7N7
The last thing on my mind is to promote high prices on ancient rectifiers. Interestingly, these high prices are not the product of a large corporation marketing campaign, but rather the market-perceived value. I also would rather spend money elsewhere, but not without acknowledging the impossibility to emulate even an old DH rectifier by a soft-recovery diode and series resistance. In fact i would gladly give up the inconvenience of complex power up of mercury (and even the blue glow) if it were possible to achieve similar sonic result with something simple, cheap and unfussy.
Have you actually listened to various rectifiers in the same circuit or are your conclusions theoretical? I always respect personal preferences for sound even if very different to mine. If you prefer the sound (extra solidity, more punch) of SS rectifiers compared to vac that's fine but please don't tell us there is no difference.
regards
peter
The last thing on my mind is to promote high prices on ancient rectifiers. Interestingly, these high prices are not the product of a large corporation marketing campaign, but rather the market-perceived value. I also would rather spend money elsewhere, but not without acknowledging the impossibility to emulate even an old DH rectifier by a soft-recovery diode and series resistance. In fact i would gladly give up the inconvenience of complex power up of mercury (and even the blue glow) if it were possible to achieve similar sonic result with something simple, cheap and unfussy.
Have you actually listened to various rectifiers in the same circuit or are your conclusions theoretical? I always respect personal preferences for sound even if very different to mine. If you prefer the sound (extra solidity, more punch) of SS rectifiers compared to vac that's fine but please don't tell us there is no difference.
regards
peter
analog_sa said:
Peter,
Please re-read my original post. At no point did I write that there is no difference. In fact I alluded to differences: e.g. 5V of ripple on the cathode of a 5u4GB.
I have just, this morning, changed the output valves (to a different type) in an experimental amplifier I have. Apart from a smidgeon of gain I can hear no difference.
Now as I said before, I do not believe that there is a difference in sound with rectifiers, apart from the fact I am not wasting tens of volts and about 15W of power in a rectifying device with a high output resistance - together with the cost and space implications. I also conceded in my previous message that others may believe that they can hear a difference. Indeed there should be a difference since the power supply will be "soggier".
As for the price aspect, well, people pay enormous sums of money for 12AU7s!
I would love to run mercury rectifiers in the new amplifier I am building, but although I have plenty of mercury vapour rectifiers, I have neither the space nor the transformers to run them.
7N7
analog_sa said:
You're just lucky. I wish they sounded the same to me, it's obvious the choice i would make
Out of curiosity: what is the system in which you audition the rectifiers? Not that it really matters; they sound very different to me even in a low-res (but minimal) system.
regards
peter
No, these are output valves; apart from the first amplifier I built (used a 5R4GY) I have never used rectifier valves - always begrudging the voltage drop and current limits.
A bit of extra information: the output valves I changed this morning: I have an experimental mid-range amplifier, called "the Dashboard" because it carries lots of meters; originally it used triode-strapped YL1071/8116. I have only a few of these and they are a terrifying price. So yesterday I installed QQV07-50s and today substituted 829Bs in their place. YL1071 has mu of 7, QQV07-50 has mu of 8 and 829B has mu of nearly 10 at the flat-out operating point I am using - it is rather weedy compared with the others. As I say, the only difference i can detect is the gain - although I should add that YL1071, being rated at 60W Pa, should have been run harder, but the chokes I have are not up to it.
Another amplifier I have uses 13E1s triode-strapped; the original design called for 300VHT at 500mA. Again there is no way I would use vacuum rectifiers in this - imagine the trouble - and instead I can use a couple of BYW56s at about 20p each!
Chacun a son gout (I cannot do accents - sorry!).
Best wishes,
7N7
Hmmm, ......
Well , we all think different, but this is a hobby, so it doesn't really matter , does it?
Personally I allways use valve rectifiers , because I think they are cool,
and they may even sound better, but that's not really important, is it?
My little 813 amps have 4 3B28s in bridge configuration per channel,
that is hardly practical, and absolutely not logical in any way.
Cheers
Now playing: King Crimson
Tweek of tonight; Kilbeggan Irish
Well , we all think different, but this is a hobby, so it doesn't really matter , does it?
Personally I allways use valve rectifiers , because I think they are cool,
and they may even sound better, but that's not really important, is it?
My little 813 amps have 4 3B28s in bridge configuration per channel,
that is hardly practical, and absolutely not logical in any way.
Cheers
Now playing: King Crimson
Tweek of tonight; Kilbeggan Irish
vacuum rectification
The sonic artifacts of vacuum rectifiers are increased intermodulation distortion. Under load the vacuum rectifier is bad news.
Any noise or byproduct, of solid state rectification, is easily removed by the power supply filter. Cheap electrolytic capacitors can get rid of a switching transient.
The solid state rectifier really shiines on low frequency definition. Bass is much tighter with the solid state rectifiers.
Again, as far as switching noise is concerned, I can't see it with my oscilloscope. Keep in mind that the last amp, that I refurbished, used cheap electrolytic capacitors for power supply filters. Any rectifier problems would have been seen in the distortion measurements. The signal to noise ratio on some of these amps is in the 80 DB region. Any noise problem, created by solid state rectifiers, is easily removed by the filtering components.
The sonic artifacts of vacuum rectifiers are increased intermodulation distortion. Under load the vacuum rectifier is bad news.
Any noise or byproduct, of solid state rectification, is easily removed by the power supply filter. Cheap electrolytic capacitors can get rid of a switching transient.
The solid state rectifier really shiines on low frequency definition. Bass is much tighter with the solid state rectifiers.
Again, as far as switching noise is concerned, I can't see it with my oscilloscope. Keep in mind that the last amp, that I refurbished, used cheap electrolytic capacitors for power supply filters. Any rectifier problems would have been seen in the distortion measurements. The signal to noise ratio on some of these amps is in the 80 DB region. Any noise problem, created by solid state rectifiers, is easily removed by the filtering components.
Re: vacuum rectification
This can be the case but it's too sweeping a generalisation to make. How does this work in my fully <i>differential</i> Class A amps? The load current is very close to constant, the only 'error' signal being from the slight mismatch between the tubes. Mine are matched on a Sofia, and have CCS in the tails of most stages to hold them in balance. So your statement does not apply. Even a simple and cheap glow diode shunt reg, if properly designed and implemented can remove almost all of the PSU signal modulation in lots of SE stages.
Best not to generate the noise in the first place than to try and filter it out later. SS rectifiers have extremely sharp switching spikes (therefore very wideband noise) which are very hard to see on a CRO, so I'm not surprised. Use a decent spec an an they're easily visible.
Most HV elcaps have a considerable inductance, so the HF component of the switching noise sails right over them, barely attenuated if at all. PSU chokes are designed to perform at <150hz, so they're not wound low capacitance, so again the HF rubbish from the SS rectifiers rides straight over them.
Using an N&D meter or a CRO? No, it wouldn't have. Besides, I'm not at all impressed with an 80dB SNR.
valveluver said:
This can be the case but it's too sweeping a generalisation to make. How does this work in my fully <i>differential</i> Class A amps? The load current is very close to constant, the only 'error' signal being from the slight mismatch between the tubes. Mine are matched on a Sofia, and have CCS in the tails of most stages to hold them in balance. So your statement does not apply. Even a simple and cheap glow diode shunt reg, if properly designed and implemented can remove almost all of the PSU signal modulation in lots of SE stages.
Best not to generate the noise in the first place than to try and filter it out later. SS rectifiers have extremely sharp switching spikes (therefore very wideband noise) which are very hard to see on a CRO, so I'm not surprised. Use a decent spec an an they're easily visible.
Most HV elcaps have a considerable inductance, so the HF component of the switching noise sails right over them, barely attenuated if at all. PSU chokes are designed to perform at <150hz, so they're not wound low capacitance, so again the HF rubbish from the SS rectifiers rides straight over them.
Using an N&D meter or a CRO? No, it wouldn't have. Besides, I'm not at all impressed with an 80dB SNR.
...Spot on Brett!!!
All rectifiers generate a switching transient at switch off and this excites the tank circuits in the psu... This is well documented, as is the difference in switching spikes between ss and tube rectifiers. Also well established is the minimal effect that normal psu components have on the ringing that these spikes excite...
The degree to which this is audiable in any system is variable. And the ability of people to identify the effect of this ringing on the sound is even more variable...
"de gustibus non est disputandum.." and all that
For those that would like to investigate ss rectifiers and snubbers please see Jim Hagermans article:
PSU Snubbers
and Voltseconds articles:
psu ringing
damping ringing
rectifiers and Xfmrs
And, as always, read Lynn Olsens writings on rectifers and ringing. BTW he briefly describes how to set up an scope to show this ringing too..
My opinion is that tube damper diodes are best followed by MV rectifiers. But I also use the good old 1N4007 when I'm pushed for B+ volts or budget - but with snubbers and they must be adjusted on test in the final amp design...
ciao
James
All rectifiers generate a switching transient at switch off and this excites the tank circuits in the psu... This is well documented, as is the difference in switching spikes between ss and tube rectifiers. Also well established is the minimal effect that normal psu components have on the ringing that these spikes excite...
The degree to which this is audiable in any system is variable. And the ability of people to identify the effect of this ringing on the sound is even more variable...
"de gustibus non est disputandum.." and all that
For those that would like to investigate ss rectifiers and snubbers please see Jim Hagermans article:
PSU Snubbers
and Voltseconds articles:
psu ringing
damping ringing
rectifiers and Xfmrs
And, as always, read Lynn Olsens writings on rectifers and ringing. BTW he briefly describes how to set up an scope to show this ringing too..
My opinion is that tube damper diodes are best followed by MV rectifiers. But I also use the good old 1N4007 when I'm pushed for B+ volts or budget - but with snubbers and they must be adjusted on test in the final amp design...
ciao
James
James D. said:
Hmm,
well, there are many diode options - some probably better than those that I have suggested; I believe that fast, soft-recovery is the way forward since obviously there will be far less RFI emission - an important factor in so many modern designs.
As a matter of course, I always by-pass each diode with a 1nF ceramic disc cap - including the Schottkys that I use for DC heater supplies.
And a glance at the mains on a reasonable 'scope will show that there are far worse problems to deal with before the diodes!
I have always found diode switching noise to be very small compared with mains-borne racket.
7N7
Neat, this can work as it lowers the main frequency of the ringing and reduces it's overall amplitude - often to the point where the rest of the psu filters can remove it's effect.
Too true - but we assume that this has been dealt with already
It certainly should be dealt with first!
ciao
James
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.