remeber how many 4558's.
It really opened up after I upgraded both the caps and the opamps.
Recently did a big Allen and Heath desk, only four channels at first. Big difference there too.
I did a Yamaha stereo 31 band graphic, one channel only. Was trying to sell the owner on the idea of some upgrades. It sounded so different he couldn't use it in his stereo FOH, he wouldn't pay, so it ended up in the foldback system.
It really opened up after I upgraded both the caps and the opamps.
Recently did a big Allen and Heath desk, only four channels at first. Big difference there too.
I did a Yamaha stereo 31 band graphic, one channel only. Was trying to sell the owner on the idea of some upgrades. It sounded so different he couldn't use it in his stereo FOH, he wouldn't pay, so it ended up in the foldback system.
After your kind suggestions and before I reopen my Yamaha,I was reading all about caps.
Improve sound quality
But I was surprised to see that Yamaha's Flagship Z9 much praised for its sound quality
Help me to choose between these MEGA MONSTERS!!! - Page 24
also use Polarized Electrolyte capacitors after DAC -
Improve sound quality
But I was surprised to see that Yamaha's Flagship Z9 much praised for its sound quality
Help me to choose between these MEGA MONSTERS!!! - Page 24
also use Polarized Electrolyte capacitors after DAC -
Even I checked,I get output voltage around 2.5V(ar per manual),but after NJM5532,there is nothing(hope my DMM read correctly).
Yes, it appears that DAC have an offset of around 2.5V. This is probably related to the 5Vdc that DAC work from.
The purpose of the capacitor is to block the 2.5V offset from affecting subsequent circuits.
The capacitor has a double duty. It blocks DC and must pass the AC audio signal. This coupling duty generally only requires that a commercial quality polar electrolytic with lowish leakage be selected such that it passes ALL the audio signal.
The purpose of the capacitor is to block the 2.5V offset from affecting subsequent circuits.
The capacitor has a double duty. It blocks DC and must pass the AC audio signal. This coupling duty generally only requires that a commercial quality polar electrolytic with lowish leakage be selected such that it passes ALL the audio signal.
Hello,
According to Cycil Batemans capacitor research (available on the net), bipolar capacitors, even with a voltage bias had much less distortion than any polar cap. If you bump up the size and voltage rating as well, you can get still less distortion. Believe it was his fourth and fifth artiles that had sections that talked about electrolytics. FWIW.
Putting two bipolars in series proved to cut the distortion even more in his later tests. Much easier and cheaper to fit bipolar caps than film caps of a reasonable size.
Regards,
Greg
According to Cycil Batemans capacitor research (available on the net), bipolar capacitors, even with a voltage bias had much less distortion than any polar cap. If you bump up the size and voltage rating as well, you can get still less distortion. Believe it was his fourth and fifth artiles that had sections that talked about electrolytics. FWIW.
Putting two bipolars in series proved to cut the distortion even more in his later tests. Much easier and cheaper to fit bipolar caps than film caps of a reasonable size.
Regards,
Greg
Good point, but even with a DC bias electrolytics still have room for improvement.
Post #11 shows back-to-back electrolytics with pads on the board for a bypass, it would be so simple to just add the bypass.
Post #19 shows polarized electrolytics with no DC bias, and pads for power supply bypass caps that are not installed.
I used to sell and service Nakamichi. Some of the models had similar pads for a bypass. The better models had a better opamp, and the bypass cap. It cost less than $1 to change the opamp and add the bypass cap. At that point they sounded the same (same DAC and you spent $500 less too).
Post #11 shows back-to-back electrolytics with pads on the board for a bypass, it would be so simple to just add the bypass.
Post #19 shows polarized electrolytics with no DC bias, and pads for power supply bypass caps that are not installed.
I used to sell and service Nakamichi. Some of the models had similar pads for a bypass. The better models had a better opamp, and the bypass cap. It cost less than $1 to change the opamp and add the bypass cap. At that point they sounded the same (same DAC and you spent $500 less too).
Last edited:
The capacitor tests show a hierarchy of signal quality.
But the test uses the capacitor as part of a filter.
The actual distortion levels which in some of the tests is extremely low ONLY applies when these capacitors are used as filters, i.e. when there is a large signal voltage across the capacitor.
In coupling duty, where there is virtually no audio voltage across the capacitor, the distortion drops by an enormous amount.
That is why the test does not cover coupling duty, the levels of distortion contributed by the capacitors would in general (all the back to back electrolytics and better) require a different and more sensitive measurement method to show any added distortion.
As far as I know, no one has shown distortion, or similar, measurement results for capacitors used as signal coupling when there is virtually no signal voltage across the coupling capacitor.
Back to back polar electrolytics when correctly sized to pass all the audio signal are perfectly good for coupling duty.
This is confusing,some agree that polarized caps do send reverse voltage in half cycle.Still they are good for coupling?
My simple question,if capacitor cant pass DC voltage across,then why most of these companies use bipolar(some are cheaper still)?
Hello,
A bipolar electrolytic cap has quite a bit less distortion than a polarized one.
Putting two electrolytic polarized caps in series with polarity switched on one of them results in a bipolar cap 1/2 the size of one of the polar ones used.
C1 C2
IE: -----------||------------||------------
+ - - +
Result cap = bipolar (C1+C2)/2
(assuming both caps are the same size.)
Regards,
Greg
A bipolar electrolytic cap has quite a bit less distortion than a polarized one.
Putting two electrolytic polarized caps in series with polarity switched on one of them results in a bipolar cap 1/2 the size of one of the polar ones used.
C1 C2
IE: -----------||------------||------------
+ - - +
Result cap = bipolar (C1+C2)/2
(assuming both caps are the same size.)
Regards,
Greg
That equation says final capacitance will be same as per cap use.But I think final value should be half of that of individual.
i.e. if 10uF caps used in series,then final capacitance should be 5uf since they are parallel.
Another question,does 2 caps used that way work actually as bipolar by holding charge or just repel DC since at both ends of those will be +ve?
I did add one small cap parallel to C4240 which has a small value of 680pF which is at entry point to NJM5532 of right channel.Immediate impression,sound lost details.Obvious reason that this cap connects signal path to ground.May be that's why Yamaha sounds very warm in pure direct.
Now my question is,if I lower that value,can it help to improve details(say high frequencies?).Some of signal is certainly getting grounded.Is it safe to lower value?Just need experts advice.
Yo keep poking around at this circuit which is actually quite complex, expecting to somehow 'simply' learn what each capacitor will do. Save yourself and us a lot of time and effort and read up some on the basic circuits the overall circuit is composed of.
The capacitor you ask about is part of an output filter, which also has it's balanced counterpart. It defines the low pass cut-off of the circuit, however it needs to have a specific relationship with other caps in the circuit to operate correctly. Mucking around with values blindly will just get you a mucked-up circuit that does not operate correctly.
Yes, changing the caps in the filters in VERY specific ratios can change the sound but the changes have to be fairly subtle. The filters are also part of a complex filter system which is partly implemented digitally ant dhis MUST be taken into account. Leave it alone unless you RELALLY know what you are doing, and since we are discussing the very basic basics of circuits such as series or parallel caps, you DON'T.
The critical capacitors in this circuit are the output caps, C4225, 4242, 4251, 4258, 4266, 4273. THese block the DC offset of the NJM5532 OPamps, which is fairly small, unlike the inputs along the top edge of the schematic. Regular polarized electros have problems with small or zero offsets (which is what 'audio grade' caps do much better but not as good as non-polarized ones!). These can in some cases be removed and a short fitted instead, if the actual offset is verified to be low enough. If not, these could be replaced with bipolar caps. After that the 10u/50V parts along the top edge of the schematic would be the next candidates for replacement by bipolars - these CAN'T be replaced by shorts, there is about 2.5 or 1.6V on each, this is the centerpoint bias of the DAC output driving the channels.
Regarding bipolar caps, two anti-series connected electrolytic caps are NOT a bipolar cap. This fallacy keeps being propagated over and over and I suggest whoever thinks so, make such a 'bipolar' cap and expose it to over about 3V of AC voltage. Hopefully the magic smoke escaping will finally bury this myth.
Why? Because in such a connection, one cap is always reverse connected. Most electrolytic caps (aluminium types) can withstand small reverse voltages (around up to 1.5V but this will eventually degrade the cap). SO what happens if the voltage across such a connection is more than twice 1.5V? Because the reverse polarized one will leak, the correctly polarized one will carry a higher voltage. This is all 'fine' if the AC component is less than about 2x1.5V - if not, both caps will be reverse polarized over 1.5V alternatively as the AC signal changes polarity and what you get is aVERY leaky and short lived capacitor. IF by chance you used tantalum caps instead, you would be lucky to even get to some 1V AC total as tantalums are NOT tolerant to reverse voltages.
Such 'substitution' for a bipolar cap works ONLY if the total voltage across the caps remain VERY small, perhaps in the 100mV range. In other words, to cater for slight DC offsets between circuit stages. For anything more, you need a REAL bipolar cap.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.