Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
I knew I had the pin #'s all wrong. It's been a while since I built cables.
Your SE (unbalanced) explanation was perfect. The signal is referenced to ground. No wonder Ralphs gear sounds so good. SE is just plain wrong, in my opinion.
Balanced cabling was developed to combat issues related to transmissions over very long runs of cable (hundreds of feet to hundreds of miles).
The short runs of cable in your system may not show any measurable noise reduction from using a balanced run of cable.
The effort to utilize balanced lines may require vastly more components than otherwise necessary (even transformers in some cases), it may cause other impedance issues, as well as inducing more self noise in the actual circuit itself.
For example, if you take the signal off your phono cartridge and send it into the grid of a half of a 12AX7 with a CCS loading the plate, you will get more gain and less tube noise than running a differential long tailed pair. In my experience, this additional noise is more audible than noise picked up by the 3 feet of cable coming from your tonearm (unless you run it right on top of a power transformer).
There's no free lunch in circuit design.
Did you know that the input stage of both our amps and preamps is a differential cascode with CCS? The top portion of the circuit is arranged as a CCS.
In the preamp, what this does is give us low enough noise and enough gain that even with passive EQ, only two stages of gain are needed for the phono section.
From what I have seen, the top portion of the circuit is not arranged as a CCS (I don't consider a cascode configuration to be related to a CCS since the load seen by the bottom triode is very low).
Such topology is reminiscent of Allen Wright's old circuit cookbook.
Still, you're going to get more self noise from the cascode differential pair than you would from a single cascode stage. This doesn't matter all that much in a linestage or power amp, but will start to bite you in a phono preamp.
-We've been using that topology since the mid 1980s.
Actually we've not been able to get SE implementations to be as quiet. Mind you, they come pretty close. The key to getting differential amplifiers to really work they way they are supposed to is through an effective CCS- and FWIW, most of the CCS circuits I've seen really don't perform all that well, leaving a lot of performance in the actual amplifier on the table.
It's not "wrong"; it's just simpler, cheaper, and easier to build with fewer parts. Proponents of SE vs balanced argue that the additional parts needed to build balanced circuits can muck up the sound. I don't hear that. I do hear fewer problems with hum and noise that can plague single-ended circuits; that's enough alone to convince me. But I do have some single-ended gear in my second system, and it can sound excellent too.
I've developed a fully balanced dual cascade series 4-way crossover. Even though I am still in the testing stage, and only one test box has been build, the sound is out of this world.
I asked my question regarding RCA's because I had a fear.
This fear is that my crossover will only shine if the entire audio chain is balanced.
Once my speakers are finished, I will fashion an XLR to RCA connector.
Lew; when I plug my new cable into my MA-1 amp, will this now be a single ended signal. Or does Ralph somehow preserve the balanced nature of the waveform?
I'm not sure I understand the question. If the preamplifier is balanced (presumably an Atma-sphere product) and if the amplifiers downstream from the new balanced electronic crossover have a balanced input, why would you need to use an RCA interface or worry about SE connections anywhere in that chain?
Let's start with some basics;
There are manufacturers that produce single ended (RCA connection) gear.
There are manufacturers that continue to produce balanced (XLR) gear.
In theory the XLR gear should blow the RCA gear out of the water. It should be no contest.
Why is the difference so hard to hear?
I have developed a crossover which is balanced.
If I feed my speakers a balanced signal, they blow me out of the water.
I believe a balanced pre amplifiers, and a balanced amplifiers, should blow the competition out of the water. There should be no such thing as a SE High End pre or power. SE should be relegated to mid-fi.
I put forth that my balanced crossover will change the landscape of the high end speaker.
Until you all hear it, these are just keystrokes.
I need to know if my speakers will sound as fabulous with single ended gear, as it does with balanced gear.
I am hopeful a pair will be at AXPONA 2016.
I ask my question again: if I make a new interconnect, one with an RCA at the end, and connect it to my MA-1, will the 'balanced' nature of the audio waveform stay in tact. Or will it now be a single ended signal?.
"In theory the XLR gear should blow the RCA gear out of the water. It should be no contest." What theory says this? Worked in a recording studio for many years. For long mike cables we went balanced to diminish noise problems Once we got to the console everything was converted to single ended. Talking about a large Neve console. At audio frequencies and short cable lengths there is no advantage to balanced. More components, more to go wrong. Believe me a Shindo single ended pre is as good or better than any balance pre. Also the balanced stuff cost more money for no real improvement in performance
The noisier the environment, the more important it is to go balanced. In the home the presence of a digital source might be all the justification needed.
More components, more to go wrong. Believe me a Shindo single ended pre is as good or better than any balance pre. Also the balanced stuff cost more money for no real improvement in performance
I don't believe you. Our preamps are OTL preamps, in that they don't use an output transformer to drive a balanced line (nor do they use coupling caps). With regards to the signal path, it is simpler and lower distortion/wider bandwidth at the same time. There is only one stage of gain in the line section and two in the phono section.
The coupling cap at the output of traditional preamps, such as the Shindo, is a limiting device no matter how good the capacitor (the limit is imposed by stray capacitance and the inductance of the part, not the materials from which it is made; IOW its a mathematical thing for which there is no workaround by going with better materials).
In practice, we've had plenty of Shindo customers tell us our preamps sound better; being anecdotal that should not be taken as actual proof by any means. However, the fact of the coupling cap (and the additional complexity it represents to the signal) can. The result of direct coupling the output is improved performance which is instantly audible, particularly in the bass region, but additional transparency is also heard in the mids and highs. Its been our opinion for many years that it makes a bigger difference for the preamp to be OTL-OCL than it does for the amps! If you loose information upstream it does not matter how good your amps and speakers are- you won't recover the lost resolution.
This may explain why the MP-1 is our most popular product, often outselling all of our amplifier products put together.
It is too bad the late Ken Shindo isn't around to discuss theis with you. I always have problems with the OTL forum because it really is more the Atmasphere forum. So many ofRalph's answers to questions turn out to be a sales pitch for his products. Wonder how many Shindo owners have told Shindo that they like there preamps better than the Atma-Sphere preamps.
Being anecdotal that should not be taken as actual proof by any means. I have heard both. Your preamp is wonderful, I prefer My Shindo.
Yes, well, that's nice. Are there really bushels of folks who have owned both an MP1 and a Shindo preamplifier, at the same or at different times in their audio lives? I would think not, but perhaps I am mistaken. Nevertheless, I doubt that the late Ken Shindo could prove on the sheer basis of polemics that his preamplifier was superior to some other well designed preamplifier, balanced or otherwise. By saying this, I mean him no disrespect; such a debate rarely leads anywhere except back to the listening seat.
Like I said elsewhere in this thread, I am not fixed on a balanced topology for preamplifiers (or amplifiers) over all else, but I have heard a wide variety of preamplifiers in my home system, and I find I prefer the MP1 over any of them, still after 20 or so years of ownership of the same MP1, which I've updated myself in parallel with some of Ralph's ideas and some of my own crazy ones. Is the MP1 great because it is balanced or for other reasons? I don't know. Of course, my experience and preferences are colored by the fact that I am an ESL type of guy. OTLs (amplifiers of course) and ESLs mate particularly well, as you may know.
The only reason for you not to post here might be that none of Shindo's designs were OTL amplifiers, let alone OTL preamplifiers. There are several guys who do post here who prefer Futterman type (SE) OTLs or Transcendent OTLs. And now and then, Graaf. All commercial OTL amplifiers that are "not circlotron" in concept are derivative of the Futterman circuit. I think it would be great if such folks were to post more frequently. I myself owned nothing but Futterman amplifiers between about 1972 and the mid-1990s, and I loved them matched with KLH9s or M-L ESLs. You cannot blame Ralph for posting in order to answer questions posed on this site that are specifically about his products, and in the process, his particular bias always does come through. So does yours.
I also happen town a pair of David Berning ZH270 power amps driving Maggie 3.6 speakers. The Bernings use a large output capacitor instead of output transformers. I don't think I hear any downside to this design. They are spectacular amps. I realize that the planar forum has become basically a Maggie forum and although I own Maggies I wish there was a greater diversity of planar speakers discussed on that site.
Futterman amplifiers are capacitor-coupled. The Berning ZOtl amplifiers are not; they work by raising the audio signal so it "rides" on a frequency around 250 KHz, then the carrier frequency is filtered out at the output by use of a small inductor, rather than an output transformer. Because the carrier frequency is so high, only a small inductor is required which makes it possible to achieve very wide bandwidth and low distortion. I'm sure this is a gross over-simplification, but for sure David Berning is a brilliant guy. Pretty much nothing he designs is ordinary. Absent that small inductor, the Berning ZOtl amps are direct-coupled.
EDIT. I'm sure there's an electronic network involved too, in filtering out the carrier frequency, not just an inductor.
Of course you are correct on all counts. I was thinking of the single ended OTL by Transcendent sound that uses a big output capacitor. The Berning just has a small input capacitor to keep DC out. Direct coupled the rest of the way
"Of course you are correct on all counts. I was thinking of the single ended OTL by Transcendent sound that uses a big output capacitor. The Berning just has a small input capacitor to keep DC out. Direct coupled the rest of the way"
Looking at the ZOTL patent, it seems that the MOSFETS in the output stage, which are actually switching the high currents that pass through the loudspeaker, are DC isolated from the previous stage (which is vacuum tube), by the high-frequency transformer.
I think one way one can view the ZOTL is that what is essentially a switching-mode power supply is producing a high voltage output, which is then loaded by the "output" tubes. As the anode current drawn by the output tubes increases, so too the current draw in the low-voltage MOSFET switching circuit side of the "switching-mode power supply" must increase. The change in current on the low-voltage side is correspondingly much bigger than the change in current on the high voltage side. The high current on the low voltage side passes through the loudspeaker, and thus a large current change through the loudspeaker is produced as a result of a much smaller current change through the "output" tubes. In effect, the "switching-mode power supply" acts as an impedance matcher between a high impedance "output" tube and the low impedance speaker.
It is true that the impedance matching operates all the way down to zero frequency, so I suppose in a sense one could say it behaves as if the output side with the loudspeaker is DC coupled to the input side with the "output" tubes, even though there is a high-frequency transformer isolating one side from the other.
It's an ingenious design, and although I don't think it is fair to call it OTL, it certainly has a novel way to impedance match a tube "output" stage to a low--impedance speaker, without requiring the tubes to pass the high currents needed by the speaker. So it achieves the impedance matching without a conventional OPT.
What is not clear to me, though, is why such a MOSFET/tube hybrid amplifier is likely to be any better than a more conventional type of MOSFET/tube hybrid amplifier, in which vacuum tubes are used in order to drive a conventional MOSFET output stage.
People say it sounds very good, and I'm sure they're right. But probably conventional MOSFET/tube hybrids can sound very good too.
For your far superior explanation of the ZOtl circuit. When I wrote my own response, I was in no position to research the subject, was relying on fragmentary memory of an early review of the first such amplifier to come from DB; I think it was in Positive Feedback. This is not to say that I understand the circuit as well as you apparently do.
There are single ended (SET)amps. A single tube for amplification. 2A3, 45, 300B, 1605, etc.
There are parallel single ended amps. They use two tubes, paralleled, for amplification. Twice the power.
If this is true; a single ended OTL can use only one tube for amplification.
Question. Is the 2A3, 45, etc. a gain stage?
Could a 6SN7 be replaced in one of Ralphs amps by a 2A3? (I know there's wayyyyyy more to it then that, but in theory, could it be done?)
-Question. Is the 2A3, 45, etc. a gain stage?
- In conventional SET amplifiers 2A3 ,45 ,and similar output power triodes connected as anode follower gains both voltage and current gain , resulting in specific output power gain ,
however those SET amps usually needs at least one more voltage gain stage to work properly , since voltage gain only from those output power triode tubes is not enough to interface with low level signals coming from standard sources , as CD player , pre-amp ,PC card , etc ,....
-Could a 6SN7 be replaced in one of Ralphs amps by a 2A3? (I know there's wayyyyyy more to it then that, but in theory, could it be done?)
in some way all Atma-Sphere OTL power amps can be considered as two same SET-OTL power amp units joined together forming unique OTL Circlotron Push-Pull balanced power amplifier unit .
"Art which does not have the appearance of art is true art."
- Old Roman saying -
Your explanation was perfect.
When I read about SET OTL's, I scratch my head. By their very nature, A-S OTL's are a type of SET.
I remember once asking another manufacturer 'your SET amplifier has an XLR, but is it a true balanced circuit?'. 'No', was the response, 'it would be too expensive to build. It would need another amplifier built in, and another output transformer'.
The single ended OTL by transcendent sound is no longer made Each channel had a driver tube and 4 output tubes in Parallel. 1-1/2 watts per channel. You can do a search for single ended OTL and see a lot of info about this amp
David Berning makes an SE OTL, called the "Seigfried". I think it utilizes his ZOtl technology, but I am not sure.
I have the Atma-Sphere MA-1 set up. 140 watts.
I have the Tenor WP75 waiting for another chance. 75 watts.
If the Transcendent was 1.5 watts, with 4 output tubes, then SET's are most often low powered.
The obvious exceptions being the 211, 845, 1605 etc tubes.
I believe I'm starting to understand.
by paralleling more and more output tubes, but inevitable mismatching among the output tubes and additional requirements of power handling in the output transformer, are said to result in a deterioration and loss of SE "magic", such as it is. 100W would be about the practical limit, though. Those Kron amplifiers from Europe were an interesting attempt at high power SET. SE amplifiers are best mated with very high efficiency, and high input impedance, speakers, such that low power output is more than adequate. Real aficionados seem to feel that the lower the power output, the better the amplifier will sound.
Have you auditioned those Tenor amplifiers? I am curious to hear them some time.
You are right ! , since my experience with DIY SET amps are that 10W max is top optimum limit , SET with more output power than 10W cause SET magic to gone ,
and of course that those low power SET amps have to be very carefully matched with very high-efficiency LS for best sonic results ,
IMHO the best SET amps are those simple two stage based ,RC coupled ,
any way , except hyper liquid treble sound which come from typical SET amp I was never impressed or big fun of that sort of amps , there is so many things that miss for me , special at reproduction of lower bass range same as at highest sonic range ,
Best Regards !
General Warning , I have no intention to start OTL vs SET war with nobody !
"Art which does not have the appearance of art is true art."
- Old Roman saying -
Tenors claim to fame was 'Harmonic Structural Integrity'.
To that end they succeeded.
However, the Atma-Sphere's do most of the other stuff better.
Once my new speakers are complete, I'll try the Tenor's again.
My second hand information leads me to believe the Tenors used voltage regulation on the output stage, unlike most tube amplifiers, let alone Atma-sphere amplifiers. And they use a 12AX7 for voltage gain at the input, instead of a cascode a la Atma-sphere. And they are not dual-differential at the input, I don't think. But it's dangerous to assume from design or specs that one can therefore judge the amplifier. It may be excellent and does have its adherents.
You are probably correct.
I am not an electrical engineer. I know very little on the subject.
This is how I look at it;
A phono cartridge has two leads per channel. The red lead is the positive portion of the waveform, analogous to the speaker driver moving outwards. The white lead is the negative portion of the waveform, analogous to the speaker driver moving inwards.
A truly balanced signal preserves both sides of the waveform through the entire amplification process.
A single ended amplification chain uses the positive portion (red lead) of the waveform, but what happens to the negative portion of the waveform?. I believe half of the music is relegated to ground?
Or am I completely out to lunch?
Sorry, Billy, hope lunch was good.
The red and white sides of a channel are not the positive and negative portions of the waveform. They are the positive and negative portions of the current flow.
Any circuit has to provide a closed path for current to flow. Look at the one channel diagram: if the cartridge is our signal generator (the circle with the sine wave inside), and it has colored leads, in this case red and blue, it connects to the phono stage (here shown as a resistor) from both its red and blue sides. Yes, I know the load a phono stages presents to a cartridge is more complicated than a simple resistor, but let's leave that aside for now.
If it is generating a sine wave, during one half of the cycle current will flow clockwise around the diagram, and during the other half of the cycle current will flow counter-clockwise. But at all times that current is flowing, it flows through both the red and blue wires. There are no dead ends. Also, while current in one direction will push the woofer out, and current in the other direction will suck it in, current must flow through both the red and blue wires whenever current flows at all, and regardless of the direction of current flow.
For single-ended gear, the red and blue connect to the pin and shield of the RCA plug. Whenever current is flowing, both the pin and shield carry the signal, i.e. the current.
The same diagram can be used to show the phono stage (signal generator) and amplifier (resistor). Again, current flows through both the pin and shield of the RCAs. If using balanced gear, it flows through XLR pins 2 and 3 instead of the RCA pin and shield. It's like putting you, your turntable, and your electronics inside a big metal box that shields you from all the nastiness that can intrude from the outside world.
Because single-ended gear sends the signal through the shield, it is more susceptible to picking up various interference from the outside world, particularly when dealing with the tiny currents involved in phono cartridge outputs. But there are ways to design ICs from tonearm to phono stage that minimize this, such as a second outer shield. In my system, that second shield is connected at one end to the tonearm housing and turntable metal, and at the other end is connected to the preamp chassis ground.
For single-ended ICs, I connect the outer shield to the RCA shield at one end only, which lets the outer shield actually do some shielding from RF et al. The isolation it provides is not as robust as balanced gear, but I haven't yet had a home situation that could not be made completely quiet. (Hint: with the outer-shield-at-one-end configuration, where you place the shield "drains" can make a difference. In my case, they are all on the preamp end of the ICs.)
"A man need merely light the filaments of his receiving set and the world's greatest artists will perform for him." Alfred N. Goldsmith, RCA, 1922
I admit this can be a difficult concept to grasp, but please re-read what I wrote. It is not the case that "half the wave form" is lost in SE connections. This is not to say that I completely agree with Alan that there is nothing to be gained in home audio systems by using balanced connections. Perhaps Ralph can chime in with a more erudite discussion that will edify you.
Anyway, in your re-statement of a specific question regarding an MA1, it is possible to give a better answer than I did the first time. In Atma-sphere amplifiers that provide an RCA jack input as well as an XLR input, the user performs an act that connects the grid of one side of the dual-differential cascode input topology to ground,when he or she elects to use the RCA jack. The other grid of the input stage is driven by the input coming via the RCA jack. (I cannot recall whether connecting the unused grid to ground is done with an external switch or with a modified XLR plug, because it's been so long since I've even thought about driving the amp with an SE input.) However, the amplified signal still appears at the plates on both sides of the dual-differential cascode, which via cathode-followers drives the output stage still in balanced mode. So, I think it's correct to say that in an MA1, SE input is converted to balanced output. But it's the balanced output of what was originally an SE signal, so there was no noise cancellation (Common Mode Rejection) operating on the signal that reached the RCA jack. Some of that will be cancelled in the balanced circuit, I think.
Note. This response is very specific to Atma amplifiers. I daresay most other commercial amplifiers that offer both XLR and RCA inputs will handle the signal differently. That is to say that many products are "feaux balanced". For this reason also it is misleading to think that you can look for XLR input jacks and assume that the product contains a balanced circuit design inside.
I admit this can be a difficult concept to grasp, but please re-read what I wrote. It is not the case that "half the wave form" is lost in SE connections. This is not to say that I completely agree with Alan that there is nothing to be gained in home audio systems by using balanced connections. Perhaps Ralph can chime in with a more erudite discussion that will edify you
Its best if you think of an non-inverted signal and an inverted signal, rather than 'half the waveform' which only introduces confusion. IOW, both phases are being dealt with at the same time.
When you drive the MA-1 through the RCA input, the input is single-ended even if the source is balanced. Single-ended and balanced are inherently incompatible so you usually have to jump through some hoops to make the components work together- special adaptors, transformers and the like. But the thing you want to remember is that the signal is either balanced, or it isn't, no in-betweens.
Suffice it to say, Atma-Sphere for life.
Post a Followup:
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: