So im looking at some options for a first tube amp build. I was looking for kits. A few that seem to have alot of support are the millet dcpp and tubelabs sse and spp. I know the SPP and dcpp are both push pull, vs single ended sse.
I guess where some of my confusion comes in is reliable information about the differences. The actual sonic differences in character.
My impression is pp has higher odd numbered harmonics, and the se has higher 2nd order harmonic distortion. Preliminary reading suggest 2nd harmonic is much harder to hear than higher orders like 3rd and 5th which im under the impression pp has more of?
So off the bat in a nut shell im trying to compare which of these designs is the theoretically cleaner design. And the sonic differences.
Next would be comparing them to a solid state amp. What exactly am i getting with a 250-350$ tube kit vs a typical decent solid state amp? Everyone on the internet seems to have a different opinion, without any numbers for distortion and other things backing it up so im trying to understand what is the actual difference or benefit of tube kits?
And lastly has to do with speaker choice. Im looking at 2 speaker kits. the SEOS high efficient horn kits from DIYsound group, and a MTM kit from vapor audio that is about 94db efficient but a 4ohm load. Can tubes handle a 4ohm load?
I appreciate the help!
I guess where some of my confusion comes in is reliable information about the differences. The actual sonic differences in character.
My impression is pp has higher odd numbered harmonics, and the se has higher 2nd order harmonic distortion. Preliminary reading suggest 2nd harmonic is much harder to hear than higher orders like 3rd and 5th which im under the impression pp has more of?
So off the bat in a nut shell im trying to compare which of these designs is the theoretically cleaner design. And the sonic differences.
Next would be comparing them to a solid state amp. What exactly am i getting with a 250-350$ tube kit vs a typical decent solid state amp? Everyone on the internet seems to have a different opinion, without any numbers for distortion and other things backing it up so im trying to understand what is the actual difference or benefit of tube kits?
And lastly has to do with speaker choice. Im looking at 2 speaker kits. the SEOS high efficient horn kits from DIYsound group, and a MTM kit from vapor audio that is about 94db efficient but a 4ohm load. Can tubes handle a 4ohm load?
I appreciate the help!
Hi. I'll try and address a few of your questions.
Regarding single-ended vs. push-pull:
You've pretty much got it in technical terms, so I'll give you the subjective.
Push-pull:
The bulk of the tube amps you've heard from commercial manufacturers are probably push-pull. They develop decent power - anywhere from 25 to 300W, although commonly about 50 - so you needn't have especially efficient speakers. The sound is closer to a (very) high-quality solid-state amp than a Single Ended Triode amp would be, but it's still got the warmth, depth, and effortless quality associated with tubes.
Construction is a little more involved because, aside from your input, driver, and output stages, you need a phase splitter. Still, in kit form, you shouldn't have too much trouble.
Push-pulls have oomph, extension, and detail - they lack a little of the immediacy that a SET will give you, but they can be tailored to drive pretty much anything through judicious application of negative feedback - (although some designs use none).
Single-ended:
Most single-ended amps are best married to high-efficiency (≥90dB/W/m) speakers, as they are typically "flea-powered" affairs (2-10 watts is typical), although they generate a good deal more volume than their power would suggest because of their distortion spectrum. The beauty of the single-ended amp lies in its midrange; they don't do bass (in general) as well as p-p designs, nor do they have the high-end extension of a well-designed p-p amp. Their midrange, however, is quite simply, "to die for." Voices gain an intimacy and realism that you'd be hard-pressed to match with any other amp topology. They are also, by nature, the simplest amplifier circuits in existence - input stage, driver stage, and output - (and sometimes, two sections of the same tube are used for input and driver). Most of them famously use NO negative feedback anywhere in their design as well.
My personal predisposition is toward sensitive speakers - they just reveal microdynamics and subtle details so readily. The presentation is truly effortless - ( as the speaker requires so little power that every tiny electrical impulse causes sound to be heard) - hence great retrieval of detail. The problem is, because they are so sensitive/revealing, they can really show up shortcomings elsewhere in one's system.
SET's are the obvious choice for this type of speaker, and the single most important component in the amp will be the output transformer, as it's the "bottleneck" (relative to all the other circuitry) which limits frequency response and contributes most of what little odd-order distortion there is. Nonetheless, any half-decent OPT will at least give you a window into the magic of SE amps. Once hooked, you can always upgrade the output trafo (or build a better amp later).
Tubes vs. Solid State:
Well, here we come to the crux of "the great debate" - what is the advantage of a tube amp? (or a solid state amp)? Tubes are inherently more linear (for the most part) than transistors, but they are voltage devices that don't handle current all that well. Transistors are current devices that don't handle voltage all that well. It takes current to drive speakers, but we hear in a manor which most mimics voltage. So which is better? For bass, transistors - nothing controls a woofer like a bunch of current at a real low impedance. For midrange, tubes - they just sound more "real" compared to all (but the very best) solid state designs. Highs, you can go either way; transistors are faster, and can give you greater high end extension, but tubes exhibit less "overshoot", which is really noticeable at high frequencies. Transistors can be hooked right to speakers, so that part is easy, but keeping them in their "linear region" (between being switched on and switched off) is a difficult balancing act. Most solid state amps use negative feedback to correct non-linearities in the signal; (negative feedback reduces distortion, but it also kills dynamics - almost any aspect of audio design involves compromising between two evils). Tubes love to operate in a linear region, but generally have to be coupled to the speaker through an "output transformer" to turn their high voltage into high enough current to drive the speaker. That adds cost - (often more than the tubes themselves) - and weight, and the transformer will never have the frequency response of the tube(s) driving it, but if you love the sound of tube amplification (and I do!), it's worth it.
BTW, 4Ω is not a problem; most tube amplifiers' output transformers have 4Ω and 8Ω (and often 16Ω) outputs. So, unlike solid state where impedance alters the power curve, a tube amp will put out the same rated wattage (at roughly the same distortion) regardless of your speakers' impedance, as long as the OPT's tapped to match that impedance).
Regarding single-ended vs. push-pull:
You've pretty much got it in technical terms, so I'll give you the subjective.
Push-pull:
The bulk of the tube amps you've heard from commercial manufacturers are probably push-pull. They develop decent power - anywhere from 25 to 300W, although commonly about 50 - so you needn't have especially efficient speakers. The sound is closer to a (very) high-quality solid-state amp than a Single Ended Triode amp would be, but it's still got the warmth, depth, and effortless quality associated with tubes.
Construction is a little more involved because, aside from your input, driver, and output stages, you need a phase splitter. Still, in kit form, you shouldn't have too much trouble.
Push-pulls have oomph, extension, and detail - they lack a little of the immediacy that a SET will give you, but they can be tailored to drive pretty much anything through judicious application of negative feedback - (although some designs use none).
Single-ended:
Most single-ended amps are best married to high-efficiency (≥90dB/W/m) speakers, as they are typically "flea-powered" affairs (2-10 watts is typical), although they generate a good deal more volume than their power would suggest because of their distortion spectrum. The beauty of the single-ended amp lies in its midrange; they don't do bass (in general) as well as p-p designs, nor do they have the high-end extension of a well-designed p-p amp. Their midrange, however, is quite simply, "to die for." Voices gain an intimacy and realism that you'd be hard-pressed to match with any other amp topology. They are also, by nature, the simplest amplifier circuits in existence - input stage, driver stage, and output - (and sometimes, two sections of the same tube are used for input and driver). Most of them famously use NO negative feedback anywhere in their design as well.
My personal predisposition is toward sensitive speakers - they just reveal microdynamics and subtle details so readily. The presentation is truly effortless - ( as the speaker requires so little power that every tiny electrical impulse causes sound to be heard) - hence great retrieval of detail. The problem is, because they are so sensitive/revealing, they can really show up shortcomings elsewhere in one's system.
SET's are the obvious choice for this type of speaker, and the single most important component in the amp will be the output transformer, as it's the "bottleneck" (relative to all the other circuitry) which limits frequency response and contributes most of what little odd-order distortion there is. Nonetheless, any half-decent OPT will at least give you a window into the magic of SE amps. Once hooked, you can always upgrade the output trafo (or build a better amp later).
Tubes vs. Solid State:
Well, here we come to the crux of "the great debate" - what is the advantage of a tube amp? (or a solid state amp)? Tubes are inherently more linear (for the most part) than transistors, but they are voltage devices that don't handle current all that well. Transistors are current devices that don't handle voltage all that well. It takes current to drive speakers, but we hear in a manor which most mimics voltage. So which is better? For bass, transistors - nothing controls a woofer like a bunch of current at a real low impedance. For midrange, tubes - they just sound more "real" compared to all (but the very best) solid state designs. Highs, you can go either way; transistors are faster, and can give you greater high end extension, but tubes exhibit less "overshoot", which is really noticeable at high frequencies. Transistors can be hooked right to speakers, so that part is easy, but keeping them in their "linear region" (between being switched on and switched off) is a difficult balancing act. Most solid state amps use negative feedback to correct non-linearities in the signal; (negative feedback reduces distortion, but it also kills dynamics - almost any aspect of audio design involves compromising between two evils). Tubes love to operate in a linear region, but generally have to be coupled to the speaker through an "output transformer" to turn their high voltage into high enough current to drive the speaker. That adds cost - (often more than the tubes themselves) - and weight, and the transformer will never have the frequency response of the tube(s) driving it, but if you love the sound of tube amplification (and I do!), it's worth it.
BTW, 4Ω is not a problem; most tube amplifiers' output transformers have 4Ω and 8Ω (and often 16Ω) outputs. So, unlike solid state where impedance alters the power curve, a tube amp will put out the same rated wattage (at roughly the same distortion) regardless of your speakers' impedance, as long as the OPT's tapped to match that impedance).
Last edited:
Great reply. Thats helps a lot!
So i hear that SET have this great midrange. Perhaps my questions should be what causes this? Why is it better? And what kind of roll off are we talking about for the highs and lows?
One thing that would be disappointing would be to have a high end speaker using something like raal ribbon transducers, noted for their incredible high end spectrum, and have it rolled off. Anything I drive also would need to exhibit nice control into the 40-50hz region as well.
So i hear that SET have this great midrange. Perhaps my questions should be what causes this? Why is it better? And what kind of roll off are we talking about for the highs and lows?
One thing that would be disappointing would be to have a high end speaker using something like raal ribbon transducers, noted for their incredible high end spectrum, and have it rolled off. Anything I drive also would need to exhibit nice control into the 40-50hz region as well.
The decision also needs to consider your listening style and preference. If you're into rock concert loud with chest thumping bass then you want to go with push pull tube or solid state depending on the detail desired, tube will bring out more detail and warmth. if you listen at reasonable volumes and are looking for incredible detail then go with single ended tube. I'm listening to a 3/4 Watt SE tube amp right now at low to medium volume and it's incredibly detailed yet relaxing to listen to. Solid State tends to be fatiguing after a little while.
What type of music / source do you normally listen to?
Paul
What type of music / source do you normally listen to?
Paul
Music is usually listened to at moderate volumes but I would want the ability to get fairly loud on occasion. Not run you out of the house loud but pretty good. I guess this is more speaker dependent as well. Im not sure at what decibel people classify it as rocking loud.
Music choice is pretty much a blend of all modern music. Not much classical or jazz. lots of rock, hip hop, country.
I do listen to a lot of instrumental though. Acoustic and electric guitar instrumental. And as a musician i certainly like micro detail and fantastic vocals.
The idea of high end detail being rolled off, and slower less controlled bass bass is a bit frightening from single ended. But then people say its more detailed and glorious midrange? Hard to picture almost seems like an oxymoron to be more "detailed" yet have less dynamics and rolled off extended range.
Music choice is pretty much a blend of all modern music. Not much classical or jazz. lots of rock, hip hop, country.
I do listen to a lot of instrumental though. Acoustic and electric guitar instrumental. And as a musician i certainly like micro detail and fantastic vocals.
The idea of high end detail being rolled off, and slower less controlled bass bass is a bit frightening from single ended. But then people say its more detailed and glorious midrange? Hard to picture almost seems like an oxymoron to be more "detailed" yet have less dynamics and rolled off extended range.
Can I suggest maybe contacting an audio club in your area and having a listen to a few types of amps and speakers? Could save some disappointment down the road...
I have six amps, pp, se and solid state. My SSE is by FAR far the most enjoyable to listen to. If I could only keep one, that would be it.
Good luck in your search!
I have six amps, pp, se and solid state. My SSE is by FAR far the most enjoyable to listen to. If I could only keep one, that would be it.
Good luck in your search!
Probably a good idea. I'll have a look around. Would be nice if there were any high end shops.
If anyone knows any good places in Sacramento California let me know 😀
If anyone knows any good places in Sacramento California let me know 😀
I just finished a tubelab sse build and have had the opportunity to compare it to my Rotel/Adcom pre/power combo.
While the low power single ended amp lacks the impact of the far more powerful solid state combo, I can honestly say that the natural delivery it provides is, simply put, far more musical. There is certainly something to be said for the effortless smoothness a well implemented tube amp provides (music genre willing). With 90db speakers, I have plenty of volume with the sse and can't imagine myself going back to the (until now respectable) solid state separates.
Dollar for dollar, these two approaches are at the same level. One could probably best the rotel kit in solid state DIY builds though. I imagine a pass labs solid state would be near the cost and a big improvement.
In the end, I suppose that if this is your first build and you're coming from commercial products, you're hardly likely to be let down with any of the proven tube or solid state documented on this forum. If my DIY itch is any indication, you'll probably have a hard time stopping at just one anyways!
Edit: with regards to frequency response, I can't say I've been let down. I'm using speakers based on fostex fe103s. The highs are not lacking (and I have young ears). If I do feel the need for bass, I will provide it with an active subwoofer, as would most people.
While the low power single ended amp lacks the impact of the far more powerful solid state combo, I can honestly say that the natural delivery it provides is, simply put, far more musical. There is certainly something to be said for the effortless smoothness a well implemented tube amp provides (music genre willing). With 90db speakers, I have plenty of volume with the sse and can't imagine myself going back to the (until now respectable) solid state separates.
Dollar for dollar, these two approaches are at the same level. One could probably best the rotel kit in solid state DIY builds though. I imagine a pass labs solid state would be near the cost and a big improvement.
In the end, I suppose that if this is your first build and you're coming from commercial products, you're hardly likely to be let down with any of the proven tube or solid state documented on this forum. If my DIY itch is any indication, you'll probably have a hard time stopping at just one anyways!
Edit: with regards to frequency response, I can't say I've been let down. I'm using speakers based on fostex fe103s. The highs are not lacking (and I have young ears). If I do feel the need for bass, I will provide it with an active subwoofer, as would most people.
Last edited:
Thanks. Hope it cleared things up a little for you.
Two things: the distortion spectrum (predominantly even order) of single-ended triode amplifiers, and the fact that they are, by nature, Class A amplifiers. Most (though not all) solid state and push-pull amplifiers are Class A/B by nature, meaning that at least two devices amplify the signal in turns, one handling the positive-going half, the other handling the negative-going half. This creates a small "glitch" at the 0 point known as crossover distortion, which is a small (but noticeable) 3rd order distortion created by a little turn-on "spike" to either side of the transition.
There's no crossover distortion, and what distortion the tube itself produces is mostly 2nd order – exactly an octave away from the note(s) being produced.
That depends almost entirely on the quality of the output transformer. Typically, you'll get about 30Hz to 20KHz within a 1 dB and a few Hz lower to a few KHz higher within 3.
It's po$$ible to buy a single-ended output transformer that's flat from below 20Hz to above 35KHz, but you'll need a second mortgage to afford it.
I don't think you'd experience issues with either. The more revealing the tweeter is, the more you'll appreciate the high-end smoothness and neutrality of a SET. The beauty of having tweeter response to, say, 28KHz is not so much that you'll hear it as it is that the tweeter will be that much more effortless at 20K, and will exhibit excellent transient and phase response even at 3K. And, as far as bass control goes, any decent SET will have good control at 40-50 Hz; it's down at 20 or 30 where you'd need to worry.
Last edited:
Actually, given the speakers he's considering, volume shouldn't be an issue, even with a SET (nor would the 40-50Hz bass he's asked about, which is hardly "chest thumping").
However, should he want the ability to drive "conventional" (i.e. low-sensitivity mass hi-fi) speakers, that would be a consideration.
Last edited:
Never said that the high end was rolled off, just that it doesn't extend into the ultrasonic region to the extent that the best SS (or push-pull) designs do.
Sorry; wrong again. SET's are not "more detailed", they are more accurately detailed. It's not that they have "microscopic" definition, but that they define everything as if it were right there in front of you – no more, no less. After listening to a well-implemented SET for a while, one can recognize, much more readily, an amp which has artificially exaggerated detail - (these amps are usually built for less sensitive speakers which don't react as quickly to micro-detail, so they have to).
Also, SETs do not have "less dynamics"; in fact, if anything, they are more dynamic than their SS or P-P counterparts. (Remember the little treatise re:negative feedback in my first post?)
Hope that clears things up a little.
Last edited:
You guys are great, this is all really helpful
I found this an interesting read. Damping factor
Going by that, it would indicate the control of bass via damping factor is less a problem vs the non flat freq response. Basically bass gets louder. Possibly the sourcing of the boomy loose bass?
I'm leaning towards the sse at this point. But I know loose boomy bass is not something I like. I remember when I got my first nice amp in car audio. The radical change in bass attack, decay, impact was night and day, and had nothing to do with volume.
I found this an interesting read. Damping factor
Going by that, it would indicate the control of bass via damping factor is less a problem vs the non flat freq response. Basically bass gets louder. Possibly the sourcing of the boomy loose bass?
I'm leaning towards the sse at this point. But I know loose boomy bass is not something I like. I remember when I got my first nice amp in car audio. The radical change in bass attack, decay, impact was night and day, and had nothing to do with volume.
In a SET amp, your damping factor is also determined by the output transformer. The greater the ratio of the transformer input impedance (to the tube's plate impedance), the higher the DF. One method to "cheat" a higher damping factor out of a SET (at the expense of some output power) is to use the 4Ω tap on an 8Ω speaker, effectively doubling your DF.
That being said, if you like a super tight bass, you're never going to beat solid state. Nonetheless, a SET will give you decent damping down to about 40 or 50Hz in most cases. I wouldn't say they exaggerate bass, as the output transformer (typically) works at an inverse to the speaker's impedance "down there", but they are a warm (as opposed to "slamming") bass. (Also, as I said before, down below 30Hz, a SET can be a little weak).
That being said, if you like a super tight bass, you're never going to beat solid state. Nonetheless, a SET will give you decent damping down to about 40 or 50Hz in most cases. I wouldn't say they exaggerate bass, as the output transformer (typically) works at an inverse to the speaker's impedance "down there", but they are a warm (as opposed to "slamming") bass. (Also, as I said before, down below 30Hz, a SET can be a little weak).
Last edited:
I like SE-UL (single ended ultra linear) running KT88/6L6 tube types with either low efficiency (88dB or so) or high efficiency (+100dB horn) speakers both sound good to me with plenty of bass to. That said it's all in the output iron. Don't skimp there. I like Edcor CXSE 25W. plenty of power to drive whatever you want to thro at them speaker wise with good prices. I've tried many speakers even 2Ohm Infinity QLS-IIIs (with ribbon tweets) and they sound great as does my horn loaded EVs (high efficiency). No lack of bass with either end of the spectrum. I've got some decent PP tubage knocked to the shelf after foray into SE. Not night and day diff but enough, that's all subjective tho. Some love SE some hate it. Either PP or SE are great given a good design and good iron. With SE you need a VERY good power supply with low ripple to make it work, PP requires a bit more tweaking to get to the sweet spot but worth the effort. For a beginner build SE is hard to beat and pretty easy as long as you understand your limitations with regard to safety which is very important. 500V caps can get you good if you aren't paying attention. My experience with safety and beginner build, just a matter of time till you get a jolt, most times it's better to walk away when you get tired or especially, distracted. Whatever you decide to build remember safety 1st and there's no 2nd. DMM with clip leads not probes. You'll be testing power supply before you have a loaded circuit ~500V, that's nothing to take lightly. One hand only in the box once you have power the 1st time.
Cheers,
Bob
Cheers,
Bob
+1
Pretty much what we've all been sayin'.
Yeah, Edcor and Heyboer are both great "starter" trafo's – especially for the money. Hammonds have slightly better bass, but you probably won't notice a difference until you get up into the ($200+) ElectraPrint area, which is where you might start your "upgrade path", somewhere down the road.
(From there, you could move into some Asian iron like Tango or Silk, then – depending on how badly bitten by the bug you get – you might go so far as to investigate AudioNote, or Monolith Magnetics – their "Summit" series is the best I've heard personally, but I haven't yet experienced an AudioNote firsthand).
Don't worry about that now, though – there'll be plenty of time for upgrading (if you so choose) later.
Bear in mind, though, that every time you do, it's like getting a whole new amplifier – so profound is the effect of the OPT on overall sound in a SET (or in Bob's case, a Single-Ended Pentode).
Good advice.
Yes, make sure you keep one hand in your pocket (and work with only one hand) at all times when working "live".
I have two SSE's, two TSE's, three SPP's and two and a half DCPP's as well as several other one-of-a-kind amps. I might be a bit biased, but I like them all...for different reasons.
I have a system in the living room that uses a pair of 15 inch Silver Iris OB coaxial drivers mounted in 1941 vintage Zenith console radios to achieve acceptable WAF. These speakers are 96db efficient and go down into the low 40Hz region. They can eat 100+ watts too. There is now a flat screen TV in the wall unit where the CRT TV was so the space behind the TV is available for just about any amplifier regardless of ugliness, it's not visible. I have swapped dozens of amps through this system. The usual amp here is an SSE with the big Edcor CXSE output transformers. This combination will shake the whole house with KT88's in UL or deliver the sweet sounds of a female vocalist with EL34's in triode. When it's time for a rock concert in the living room, or rattling the neighboorhood with tecno or EDM, there is only one valid choice, my hotrodded version of Pete's DCPP that cranks out 125 WPC. I can open the house windows, crank this thing to 11 and go outside to mow the lawn.....without hearing the lawnmower. I also like the TSE with 307A tubes in it on these speakers.
I have another system (actually two) in my work room. This is a small room (10 X 11 feet) that is my lab, computer room, music production studio, and primary listening location. Heat is a big factor here. The house AC unit AND a small wall unit can not keep this room cool with a lot of stuff turned on in the south Florida summer. The go to amp here has been the SSE with smaller Transcendar OPT's through Yamaha NS-10 Studio monitors. These speakers don't go below 75 Hz so I use a 10 inch powered subwoofer. This has been the case for 6 years. The Yamahas are only 87db so they don't go terribly loud with 7 watts, and neither does the wimpy sub, but it is plenty in the small room. I often swap in an SPP here. It has a different sound, which I will try to explain later. The DCPP has a similar sound, and goes far louder, but generates more heat, so it doesn't see much use. I have a 40 WPC SE amp, but I don't use it because the room hits 85 degrees in about an hour when it's on. I am in the process of rebuilding the studio yet again, so the amp line-up will change, but an SSE will always be a part of it....so will some sort of P-P amp. A new pair of FH-3 horn speakers is being added.
The sound.....
Both of these statements are mostly true, but not the only factors. I don't like flabby bass.
Most speakers require a good DF for solid bass. The DF requirement depends greatly on the woofer design. Ported cabinets generally will go lower than the same woofer in a sealed box, but the DF requirements are greater. The drivers in my OB cabinets are designed for open baffle systems and do not need a huge DF.
A SE tube amp generally does not use much feedback if any. This guarantees a low DF, how low.....1. OK 1 is unusual, but not uncommon. Usual is between 3 and maybe 10.
The impedance of the OPT is the biggest factor here. As you increase the impedance the tube sees, the DF goes up, and the distortion goes down, but so does the output power. The DF can be increased, and distortion decreased, by increasing the tube current, at the expense of tube life. Bigger tubes tend toward higher DF over smaller tubes in the same amp (KT88 VS EL34 VS 6L6GC). These are all compromises made during amp design or build time.
This is why I recommend a 5 K OPT over the usual 3.4K OPT in the SSE.
I recover the lost power by increasing the B+. The SSE runs around 440 volts. The SSE can run triode mode which has the highest DF, lowest distortion, but lowest power. It can also run UL for nearly double the power with a slightly higher distortion, and loss of some DF. A few users have tried pentode mode, but the DF IS near one which doesn't work on most speakers.
I am an electrical engineer by trade, so it is my nature to over analyze things to figure out what is going on. The explanation that follows is somewhat simplified for understanding.
The SE has two attributes that make it stand out. It is assymetrical by nature. There is only one tube that must "push" and "pull" the speaker cone toward you, or away from you. The tube can actively pull its plate toward ground, but it can not force it toward B+, it can only reduce conductivity allowing the plate voltage to return toward B+. This return can be impeded by the magnetic properties of the OPT / speaker system. This means that the dynamic ability to abruptly move the speaker cone is assymetric, and the overall system dynamics of an SE amp will not be as good as a P-P amp of similar design.
All linear systems in nature are never perfectly linear. The magnetics in the OPT are no exception. The ability to turn current into magnetic energy and back to current operate fairly linearly over a large range except around zero net current, and as magnetic saturation is approached. Near zero OPT current some energy is expended magnetizing the core before any energy is succesfully transfered to the secondary. An SE amp runs a DC current through the OPT. This "idle current" moves the idle magnetic flux in the OPT to the middle of it's most linear range. Small "details" move this idle point around in the middle of its linear range so very little information is lost. A P-P amp runs identical idle currents through two opposing windings which create opposing magnetic fields that cancel each other. At zero signal the idle magnetic flux in the OPT is zero. Some energy must be expended to magnetize the core, so small details can be lost. The OPT spends a considerable time in the zone around zero magnetic flux where linearity is poor.
An SE amp MUST operate in class A since there is only one tube. It is the least efficient and will make the most heat. The output tube must be run continuously at somewhere between 70 and 100% of maximum, so tube life will be shorter that a P-P amp. Sonically an SE will do better expressing detail The bass control is dependant on OPT, tube, and speaker combination. Dynamics is good, but many P-P designs can be better. The distortion is predominately second harmonic which is sonically pleasing or neutral. The higher order harmonics are quite low.
A P-P amp can be class A or AB and can even approach true class B.
A class A P-P amp is essentially two SE amps operating on opposite sides of the same signal. The efficiency is still poor, the DF is doubled over the SE amp and the even order harmonics are mostly cancelled. The higher order harmonics are quite low, and the third harmonic (also sonically neutral in low amounts) is dominant. The dynamics are quite good since the amp is symmetrical.
A class AB P-P amp is the same design as the class A P-P amp, except that the operating conditions are adjusted such that one output tube shuts off for a portion of the time, while the other one does all the work. This leads to higher efficiency, longer tube life, but increases distortion and reduces the DF. The higher efficiency allows for much greater power output for a given tube size.
Most class AB amps, tube or solid state generate too much distortion to be used for audio without some form of negative feedback to reduce the distortion. Negative feedback reduces the distortion and increases the DF, it also reduces the total system gain, so an additional amplification stage or two may be required to make up for the feedback. Small amounts of feedback can help an amplifier sound better, but large amounts make the (DF, distortion) numbers look good but the amp sounds sterile.
The choice of tubes governs the amount of reserve current capacity that an amplifier can deliver. Consider the case where the woofer cone is moving in one direction when a large transient commands it to instantly reverse direction. At this instant the dynamic impedance of the speaker is very low. The amp must be capable of pumping a large amount of current into the speaker for a brief period of time. A solid state amp will generally have an edge here. In this case, all tubes are not created equal either. The average audio tube has a cathode structure capable of supporting the maximum power output of its design with a small reserve. If the amp is already operating near this limit it can not deliver the extra current for this transient and the transient response will suffer. The venerable 6L6GC is one of the worst in this case, with the KT88 being much better.
There was however a special class of tubes designed for the sweep circuits in tube type TV sets that have HUGE reserve current capability. These tubes, called "sweep tubes" are no longer made and some versions are getting scarce, but some are still quite common. They can put out large power levels and still have ample reserve capability. I have been using them in one off amp designs for about 20 years. The DCPP amp uses sweep tubes.
I got my first DCPP board from Pete for experimentation. It is still used for experiments, hence it is only counted as 1/2 of an amplifier. Pete claims 18 WPC. That looked like a conservative claim. I stated in the build thread when I bought it that I would be dissapointed if I couldn't get at least 50 WPC. Boy, was I wrong. One of my designs which involved minimal modifications to the board, but added BIG tubes and power supply reached 250 WPC and 525 watts in paralleled mono. Enough reserve?
So, the SSE and TSE are both class A SE amplifiers. They will deliver the best detail of any of the amps you asked about. Bass can be quite good with a big output tube (KT88's in the SSE) and good sized OPT's (Edcor CXSE). The dynamics and transients are good, but a P-P amp will do better on some demanding music. If a powered sub is used to handle the frequencies below 100Hz, then the TSE can be built with smaller OPT's and the results are outstanding.
The SPP, and DCPP are both class AB1 P-P amps.
The SPP is a 15 WPC design based on the 6BQ5/EL84 tube. There is a version that used the 6CW5 tube. The design has been around since the 1950's and has been used by many major amp manufacturers. Many people like this sound and build the amp because of the sound quality. There is more reserve capability in the SPP than either the SSE or TSE because it is a P-P amp and the EL84 can put out 30 to 35 WPC if squeezed a bit.
The DCPP is a "18 WPC" design based on TV sweep tubes. The tubes that Pete specifies have the capability of putting out 50 to 75 WPC. Plenty of reserve there. In fact the DCPP can be a 30 WPC amp just by swapping OPT's. I think it has the best transient capability of the bunch.
Which amp do you build? What are your priorities?
I have ripped my work room completely apart, and am in the process of rebuilding it. I plan to have 3 sets of speakers, and 3 or 4 amps plus a powered sub. I am space limited and have heat issues so these limit what I can use. I plan to build a new SSE and a new TSE in rack mount chassis to conserve space. I am working on a new design sweep tube amp that will make 100+ WPC, also in a rack mount chassis. There will also be a small form factor P-P amp also based on sweep tubes. This one will be on whenever the computer is on, so power consumption muse be minimized to reduce heat.
I will be moving in a year or two, and I plan to have a bigger work space in a cooler climate. The SSE, TSE and the big P-P amps will be kept, maybe the 40 WPC SE, and a bigger powered sub.
Note that I have been swapping out amps as fast as I build them over a 7 year period. There is only ONE that has remained a constant. It is actually the SAME industrial SSE that I built 7 years ago on a diamond plate chassis, and it still has the same set EH KT88's in it from 2007.
I have a 2 WPC TSE that IS the most detailed amp I have. Even with the 87db Yamahas it sounds great on some music. Norah Jones, Diana Krall, they are in the room with you. Try to play Metallica, Depeche Mode, or Pink Floyd.....not pretty. My high efficiency FH-3's are nearly finished, and I might change my mind about flea powered SET. That same amp on 106db Lowther based horns will crank anything at any reasobnable volume with amazing detail.
The Edcor CXSE transformer is by far the best deal in "budget" OPT's. The big Hammond does have better bass, but for some reason it gives up detail especially on the 106db Lowthers. For me that is my go to transformer since my budget is in the "budget" price range. I do have a set of custom Electra Prints in the 2 WPC TSE, and maybe that's the reason that the little TSE has kicked the butt of many expensive amps in showdowns.
It is possible to have too big an OPT. In fact if you are using a powered sub so that the SE amp doesn't need to work below 100 Hz, then its hard to beat the $29 Edcor.
I have a system in the living room that uses a pair of 15 inch Silver Iris OB coaxial drivers mounted in 1941 vintage Zenith console radios to achieve acceptable WAF. These speakers are 96db efficient and go down into the low 40Hz region. They can eat 100+ watts too. There is now a flat screen TV in the wall unit where the CRT TV was so the space behind the TV is available for just about any amplifier regardless of ugliness, it's not visible. I have swapped dozens of amps through this system. The usual amp here is an SSE with the big Edcor CXSE output transformers. This combination will shake the whole house with KT88's in UL or deliver the sweet sounds of a female vocalist with EL34's in triode. When it's time for a rock concert in the living room, or rattling the neighboorhood with tecno or EDM, there is only one valid choice, my hotrodded version of Pete's DCPP that cranks out 125 WPC. I can open the house windows, crank this thing to 11 and go outside to mow the lawn.....without hearing the lawnmower. I also like the TSE with 307A tubes in it on these speakers.
I have another system (actually two) in my work room. This is a small room (10 X 11 feet) that is my lab, computer room, music production studio, and primary listening location. Heat is a big factor here. The house AC unit AND a small wall unit can not keep this room cool with a lot of stuff turned on in the south Florida summer. The go to amp here has been the SSE with smaller Transcendar OPT's through Yamaha NS-10 Studio monitors. These speakers don't go below 75 Hz so I use a 10 inch powered subwoofer. This has been the case for 6 years. The Yamahas are only 87db so they don't go terribly loud with 7 watts, and neither does the wimpy sub, but it is plenty in the small room. I often swap in an SPP here. It has a different sound, which I will try to explain later. The DCPP has a similar sound, and goes far louder, but generates more heat, so it doesn't see much use. I have a 40 WPC SE amp, but I don't use it because the room hits 85 degrees in about an hour when it's on. I am in the process of rebuilding the studio yet again, so the amp line-up will change, but an SSE will always be a part of it....so will some sort of P-P amp. A new pair of FH-3 horn speakers is being added.
The sound.....
Both of these statements are mostly true, but not the only factors. I don't like flabby bass.
Most speakers require a good DF for solid bass. The DF requirement depends greatly on the woofer design. Ported cabinets generally will go lower than the same woofer in a sealed box, but the DF requirements are greater. The drivers in my OB cabinets are designed for open baffle systems and do not need a huge DF.
A SE tube amp generally does not use much feedback if any. This guarantees a low DF, how low.....1. OK 1 is unusual, but not uncommon. Usual is between 3 and maybe 10.
The impedance of the OPT is the biggest factor here. As you increase the impedance the tube sees, the DF goes up, and the distortion goes down, but so does the output power. The DF can be increased, and distortion decreased, by increasing the tube current, at the expense of tube life. Bigger tubes tend toward higher DF over smaller tubes in the same amp (KT88 VS EL34 VS 6L6GC). These are all compromises made during amp design or build time.
This is why I recommend a 5 K OPT over the usual 3.4K OPT in the SSE.
I recover the lost power by increasing the B+. The SSE runs around 440 volts. The SSE can run triode mode which has the highest DF, lowest distortion, but lowest power. It can also run UL for nearly double the power with a slightly higher distortion, and loss of some DF. A few users have tried pentode mode, but the DF IS near one which doesn't work on most speakers.
I am an electrical engineer by trade, so it is my nature to over analyze things to figure out what is going on. The explanation that follows is somewhat simplified for understanding.
The SE has two attributes that make it stand out. It is assymetrical by nature. There is only one tube that must "push" and "pull" the speaker cone toward you, or away from you. The tube can actively pull its plate toward ground, but it can not force it toward B+, it can only reduce conductivity allowing the plate voltage to return toward B+. This return can be impeded by the magnetic properties of the OPT / speaker system. This means that the dynamic ability to abruptly move the speaker cone is assymetric, and the overall system dynamics of an SE amp will not be as good as a P-P amp of similar design.
All linear systems in nature are never perfectly linear. The magnetics in the OPT are no exception. The ability to turn current into magnetic energy and back to current operate fairly linearly over a large range except around zero net current, and as magnetic saturation is approached. Near zero OPT current some energy is expended magnetizing the core before any energy is succesfully transfered to the secondary. An SE amp runs a DC current through the OPT. This "idle current" moves the idle magnetic flux in the OPT to the middle of it's most linear range. Small "details" move this idle point around in the middle of its linear range so very little information is lost. A P-P amp runs identical idle currents through two opposing windings which create opposing magnetic fields that cancel each other. At zero signal the idle magnetic flux in the OPT is zero. Some energy must be expended to magnetize the core, so small details can be lost. The OPT spends a considerable time in the zone around zero magnetic flux where linearity is poor.
An SE amp MUST operate in class A since there is only one tube. It is the least efficient and will make the most heat. The output tube must be run continuously at somewhere between 70 and 100% of maximum, so tube life will be shorter that a P-P amp. Sonically an SE will do better expressing detail The bass control is dependant on OPT, tube, and speaker combination. Dynamics is good, but many P-P designs can be better. The distortion is predominately second harmonic which is sonically pleasing or neutral. The higher order harmonics are quite low.
A P-P amp can be class A or AB and can even approach true class B.
A class A P-P amp is essentially two SE amps operating on opposite sides of the same signal. The efficiency is still poor, the DF is doubled over the SE amp and the even order harmonics are mostly cancelled. The higher order harmonics are quite low, and the third harmonic (also sonically neutral in low amounts) is dominant. The dynamics are quite good since the amp is symmetrical.
A class AB P-P amp is the same design as the class A P-P amp, except that the operating conditions are adjusted such that one output tube shuts off for a portion of the time, while the other one does all the work. This leads to higher efficiency, longer tube life, but increases distortion and reduces the DF. The higher efficiency allows for much greater power output for a given tube size.
Most class AB amps, tube or solid state generate too much distortion to be used for audio without some form of negative feedback to reduce the distortion. Negative feedback reduces the distortion and increases the DF, it also reduces the total system gain, so an additional amplification stage or two may be required to make up for the feedback. Small amounts of feedback can help an amplifier sound better, but large amounts make the (DF, distortion) numbers look good but the amp sounds sterile.
The choice of tubes governs the amount of reserve current capacity that an amplifier can deliver. Consider the case where the woofer cone is moving in one direction when a large transient commands it to instantly reverse direction. At this instant the dynamic impedance of the speaker is very low. The amp must be capable of pumping a large amount of current into the speaker for a brief period of time. A solid state amp will generally have an edge here. In this case, all tubes are not created equal either. The average audio tube has a cathode structure capable of supporting the maximum power output of its design with a small reserve. If the amp is already operating near this limit it can not deliver the extra current for this transient and the transient response will suffer. The venerable 6L6GC is one of the worst in this case, with the KT88 being much better.
There was however a special class of tubes designed for the sweep circuits in tube type TV sets that have HUGE reserve current capability. These tubes, called "sweep tubes" are no longer made and some versions are getting scarce, but some are still quite common. They can put out large power levels and still have ample reserve capability. I have been using them in one off amp designs for about 20 years. The DCPP amp uses sweep tubes.
I got my first DCPP board from Pete for experimentation. It is still used for experiments, hence it is only counted as 1/2 of an amplifier. Pete claims 18 WPC. That looked like a conservative claim. I stated in the build thread when I bought it that I would be dissapointed if I couldn't get at least 50 WPC. Boy, was I wrong. One of my designs which involved minimal modifications to the board, but added BIG tubes and power supply reached 250 WPC and 525 watts in paralleled mono. Enough reserve?
So, the SSE and TSE are both class A SE amplifiers. They will deliver the best detail of any of the amps you asked about. Bass can be quite good with a big output tube (KT88's in the SSE) and good sized OPT's (Edcor CXSE). The dynamics and transients are good, but a P-P amp will do better on some demanding music. If a powered sub is used to handle the frequencies below 100Hz, then the TSE can be built with smaller OPT's and the results are outstanding.
The SPP, and DCPP are both class AB1 P-P amps.
The SPP is a 15 WPC design based on the 6BQ5/EL84 tube. There is a version that used the 6CW5 tube. The design has been around since the 1950's and has been used by many major amp manufacturers. Many people like this sound and build the amp because of the sound quality. There is more reserve capability in the SPP than either the SSE or TSE because it is a P-P amp and the EL84 can put out 30 to 35 WPC if squeezed a bit.
The DCPP is a "18 WPC" design based on TV sweep tubes. The tubes that Pete specifies have the capability of putting out 50 to 75 WPC. Plenty of reserve there. In fact the DCPP can be a 30 WPC amp just by swapping OPT's. I think it has the best transient capability of the bunch.
Which amp do you build? What are your priorities?
I have ripped my work room completely apart, and am in the process of rebuilding it. I plan to have 3 sets of speakers, and 3 or 4 amps plus a powered sub. I am space limited and have heat issues so these limit what I can use. I plan to build a new SSE and a new TSE in rack mount chassis to conserve space. I am working on a new design sweep tube amp that will make 100+ WPC, also in a rack mount chassis. There will also be a small form factor P-P amp also based on sweep tubes. This one will be on whenever the computer is on, so power consumption muse be minimized to reduce heat.
I will be moving in a year or two, and I plan to have a bigger work space in a cooler climate. The SSE, TSE and the big P-P amps will be kept, maybe the 40 WPC SE, and a bigger powered sub.
Note that I have been swapping out amps as fast as I build them over a 7 year period. There is only ONE that has remained a constant. It is actually the SAME industrial SSE that I built 7 years ago on a diamond plate chassis, and it still has the same set EH KT88's in it from 2007.
I have a 2 WPC TSE that IS the most detailed amp I have. Even with the 87db Yamahas it sounds great on some music. Norah Jones, Diana Krall, they are in the room with you. Try to play Metallica, Depeche Mode, or Pink Floyd.....not pretty. My high efficiency FH-3's are nearly finished, and I might change my mind about flea powered SET. That same amp on 106db Lowther based horns will crank anything at any reasobnable volume with amazing detail.
The Edcor CXSE transformer is by far the best deal in "budget" OPT's. The big Hammond does have better bass, but for some reason it gives up detail especially on the 106db Lowthers. For me that is my go to transformer since my budget is in the "budget" price range. I do have a set of custom Electra Prints in the 2 WPC TSE, and maybe that's the reason that the little TSE has kicked the butt of many expensive amps in showdowns.
It is possible to have too big an OPT. In fact if you are using a powered sub so that the SE amp doesn't need to work below 100 Hz, then its hard to beat the $29 Edcor.
Question for tubelab:
40's good output for an SE. Is that pentode, or parallelled triodes? Or an 833 (or GU-48)? Or are you running a GM-70(or 572 into A2? Inquiring minds (and fellow EE's) want to know.
The amp runs a single 845 transmitting tube in each channel. It runs in A2 to get 40 WPC. B+ is 1050 volts. I built this about 6 years ago. It is essentially a TSE with another stage added. It uses the TSE PC board. Schematic is here....the third one down on the page.
845 SE | Tubelab
845 SE | Tubelab
Tubelab, your reply was awesome. I've been diving into tubes head first and that really explained and pulled together a lot of concepts for me.
You asked a good fundamental question. "Which amp do you build. What are your priorities."
Fundamental priority would be best detail retention with lowest possible distortion.
The problem lies in that while this is a priority I would certainly want to have good bass, bass/driver control, dynamics, and transients.
Given your description it would seem that best choice would be an sse with big high impedance opt, negative feedback, and big tubes in triode? Would adding a bigger opt like csxe25 improve power output? Are there any other opt, or perhaps an entire combo you might recommend to get typical SE detail while coming as close as possible to PP type dynamics, transients, and df?
The pure class a p-p sounds intriguing. Highest df, lowest distortion, pp dynamics. But it still doesn't get around the design parameter of a pp where small details can be lost according to your above description correct?
You asked a good fundamental question. "Which amp do you build. What are your priorities."
Fundamental priority would be best detail retention with lowest possible distortion.
The problem lies in that while this is a priority I would certainly want to have good bass, bass/driver control, dynamics, and transients.
Given your description it would seem that best choice would be an sse with big high impedance opt, negative feedback, and big tubes in triode? Would adding a bigger opt like csxe25 improve power output? Are there any other opt, or perhaps an entire combo you might recommend to get typical SE detail while coming as close as possible to PP type dynamics, transients, and df?
The pure class a p-p sounds intriguing. Highest df, lowest distortion, pp dynamics. But it still doesn't get around the design parameter of a pp where small details can be lost according to your above description correct?
- Status
- Not open for further replies.