Mark
How much do you think the results of your tests with these 48 diodes is dependent on the electrical parameters (leakage inductance and stray capacitance?) of the transformer that precedes them? Presumably the optimum values for the snubber components are also dependent on this? If so, how does that help the DIYer who will be using different transformers to you?
Does your article specify the transformer you used for these tests?
How much do you think the results of your tests with these 48 diodes is dependent on the electrical parameters (leakage inductance and stray capacitance?) of the transformer that precedes them? Presumably the optimum values for the snubber components are also dependent on this? If so, how does that help the DIYer who will be using different transformers to you?
Does your article specify the transformer you used for these tests?
In this case, I still personally prefer RC snubbed rectifiers of any stripe, but I
bow to the customers who prefer fast/soft diodes.
Yes, the Linear Audio article found that, among the 48 rectifiers tested,
- The very best rectifier turned out to be a soft recovery diode
- The very best rectifier was 20X better than the very worst rectifier
- The very worst rectifier, plus snubber, was better than the best rectifier without snubber
_
Then how good was the best diode with a snubber ? Go to the jumpsuit and not need the belt ,suspenders and the elastic wasteband (pun) . Though my father wore one with a belt and waistband.
I predict that if you spend EUR 0,99 to download the article from Linear Audio's website, when you read it you will come to the conclusion:
how good was the best diode with a snubber ?
Sure does appear to be the very best of the best. It is after all a second order linear system. The belt helps the suspenders and vice versa. DIYers now know what to do.
Mark,
Is the problem before, at or after standard diodes? If a snubber makes everything nice, then there's a filtering issue here, right? Maybe your diodes are more ideal diodes, but not the whole answer?
I got the same euphoric sound using balanced AC power. A close second is listening between 1 and 3 AM when AC is cleaner. Listening at 1 AM is only second, well because I'm usually asleep at that time.
Plitron Torus and BPT come to mind. I made a balanced power unit good to 15 amps with a Toroid of Maryland +60/-60V AC transformer designed for this purpose. It was like upgrading an entire system. I couldn't shut up about it, but no one listened.
Interestingly enough, the 1994 issue of Audio Amateur when the Zen Amp first appeared had an article on snubbers. It was my first issue and still have it.
Good luck, I'm glad someone is getting to the bottom of this.
Thanks,
Vince
Is the problem before, at or after standard diodes? If a snubber makes everything nice, then there's a filtering issue here, right? Maybe your diodes are more ideal diodes, but not the whole answer?
I got the same euphoric sound using balanced AC power. A close second is listening between 1 and 3 AM when AC is cleaner. Listening at 1 AM is only second, well because I'm usually asleep at that time.
Plitron Torus and BPT come to mind. I made a balanced power unit good to 15 amps with a Toroid of Maryland +60/-60V AC transformer designed for this purpose. It was like upgrading an entire system. I couldn't shut up about it, but no one listened.
Interestingly enough, the 1994 issue of Audio Amateur when the Zen Amp first appeared had an article on snubbers. It was my first issue and still have it.
Good luck, I'm glad someone is getting to the bottom of this.
Thanks,
Vince
What Mr. Johnson proposes is AFTER the AC enters the transformer so I would think anything you do to improve your AC would still be just as important with the secondary snubbed.
He is quieting the rectifier's interaction with the secondary.
Balanced power should be a plus but toroids are not the way - as has been said the toroid welcomes in all of the noise on the AC line. The inherent losses in an IE transformer tend to block much of that noise. Something to consider ...
He is quieting the rectifier's interaction with the secondary.
Balanced power should be a plus but toroids are not the way - as has been said the toroid welcomes in all of the noise on the AC line. The inherent losses in an IE transformer tend to block much of that noise. Something to consider ...
Hello,
anybody seen this pdf?
http://www.shaoguang.com.cn/data/sgkj09010.pdf
A Fairchild Stealth 4A rectifier has less peaks than the Vishay Hexfred 8A type.
And the Stealth diodes have better price/performance
anybody seen this pdf?
http://www.shaoguang.com.cn/data/sgkj09010.pdf
A Fairchild Stealth 4A rectifier has less peaks than the Vishay Hexfred 8A type.
And the Stealth diodes have better price/performance
I paid for and downloaded the diode article. Also picked up Stan Curtis' article on balanced power.
Thanks,
Vince
Yes I've seen those scope photos. However they are applicable to >100 kHz switching mode power supplies, NOT fullwave bridge rectifier + power transformer linear power supplies that rectify at 2 x fmains (120 Hz). I.e. not standard diyAudio power supplies.
Those folks are measuring their diodes at dI/dt = 400 amps per microsecond (4E8 amps/second). I invite you to calculate the worst case biggest dI/dt possible in a rectifier diode driven at 2 x fmains by a power transformer in a piece of audio gear. Go ahead and assume it's a 100 watt class A amplifier with 100,000 microfarads of filter capacitance connected to the rectifier bridge output. What's the dI/dt? {prediction: your calculations will say it's orders of magnitude less than 4E8 amps/second}
What you really want to know is: which diodes perform best (and which perform worst) at the dI/dt that they will actually experience, in a piece of real audio equipment? The measurements in the Linear Audio article, are an attempt to quantify exactly that. One set of data for preamp-sized supplies & load currents, another set of data for power-amp-sized supplies & load currents.
Compared to SMPS @ 100kHz, the linear power supplies we put into audio gear are slow and scrawny. We have less Ipeak and less dI/dt, so diode switchoff generates less L*dI/dt stimulus into our RLC resonances. We also have a LOT more time to damp them thoroughly.
Those folks are measuring their diodes at dI/dt = 400 amps per microsecond (4E8 amps/second). I invite you to calculate the worst case biggest dI/dt possible in a rectifier diode driven at 2 x fmains by a power transformer in a piece of audio gear. Go ahead and assume it's a 100 watt class A amplifier with 100,000 microfarads of filter capacitance connected to the rectifier bridge output. What's the dI/dt? {prediction: your calculations will say it's orders of magnitude less than 4E8 amps/second}
What you really want to know is: which diodes perform best (and which perform worst) at the dI/dt that they will actually experience, in a piece of real audio equipment? The measurements in the Linear Audio article, are an attempt to quantify exactly that. One set of data for preamp-sized supplies & load currents, another set of data for power-amp-sized supplies & load currents.
Compared to SMPS @ 100kHz, the linear power supplies we put into audio gear are slow and scrawny. We have less Ipeak and less dI/dt, so diode switchoff generates less L*dI/dt stimulus into our RLC resonances. We also have a LOT more time to damp them thoroughly.
ISL9R30120G2
30A / 1200V / TO-247
This diode has a softness factor of 4.8 (!)
I don't think they made a mistake in the datasheet,
the slightly higher trr 269ns @ If 30A should be responsible for the softness, the freaks can begin testing out !!
I also noticed the STTH30S12 (30A / 1200V / TO-247),
has a softness factor of 2. But this diode has a higher Reverse Recovery Current
30A / 1200V / TO-247
This diode has a softness factor of 4.8 (!)
I don't think they made a mistake in the datasheet,
the slightly higher trr 269ns @ If 30A should be responsible for the softness, the freaks can begin testing out !!
I also noticed the STTH30S12 (30A / 1200V / TO-247),
has a softness factor of 2. But this diode has a higher Reverse Recovery Current
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.