Well, he's relatively "new" to audiophile nutcase world.
He associates with his asian buddies that are a group of "wannabe" collectors, who all buy used "high end" stuff and go around visiting each other to critique each other's equipment.
Seems like a tight-knit group of faux-snobs with their own beliefs learned from the internet.
Okay. Sounds like a common type of audiophile situation where they don't really know what they're doing. If they're just doing for a fun hobby at least I hope they can afford it. Some people have even more expensive hobbies like boats, planes, etc. After so many hours of operation, required overhaul of the engines on a twin-prop turbo plane can cost several hundred thousand dollars. Is it cheaper than commercial flying? No. Can it fly as far or as fast as a commercial jet? Of course not! It can be great fun for some people who can afford it however.
Typical golden ear status is given to one by oneself. Ah, those hearing ability claims they make, some examples are shown below.
"I am very sensitive to distortion"
"I know I can hear smaller changes than .5dB."
A typical human can not "hear" distortions below a certain level compared to music.
A part of this is due to a type of music.
As for dB sensitivity, I doubt that normal people can detect a 0.5dB change in volume levels.
Again, depending on the music.
Because if those people are claiming to be "that sensitive", then high-end manufacturers like McIntosh would have never allowed a 0.25% distortion level in their equipment.
Which in this day and age, is terrible compared to say, my Technics at 0.008%.
Forum member PMA who used to run listening tests here once gave me an honorable mention for sorting unity gain noninverting audio opamp buffers in order of distortion by ear from recordings he made (opamps included TL071, 5532, etc.). The results were fully double blind. Distortion levels of the opamps were measured by PMA to verify proper operation.
Regarding detection of small dB changes, it can be relevant in mixing records. Sometimes .5dB out of balance is too much to a practiced ear. No surprise then that for listening tests it is often recommended to normalize DUT A/B comparisons to within .1dB. IME that is getting small enough to become unnoticeable.
Also on the historical record I have pointed out on numerous occasions I am NOT a golden ear. Any claims to the contrary are incorrect.
Much of this has been discussed before in various other threads. The person attempting to stir things up knows better.
Regarding detection of small dB changes, it can be relevant in mixing records. Sometimes .5dB out of balance is too much to a practiced ear. No surprise then that for listening tests it is often recommended to normalize DUT A/B comparisons to within .1dB. IME that is getting small enough to become unnoticeable.
Also on the historical record I have pointed out on numerous occasions I am NOT a golden ear. Any claims to the contrary are incorrect.
Much of this has been discussed before in various other threads. The person attempting to stir things up knows better.
I have to say that any type of measuring and determination of minute levels of both distortion or volume levels "by human ear" is at best iffy, due to differences in individual hearing abilities.
The best (and really the only) way to accurately determine such small amounts is by properly calibrated testing equipment - oscilloscopes, meters, etc.
The best (and really the only) way to accurately determine such small amounts is by properly calibrated testing equipment - oscilloscopes, meters, etc.
If you want a legit status, make sure to have the listening test proctored by a qualified person, test set up verified by qualified technician and documented. Anything else is equivalent to self appointment of golden ear status.
I consider published specifications to be for comparison purposes. Anything below 0.1% is academic. Plus it is questionable to me that many, if any, amplifiers can meet these specifications in everyday use, even if it never clips.
The testing methods have been long established and use continuous sine waves. Music is more complex (this shouldn't matter with a competent amplifier) and more dynamic. It is the dynamic behavior that isn't measured with an industry standard method (if at all).
I did some tests on consumer grade equipment. To be fair, most of it came from the dumpster and some of it was hotwired back to life. A typical modern (post 2000) amplifier rated at 50 watts will indeed deliver a 50 watt transient into its rated load. But for steady state operation, it's about 12 watts. The power supply will sag significantly when delivering a steady state sine wave. I postulate that modulating the power supply can introduce an audible effect, because I have observed that beefing up a power supply makes it sound more powerful and clearer, especially in the bass register. Maybe a sagging power supply varies the damping factor; I don't know. Perhaps someone could devise a test. I know I'm not the only one to observe this. Bigger caps = less sag = subjectively clearer bass. Same goes for swapping transformers for same voltage but higher VA unit.
So some of the vintage amplifiers that are still popular (like the Krell) in spite of measuring like dogs compared to even big box AV receivers today have one virtue they don't have: beefy power supplies. That's why I say ditch the moldy old electronics, spruce up the power supply, reuse the heatsinks, and put modern electronics in there (like a Blameless clone with parallel outputs). It will never run out of current! Easy peasy and the best of both worlds.
The testing methods have been long established and use continuous sine waves. Music is more complex (this shouldn't matter with a competent amplifier) and more dynamic. It is the dynamic behavior that isn't measured with an industry standard method (if at all).
I did some tests on consumer grade equipment. To be fair, most of it came from the dumpster and some of it was hotwired back to life. A typical modern (post 2000) amplifier rated at 50 watts will indeed deliver a 50 watt transient into its rated load. But for steady state operation, it's about 12 watts. The power supply will sag significantly when delivering a steady state sine wave. I postulate that modulating the power supply can introduce an audible effect, because I have observed that beefing up a power supply makes it sound more powerful and clearer, especially in the bass register. Maybe a sagging power supply varies the damping factor; I don't know. Perhaps someone could devise a test. I know I'm not the only one to observe this. Bigger caps = less sag = subjectively clearer bass. Same goes for swapping transformers for same voltage but higher VA unit.
So some of the vintage amplifiers that are still popular (like the Krell) in spite of measuring like dogs compared to even big box AV receivers today have one virtue they don't have: beefy power supplies. That's why I say ditch the moldy old electronics, spruce up the power supply, reuse the heatsinks, and put modern electronics in there (like a Blameless clone with parallel outputs). It will never run out of current! Easy peasy and the best of both worlds.
Distortion specs on power amps are iffy things. They are given at some power level. Test at lower power, distortion will be lower (well, except in rare cases of class AB crossover distortion). Spec at higher distortion and you can claim more power. So, there's a tradeoff done by the marketing staff.
Music doesn't play "at 0.xx% distortion", that doesn't mean much, if anything -- distortion is a measurement for a given power level with a simple fixed amplitude sinewave, or a couple of sine waves in the case of IMD. Music amplitude is all over the place. Not saying that's a harder condition, just that the term "% distortion:" doesn't apply to that situation. When you use your amp for music at on average WAY below clipping level, you aren't hearing music with the rated distortion in that case.
Music doesn't play "at 0.xx% distortion", that doesn't mean much, if anything -- distortion is a measurement for a given power level with a simple fixed amplitude sinewave, or a couple of sine waves in the case of IMD. Music amplitude is all over the place. Not saying that's a harder condition, just that the term "% distortion:" doesn't apply to that situation. When you use your amp for music at on average WAY below clipping level, you aren't hearing music with the rated distortion in that case.
I've got a nice, pristine 1972 era Harman Kardon 330A receiver here for free a whiile back.
The advertizing brochure gives it wattage ratings all over the place.
Now, factory standard, it uses a 2200uf main filter cap, and has a pair of 1000uf speaker caps.
To me, that's cutting things close.
So now it's been re-capped, etc, and I installed a 6800uf main filter, upped the other filter branches as well, but kept the 1000uf's the same value driving the speakers.
I also replaced the 2 main rectifiers with TO220 style Schottky diodes.
The thing sounds utterly marvelous now!.... like a big beefy model.
Also, like in the 330B version, I added a 12v zener to the tuner's B+ line to stabilize it, among other little tweaks.
Simple, easy, no need to do more to it.
It's mainly used with 8 ohm speakers.
Putting rediculously high caps in there isn't sensible, unless your're a "DIY Fanatic".
Even with 10000uF and 4R it starts to roll off at 39Hz...
With 1000uF/8R, it starts to roll off at 190Hz so either you boost the bass or you don't care for it/have a subwoofer?
Even a cheap Radio Shack stereo from the 1980's used 4700uF for the outputs IIRC
FYI the same amp I built with 10000uF output caps has a CRC filter in the power supply of 36000uF/1R/36000uF... It's single ended, and the PSRR is trash hence the large caps.
With 1000uF/8R, it starts to roll off at 190Hz so either you boost the bass or you don't care for it/have a subwoofer?
Even a cheap Radio Shack stereo from the 1980's used 4700uF for the outputs IIRC
FYI the same amp I built with 10000uF output caps has a CRC filter in the power supply of 36000uF/1R/36000uF... It's single ended, and the PSRR is trash hence the large caps.
It works! It's not too hard to do.
I made capacitor arrays to replace the antique caps in my old Pioneer. I mounted boards where the original caps were. Man did that clean it up! I added big 0.22 uF film caps to the board to bypass the power supply tight on top of the output transistors. The amplifier sounds great! It worked when I got it but it was weak and gritty. Now it's punchy and crystal clear! That's all I did to it.
I dunno if you're posting all that for the masses, but..Even with 10000uF and 4R it starts to roll off at 39Hz...
With 1000uF/8R, it starts to roll off at 190Hz so either you boost the bass or you don't care for it/have a subwoofer?
Even a cheap Radio Shack stereo from the 1980's used 4700uF for the outputs IIRC
FYI the same amp I built with 10000uF output caps has a CRC filter in the power supply of 36000uF/1R/36000uF... It's single ended, and the PSRR is trash hence the large caps.
I've been in professional servicing for decades, lord knows I've mentioned it plenty of times on here.
And I know all the capacitor time-constants and f/3's a long time ago.
Your 190hz calculation there is off though.
As for any 1980's era amp, I pretty much doubt that they used capacitor-coupling of the speakers, let alone 4.7Kuf's, never saw them in a Rat Shack unit, unless it was in the Power Supply.
I'd have to see a schematic to prove I'm wrong.
And I've serviced literally thousands of customer units over the decades.
Koda is off by a factor of two or a factor of ten, depending on how you define the start of the roll off. The pole frequency is 3.975 Hz. It “starts to roll off“ about 5X away (5 time constants to fully charge a cap). 10kuF and 4 ohms gives full output at 20Hz.
1970’s units may have had output capacitors. 4700 uF would have been about right - full power at 8 ohms at 20 Hz. All DC coupled outputs in the 80’s, except for a few pieces that were less than 20 watts per channel. Those had small output caps (and power supply filter caps), but we’re FAR from top of the line. I had a couple of them back in those days. Couldn’t afford their $300+ receivers.
The power supply filters are really supposed to be about the same (or larger) than the output cap value, for approximately the same reason. To keep the power supply impedance sufficiently “low” at the lowest frequency you‘re trying to reproduce.
1970’s units may have had output capacitors. 4700 uF would have been about right - full power at 8 ohms at 20 Hz. All DC coupled outputs in the 80’s, except for a few pieces that were less than 20 watts per channel. Those had small output caps (and power supply filter caps), but we’re FAR from top of the line. I had a couple of them back in those days. Couldn’t afford their $300+ receivers.
The power supply filters are really supposed to be about the same (or larger) than the output cap value, for approximately the same reason. To keep the power supply impedance sufficiently “low” at the lowest frequency you‘re trying to reproduce.
According to my calculator of choice: http://sim.okawa-denshi.jp/en/CRhikeisan.htm
The pole frequency is 39Hz, not 3.9Hz. 5x 39=195. I said about 190Hz. This plot looks more like ~120Hz.
What an I missing here?
The pole frequency is 39Hz, not 3.9Hz. 5x 39=195. I said about 190Hz. This plot looks more like ~120Hz.
What an I missing here?
Attachments
He's using 1000uF The sim shows 1000uF, not 10,000uF...
0.159/(0.001x4) = 39.75 no matter how many times I enter the numbers into the calculator
I'm the one using 10,000uF for outputs
The RC filter formula fc=1/2piRC is for the -3dB point, so with the values given (1000uF, 4Ohm...) it's ~39Hz where the output is down -3dB.
Also, woofers have at some f point a resonance (bassreflex may have two resonance peaks) where the impedance goes up considerably which will not load down the C, so if the resonance is at or below the fc, the output swings back up before it goes down but at much steeper rate.
2200uF would bring -3dB down to ~18Hz a good enough compromise which is below most average joe speakers lowest f resonance (edit: assuming 4Ohms..).
And another online calculator:
https://www.learningaboutelectronics.com/Articles/High-pass-filter-calculator.php#answer1
Also, woofers have at some f point a resonance (bassreflex may have two resonance peaks) where the impedance goes up considerably which will not load down the C, so if the resonance is at or below the fc, the output swings back up before it goes down but at much steeper rate.
2200uF would bring -3dB down to ~18Hz a good enough compromise which is below most average joe speakers lowest f resonance (edit: assuming 4Ohms..).
And another online calculator:
https://www.learningaboutelectronics.com/Articles/High-pass-filter-calculator.php#answer1
Last edited:
- Status
- Not open for further replies.