|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Being new to the tube world, I have noticed that a lot of sources say that tube rectification is more musical (magical) that solid state rectification. Is this generally true despite the cost of the equipment, or would that mainly apply to less costly units?
Dear albee33,from my experience tube rectifiers are not superior to solid state rectifiers, if RF generation is avoided by small caps parallel to the diodes and if the diodes work on a small capacitor (aviod current peaks and core saturation of the tranformer) followed by a well dimensioned choke. Behind the choke big capacitors are ok. Tube rectifiers avoid saturation of the transformer due to their higher internal resistance.
Klaus
The primary circuit differences between my bottom of the line Audio Research VT100 MK II and their flagship amps, the Reference 300 and Reference 600s has to do with tube rectification. William Z. Johnson, no newcomer to the world of designing musical amps, seems to agree with the assertion that tubes work best.
Use both if you used a uf first then a tube second as a slow turn on device you will get both voltage and a reduced rf problem this set up was first shown to me in audiophile (Hong Kong ) magazine .triodethom
Albee33,Why in the world would you use anything but TUBE?
First tubes clear the RFI and other High Freq hash from the lines. Second they warm up the same way the tubes do, making the life of the tubes, caps and other components last longer. 3, they act like resistors when the amps are clipping giving the amp headroom. Whereas SS acts like a brick wall!
It's more expensive, but a must! Tube Rectification for ever!
Gordon
Why did Marantz use solid state rectification in 8B?
Hi everyone,To share in the PS design:
Thsee r my deductions from fiddling with Doc B's Forepaly pre-amp.
1) If u use SS rectification, hexfreds r a must. Even with a snubber, I think if you use a ultra-fast recovery diode, the harshness and tightness in bass improves tremendously. Snubbers provides tuned rejections of interference while the UF rec is more wideband thus effective.
2) I recomend if u can get hold of LM338 floating IC regulator, try this. This IC gives better and wider supply frequency response. Thus , it will be less harsh and hard as compared to lower current IC regs like LM350s or LM317s. Though the instrument separation is very good with this IC, the vocal is slight receeded.
3) A choke in the power supply is good in rejecting lower frequency hums but becareful of high frequency roll-offs. U can compensate this by putting a polyproperene at the B+ supply.
4) A tube rectifiy is good. So is a gas regulator. But the el caps must have good damping (or rather ESR for speed response). U can do this paralleling several small value elcaps to form the value that u desire. It is more crital to parallel at the cap nearest to the B+. Better high freq response (tighter bass too) can be achieved by paralleling polypropelene to this elcaps. Vocal presentation is nice but details with tube rec and regulators must be accomapined with good quality caps.
Overall, my preference would be hexfreds for rec with good caps at the B+. This is not secret but matching the right components with the type of circiut and thus sounds, wins the day.
Bye.
I think this question is akin to the generic question "Do tube amps sound better than SS amps?" There are great sounding ss amps (Jeff Rowland, Nelson Pass etc.), and there are terrible sounding tube amps (won't name any names). In terms of rectification, it all depends on the design, but there are several caveats;1. Silicon diodes in general generate a higher level of switching noise. This can be dealt with using caps at the output to sink the noise.
2. Silicon diodes start up very rapidly, but this can be dealt with by time relays or large value caps.
3. SS rectified PS tend to use very large value electrolytic reservoir caps, and these don't sound good unless one uses very expensive brands (Blackgates, Elna Cerefines etc)
4. The trashy caps can be bypassed with high quality, lower-value ones and therefore cutting cost.
5. SS PS is capable of very tight regulation, which is important in the HT rail (and the LT, for that matter).
6. Most commercial products use cheap silicon diodes. Try Hexfreds!However:
1. Vacuum tube rectifiers can be tightly regulated by an approproately designed regulator. No gas tubes please. For example, the 5651 voltage reference is capable of giving ss regulators a run for it's money in terms of tightness.
2. Many high-end products either use choke smoothing, or a pi filter. The quality of the choke is important, and cost is usually skimped there.
3. Commercial products tend to spec their components as close to the limit as possible to cut cost. This gives rise to "sag" and other undesirable effects. Core saturation at the choke is also commonly seen in these products.
4. If you go for the overkill approach with overrated transformer, rectifiers, chokes etc., you will have a decent tube PS.At the end, it all boils down to design and component quality. Both SS and tube PS can sound good. The Svetlana 6D22S is an excellent rectifier; you don't have to pay $80+ for a Mullard 5AR4 to get decent quality.
Hi.When different tube rectifiers give different sonic properties, then I must say that there is a difference between tube and SS rectification. Most manufactures use SS rectification for one of two reasons.
1) Tube rectification costs more.
2) High output power (50 Watts +) makes tube rectification dificult with current manufactured tubes.JMHO
Chris.
Hi,Here's something of interest regarding rectification.
Try this link, and scroll down to the bottom.
n
There are many posts on this topic in the archives.In fact, all tube equipment is best with tube rectification due to it's better tone, more flexible response, and it's more gentle handling of the start up which is critical to tube life.
I would venture as far as saying that the most costly equipment will benefit greatly from tube rectification as the tone and resolution is presumably higher. Solid state rectification, in my opinion, is just cheaper and easier for the manufacturer. Look at the real high end tube designers (not the big mass market names - the real designers). They always tube rectify.
horsepuckey
Impressive - you got everything you know on this into one word! You sure put a whole new light with your generous and illuminating post!
Sorry, my response seems to have been a knee jerk reaction to what I felt was an unfairly biased opinion of designers who use SS rectifiers in their circuits. I have found that there are compelling sonic reasons to use SS rectifiers, and the implication that guys who use them are just cheap bastards is unfair. Perhaps I can offer a further contribution on this topic based on my limited experience.There are a couple of issues that can be measured and heard which indicate to me that there are differences between the way the two types of rectifiers affect a circuit's sound.
The first issue I find important is that tube rectifiers tend to put out a smoother waveform than typical SS rectifiers. This is largely because of a VHF voltage spike, known as the reverse recovery spike, which occurs as the AC waveform entering the SS rectifer swings in the "reverse" direction, which causes the rectifier to cut off. Luckily this is measurable and has been studied for some time now, and SS rectifiers which minimize this effect, known generally as fast recovery, soft start rectifiers, have now been around for several years as a result. Interestingly, even these rectifiers still show the reverse recovery spike, but to a much less pronounced degree, and they definitely close in on the tube rectifier in terms of their effect on midrange and treble sonics of a circuit.
Luckily it is possible to find a fast recovery rectifier whose reverse recovery spike ( and other spurious VHF garbage)can be eliminated by the application of a simple filter between the rectifier and the HV secondary. John Camille showed us a few years ago that a UF 4007 used with a filter composed of 10 ohm wirewounds and .01 ceramic or VHF rated poly caps takes the spike out nicely. A good fast scope will be necessary to prove this, as while the spike has a pretty big amplitude, it's exceedingly narrow, residing in the 10mHz range and so it is a bit of a challenge to locate in the first place.
In this process we eliminate the shortcoming that has existed with the use of typical SS rectifiers in tube circuitry, an edgy sound caused by said spike riding on the audio signal and out thru the power transformer primary and in to our other gear. With its elimination the midrange and treble performance of tube vs. SS rectified circuits is pretty similar, with perhaps a small speed and resolution advantage to the filtered SS rectifier.
The second issue I find important is the effect of the two types of rectifiers on power supply impedance. A typical SS diode has a very low impedance, easily measurable at perhaps an ohm or two in the forward direction. A typical tube rectifier may have an impedance in the 100-200 ohm range, a figure easily located in tube manuals. Sonically, this affects a circuit such that a tube rectified circuit will have slower, softer bass, as it is not able to respond to quick current shifts as readily as the SS rectifier.
My limited experience indicates that a carefully implemented SS rectified circuit will outperform a tube rectified circuit in terms of a circuit's bass performance, and it will match it in the midrange and treble.
So for me it really has nothing to do with the cost of the components. It's all about getting what I perceive to be the most pleasant sound. Once again I apologize for my defensive reaction to your initial post. I welcome your sharing of your own implementations and measurements of the two types of devices in your own circuit designs, as a means to further the my own and the community's understanding of this subject.
Hi, This "bass sag" thing makes me a little curious. (Assuming class A) The current draw should be constant (mean that is) and only fluctuating from signal. A low bass 20-30Hz something is still fairly short time (0.05 seconds between "more" and "less" than average current). Do you find it a good idea to take this fluctuations all the way from the power transformer? Wouldn't it be better to store (or smooth rather) that amount of current "reserve" in a chokes or in caps? And if one stores it in caps or chokes, why should there be any "sag" over the rectifier tube? I will still see a constant current (or rather a fraction (say 1/10th) of the actual fluctuation in the output tube which may be just some 20mA: so maybee 2mA.)
/Pär
I agree that a brute force approach to filtration can certainly help create a low impedance node. I can't say for sure whether it will completely alleviate the effects of the impedance of the devices upstream.In my limited experience the most optimal setup has been when each element in the chain is as "fast" as possible. Certainly a lot of attention needs to be given to the PS filter. A shunt regulator might be an even better approach to creating a very low impedance node to feed the circuit from than a brute force flter.
But I have found that little things upstream of the big filter or regulator can still add up to affect the speed of the circuit. One of the more surprising things we've seen is improvements in speed and bass control by using large gauge solid core wire in power cords. This effect is audible even on circuits with very low current draw, and with PS filters with plenty of storage. And the SS rectifiers can give a similar effect, again, even in circuits with a relatively low current draw and beefy supplies.
I wish I could offer simple techniques for measuring the rather subtle dynamic phenomena we're discussing, but at this point we are still working up models for the behavior and I don't claim to have the larger picture of power supply dynamics all sorted. I will offer that one must probably consider the DC current loops and the AC current loops of the circuit separately as a starting point. And that in the case of the class SE amp, one should probably consider that the AC ground is way 'upstream' at the power transformer HV center tap.
Doc,I have to agree with you faster supplies make better bass. Many designers go for that brute force scenerio. I use modeling programs I wrote years ago to design my filters on the high voltage (only reason I keep that damn PC). I also use a home brew FFT analyzer to check out the effects of the supply. This can really show you what sort of impedance the amp demands on transients.
In general choke input supplies make better bass, the drawback is heavy ripple and therfore HUM. I have overcome that with some of the info supplied in my Sound Practices articles for the Baby-O and Bugle.
Gordon
Now is see your way of reason./Pär
Doc B., I'm not biased against anything, except the mediocrity of design I find in most heavily promoted 'big name' 'Stereophile Recommended' components I see today, both SS, tube, or whatever.Your comments are absolutely correct.
I read the initial post as wanting a very general, non-technical answer to go forward with investigating tube equipment. As a generalization for the relative newbies, it's not a bad one to look first at tube rectification. As a very gross generalization, I usually find that those designs have a 'wholeness' and cohesiveness that some others lack.
Bass will as you say usually be tighter and faster with good SS rectifiers. This is variable somewhat depending on the quality of the choke and the amount of capacitance in the power supply, as well as the quality of the output transformers and the actual choice of tube as the rectifier.
What isn't entirely measureable is the effect of the tube rectifier on tone, which is so often a lost concept in high end audio. My experience is that a tube rectifier can add a sense of openess, air and quality of tone that even the best SS rectification still can't quite match. I live with, and am not really bothered by, the slowness of the bass, which I actually find to be quite harmonious and pleasant. And many of the SS rectified circuits have a very hard startup because they do not use slow start rectifiers. In combination with a circuit running the output tubes at or even above their rated limits, that can get you a 'tube-eater'. There are more than a few of those out there.
You're right, it has nothing to do with cost of components. But, unfortunately, it does very often have to do with cost decisions in design and manufacture, which is legitimate also. But for me, as with you, it's about a pleasing sound.
Somewhere on the Manley website they discuss having used solid state rect. in one of their designs and switching to tube rect. by customer requests but themselves preferring the SS rect. for bass. Also the folks at Bel Canto have stated their preference for SS rects in their tube gear for similar reasons. None of my business here, really, just passing along these tidbits from my trivia-engorged memory banks, make of it what you will.
Zapper
Doc B.,Is this what is refer to as a snubber circuit? Do ypu still recommend the filter if Hexfred's are used? Does this provide the same benefit as bypassing the hexfred with a .01ufd cap?
Thank You.
The McIntosh tube amps, which are still highly-sought-after by pro audio folks and collectors, used solid-state rectifiers. They were neither tube-eaters nor were they known for a solid-state sound. There are PLENTY of tube amps with solid-state rectifiers that don't eat tubes, so perhaps Doc B's response, while rather brief, was on the money!
I didn't say that SS rectified amps were either tube-eaters or SS sounding! Nor did I say they were bad sounding amps! Some are, some aren't. But I DID say that a good design will sound better with tube rectification due to the better tonal qualities of rectifiers.As a matter of fact, right now I'm listening to a very nicely restored Marantz 8B borrowed while my amps are being worked on. It sounds very good, not quite to the level of a good modern design, but way better than most of the big name stuff out there. But my tube rectified amps still sound better.
Oh, sorry. I forgot to address part of your original post. Not all high-end designers use tube rectifiers. How about VTL and Sonic Frontiers, to name the first two that come to mind? Designers come in many flavors!
Who said Sonic Frontiers and VTL are high-end designers? I'm actually not too familiar with VTL, only that some tube dealers won't warranty tubes into their equipment.
SF from what I know actually farms out their design piecemeal, using their in-house design only for integration. There's probably 3 or 4 different outsourced people working on the project, none working together, from a concept/spec. Their in-house person puts the elements together. It's all designed to a very specific cost point. And it's also designed to feed the other side of their business, the parts distribution company which they actually started as! So - design with parts we distribute and get very cheap, design so that we have easy upgradeability ( sell a 'hall of fame' parts upgrade, which feeds the parts business again) - get the picture? By the way - that formula for SF didn't work all that well - they went bankrupt not quite 2 years ago and the assets were bought by Paradigm.But you're right, not all designers use tube rectifiers for various reasons. Their products, as with tube rectified products, range from good to not good. But I have found that the chances are much improved with tube rectification, because the designer seems to have focused on the tone and wholeness of the project, not price point.
I am amazed at the high level of expert advice and comment elicited by the simple question about tube rectifiers vs solid state. As a relative novice in the design and building of tube gear, I have a related question. I have built exactly one pair of amps so far. These are 807-based, using one pair 807s per channel connected in ultralinear mode to Acrosound TO-300 output transformers. 6SN7s are used as voltage amp (in SRPP config) and as phase-splitter/driver. For tube rectification, I chose a type 83 mercury vapor tube, because of its current capacity and the relatively low voltage drop associated with this type of rectifier. However, the MV types have the disadvantage that they need to be warmed up in advance of turning on B+, and they therefore do not "protect" the downstream tubes. Does anyone have any thoughts or opinions on the sonics afforded by MV tubes vs more conventional tube rectifiers? I want to build a pre-amp next. This will have less current demand and probably lower B+, so use of other tube rectifier types is possible. By the way, I am very happy with the amps in every way, except having to wait for those 83s to warm up.
Mercury tubes have something in common with SS rectifiers: they creates high freq. noise.
Besides you should actualy use sunglasses if the mecury tubes are visible...
(you can get a good tan out of them :-) )
/Pär
ime,the hf noise is not an issue, yes maybe running a ham rig with 872's, but even here the 'hash' is not an issue for audio circuits... the problems come when you look at transmitters and receivers but for 'normal audio' this is urban legend.
should you also use sunglasses for mercury vapor outdoor lighting to... again this is just a myth....
dave
Thanks for killing the myths.
They are still pretty much alive on rat.
/Pär
This is BS
Your 807 have warmed up by the time the 83 is ready to be turned on. So, where is your problem?- Dieter
This is how I turn my amps on from fully off: (i) Turn on AC into transformer with B+ turned off via a switch using the center-tap of the secondary of the B+ output of the transformer. Thus only my bias supply and all the filament supplies are ON. (ii) WAIT for the 83s to get warm. Actually, they get hot rather rapidly. Meantime the 807s are also warming up their filaments. (iii) Turn on B+ supply. I am told that if one turns on the 83 from dead cold and applies the B+ immediately, it's bye-bye to the rectifier. That's what's unique about using an MV rectifier; it cannot be turned on simultaneously with B+. Please tell me, if I am worrying unnecessarily, but I believe I am correct in this. Finally, it's no real problem; I like the amps very much, AND I could always insert a relay, if I get tired of having to separately flip two switches per monoblock.
With respect to some of the other responses, I also was told that 83s can be noisy. So far, that is absolutely no problemo. And that blue glow is very cool; I would never spoil it with sunglasses.
I haven't heard any noise in any AF amps yet. This is only a problem in the AM bands in radio receivers - see notes of consideration issued by RCA some 50 years ago ;-)I am using an Amperite 6-60 thermal relay in the CT of the power tranny. Works great.
- Dieter
****************************8
Excerpt issued by RCA on MV rectifiers...
Mercury-Vapor Rectifier Considerations
The MVR has very low resistance, so that the current it delivers depends on the resistance of the load and the regulation of the power transformer. Sufficient protective resisyance or reactance must always be used with this tube to limit its current to the recommended maximum value. If this value is exceeded, the tube voltage drop will increase rapidly and may permanently damage the filaments.
It is characteristic of MVRs that no appreciable plate current will flow until the plate voltage reaches a certain critical positive value. At this point the plate current rises steeply to a high value in a fraction of a second. This surge of current re-occuring each time either plate becomes conductive may exite circuits in the vincinity of the tube to damped oscillation and thus cause noisy radio receiver reception. It is usually necessary, therefore, to provide small radio-frequency chokes in series with each plate lead so that the slope of the current wave front to the filter is reduced sufficiently to eliminate impact exitation.
This post is made possible by the generous support of people like you and our sponsors: