Jump to content

SET and pushpull (newbie questions)


markw

Recommended Posts

I've been readily enjoying this forum for about two weeks now.

Although I love my cornwalls to death, I have only ever heard them on SS amps (Carver CM1090, Marantz MA500, Sherwood Newcastle R945)

Now I'm hearing that a tube amp will really make a difference. (Ingorance is bliss, I guess) So here are some questions. I haven't seen these question on the board -maybe it will serve as a primer for other newbies.

What is the difference between the technologies - heck - that ARE the technologies?

Would I be wise to go for a DIY kit (I'm pretty good with a soldering iron and think it would be cool)

What should I be listening for, anyway? (I realize that's subective, but I guess mainly with respect to differeing amp types and technologies. ie- A SET amp will sound more xxxx vs. SS and more yyyyy vs. Push-pull")

There. I hope a haven't dumbed-down this forum by asking this, but I'll bet there's more than a few lurkers out there who'd like to know these and probably other questions I've forgotten (or don't know enough) to ask.

Link to comment
Share on other sites

Musical tastes: any and all (not so much on rap) Classical, jazz, blues, alternative rock...

Room size: presently 16 X 25 ish - present config: corwalls spaced along long wall about 5 ft apart - couch directly across. (if we're getting specific - I have vert mounted drivers)

not sure what 'spl' is... If it refers to volume, My favorite way to listen is not too loud, but above just normal conversation level. Second favorite is as background music. Third is TOO LOUD! 9.gif

Link to comment
Share on other sites

I'll leave the technical babble for the engineering types on this forum. Tubes will give you a warmer & more presents presentation over SS. Notes will be more rounded and extended for the most part. At low volumes you'll find a fuller presention over the whole sound spectrum. What you will also find is that you will be able to listen to music at a louder volume with ease and not fatiquing like SS. it's an addictive kind of sound, that draws you closer to the music, very inviting. Ad vinyl to this formula and your there sort to speak. There are alot more characteristical differences between SS & tubes but I think you get the picture. As far as DIY, I ordered the Paramours & Foreplay from Bottlehead, these are suppose to be very simple to build & yield you great results when done correctly. they are a SET configuration about 3.5 watts, according to the info you gave it sounds like this could be a go. Push pull types I own a Scott 299a, which is great bang for the buck, pick one up on ebay & send it to Craig at NOSvalves for a go through, you will have a great vintage 20 watt integrated. IMHO

Tom

Link to comment
Share on other sites

That does help somewhat, thank you. (although I don't know what SET is or does - same for push pull)

I realize the end-all, be-all of this is to sound as good as possible, but right now I'm looking for a bit more theoretical and instructional - in general, moreso than specific - so that regardless of my listening tastes at the time, I'll know what to expect from various types of amps.

Tom, I hope you don't think I don't appreciate your help - far from it - It's exactly the kind of answer I'd give if I were asked. But without knowing all the different options, I could never be in a position to offer that kind of advice myself. (and boy do I like to give advice 1.gif )

Link to comment
Share on other sites

SET means Single Ended Triode. These are the types of tubes in tube amplifiers. A triode is a tube that has three parts. A pentode is a thermionic tube having five electrodes. Learning about the tubes gives you some general idea about what kind of sound and power they are capable. The less parts, the better the sound, but the less power it has. Push-pull topology divides the audio signal. It can provide more power, but the sound may be less pure.

Some generalizations: 2A3 tubes are weak, but delicate. EL-34s are more powerful, all around great performers for super-sensitive speakers, 845s are superb, but expensive, 300B are the power champs.

2.gif

Link to comment
Share on other sites

Mark, nice speakers! I love mine.

"The Complete Guide to High End Audio", by Robert Harley, is a very comprehensive book that might look funny on the outside but is very readable and informative on the inside. I picked this up a few months ago and have found it an indispensable reference. Here's the link to Amazon.

http://www.amazon.com/exec/obidos/search-handle-form/002-8279806-1219238

I too will leave the technical discussions to the engineers on this board; I still have official newbie status. But with high-efficiency horns, you only need a few clean watts for great listening. Most of the higher power amps show off their stats when running closer to their power ratings.

I've listened to my Cornwalls with a McIntosh MC2105 (100WPC of SS), and 3.5 WPC SET ... To my ears, the SET is a more wonderful sound; lots of imaging and clarity, almost halographic. Tough to beat and you can find those mentioned above, or a pair of used Wrights, for less than a grand.

Link to comment
Share on other sites

push-pull means that there's a tube at each end of the output transformer winding.

SET means single-ended triode, one tube runs the transformer.

in general, most of the vintage integrated amps (Scott, Eico, Fisher, etc) are push pull.

in general, a lot of the higher-end tube stuff is SET.

generally, a lot of the P-P stuff is 15-30 watts, SET stuff is 1.5-7 watts. Your Cornwalls will make around 100dB with one watt, so 3.5 is not as small as it seems.

Some people say one tube watt is equal to about 3-5 SS watts. I don't know how valid this is technically, but the crux of the matter is that these big old Klipsch speakers are so efficient at turning watts into decibels that wattage numbers are almost irrelevent. High powered SS amps typically sound bad at very low output, their wonderful sounding spec numbers are derived at full rated output. Tubes are clean and nice at fractional wattage output, sound much better.

At typical listening levels around 80dB or so, your Cornwalls are running on milliwatts, most SS is not good at this.

Just a few generalities for you. Personally, I'm doing great with vintage push-pull. I've learned to work on some of my own stuff, it's great and keeps me off the streets, so to speak.

Tom

Link to comment
Share on other sites

Push-Pull

The signal from the driver stage is split into 2 parts and each half of the signal is sent to a power amplifier section to be amplified. The 2 halves are then recombined to create a coherent whole. This technique allows higher power levels to be achieved:

Input signal A is divided into A1 and A2. A1 and A2 are separately amplified and added back together to create the final output signal B.

SET = Single Ended Triode

The output tube (or less commonly) tubes amplify the full waveform without resorting to any splitting and rejoining. These amplifiers are considerably lower in output power but many feel the have a nicer sound particularly in the midrange.

Input Signal A is amplified without being split and this produces the output signal B.

11.gif Hopefully I have not succeeded in making the issue more confusing. 11.gif

Link to comment
Share on other sites

To me the biggest advantages of tubes are the timbral accuracy and the huge soundstage/holographic image.

I have a pair of low-powered SETs (about 5wpc)

Push-pull designed will yield a bit more power and a more bottom-end slam, and are suited for those wanting to drive lower effecientcy speakers with tubes.

Because of Klipschs very high effecientcy, a whole lot of watts are not required to get amazing sound at respectable to LOUD listening levels.

Link to comment
Share on other sites

What is the difference between the technologies - heck - that ARE the technologies?

At the basic level of the device itself, tubes are much more linear through the audio range than transistors. Linear in this instance = low distortion.

Tubes act more like valves (just like a valve in water flow thru a pipe). Thats why they typically call vacuum tubes valves. Transistors act more like on/off switches.

SE or SET is a single-ended type of circuit design. Both tube & transistor amps can be single-end designs. In this type of amp the entire waveform is amplified by the tubes/transistors. One of the major drawbacks to SET inherent in their design is noise & hum. All things being equal, there is no way around this. A push-pull amp inherently cancels these much of these components.

In push-pull designs, you need at least 2 tubes or transistors for each stage of amplification. Each device handles ½one-half of the waveform. The waveform is switched back & forth between the 2 devices. This produces something called crossover or notch distortion. In most transistor amps this phenomena is more pronounced due to the fast on/off switching nature of the transistor. Furthermore, the amplitude of crossover distortion is constant. It stays at the same level regardless of the power output, which means as the power output decreases, the % of distortion increases. The highest % of crossover distortion is at the lowest power output levels. The only way around this is SE or class A operation which keeps all the tubes or transistors operation fully on at all times regardless of whether a signal is present or not. Yes, it burns more electricity & makes more heat.

Another major difference between tube & transistor amps is how they overload or clip when pushed past their operating limits. Transistors typically, literally shear the top/bottom of the waveform off. While this may look real nice on a graph or oscilloscope, consider what this really means. The flat part at the top & bottom of the waveform actually represent DC or direct current. Its the same as taking a battery & hooking it up to your speakers. At DC the speaker diaphragm wants to just move in one direction to maximum excursion & stay there. And now, you would be doing this back & forth as the waveform swings back & forth. Speakers dont like DC. Also consider the slope of the waveform. As the amp is overdriven into clipping, the vertical slope of the waveform becomes steeper & steeper. This is equivalent to forcing the amp to produce higher & higher frequencies, which are at this point, are probably far beyond the amps power bandwidth. Tubes overload much more gracefully. They act more like natural compressors, gently rounding off or softening the clipping.

Then there is the harmonics debate. Tube amps typically have more harmonic distortion but its of the pleasant type, even harmonics. Transistors seem to typically produce more odd harmonics which tend to sound harsh or edgy. The aforementioned clipping/overload situation would aggravate this even more.

In the old days, 1000Hz was chosen as the standard frequency to test for noise. The reason being that the human is most sensitive near this frequency. Tube amps just so happen to typically have the most noise here. As it turns out, transistors have much more high frequency noise. It doesnt show up in the specs because no one is testing for noise across the entire audio bandwidth, just at the standard 1000Hz specification. For example, I can turn my tube power amps all the way up & put my ear right next to the midrange horn on a Khorn & its absolutely dead quiet. With my SS amp which measures more than 20Db less noise according to spec standards, hiss is clearly audible at the listening position.

Another fascinating tidbit of history on the tube vs transistor sound has to do with something called negative feedback. Negative feedback is where you basically take part (a percentage) of the signal, invert it & feed it back into the signal. The inverted feedback signal cancels some of the distortion products thereby lowering distortion. Contrary to popular belief, negative feedback, in & of itself, is not necessarily a bad thing depending on how & where its used in the circuit & how much of it is applied. However, consider the following (elementary) scenario: If we take a typical bench test signal (1000Hz) & we play this through the amp. And lets say the amp can provide 20DB of amplification. If we take 2Db of this output, and apply it as negative feedback back at the input, we are left with 18Db of amplification. 2Db is 10% of 20Db so the distortion is reduced by 10%. BUT..this is with a static single frequency test signal. What happens when we play music through this setup? The negative feedback is now being mixed with a signal which is different than that of output signal from which the negative feedback came, in both its amplitude & frequency content. Its actually a distortion machine! To make matters worse, transistors have lots more gain or amplification than tubes, albeit at the price of bandwidth. Ill give you one guess what most (especially early) transistor designs did. Yes, they used that extra gain not only widen the bandwidth but also to apply even more negative feedback.

So what this all boils down to is differences hum, noise, crossover distortion, overload characteristics, & harmonic distortion content and the fact that there's just more opportunity to mess up a transistor amp design, than it is with tubes.

Ive heard superior sound from both kinds of amps, SE or push-pull, tube or transistor. However, after all is said & done, the transistor amps, when designed properly for true audio high fidelity, generally loose most, if not all, of their commonly associated advantages such as low cost, small size, less weight, less heat, high power output, etc.

As far as what you should be listening for.believe me. You wont have to listen for anything. It will all be quite obvious. Thats why tubes are still alive after almost everyone said they were obsolete over 30 years ago. Hmmm..kinda reminds ya of the Klipschorn doesnt it?

Link to comment
Share on other sites

Tube watts are not SS watts. The saying is that tubes watts are twice as powerful as SS watts. Now it turns out that Tubes Do Something Special (Peter van Willenswaard, Stereophile, September 2000, http://www.stereophile.com/showarchives.cgi?357). They put out a lot more voltage than we are aware of. Willensward measured recorded music on a 9-watt amp with 300B tubes. He found that You'd need a 50W transistor amp to realize the same peaks my 9W 300B launched without wincing at my speakers. That is more like 51/2 times the rated power of the tube amplifier.2.gif

Link to comment
Share on other sites

Like I said, it was an "elementary" explanation for our newbie friend Mark. I didn't really want to get into all the phase shift, current differential & biasing stuff, etc. At this stage I was trying to get the basic concept(s) across for him. But thanks for expanding on it.

Link to comment
Share on other sites

Understood mdeneen. I've had my triode monoblocks for over 25 years now. They are push-pull, class AB1 biased. I'm quite aware that up to several watts they are essentially operating in class A with Klipschorns 99.99% of the time.

While I have a technical background, I'm not an EE (electrical engineer). Thought I'd just try to pass these concepts along the way I learned them from the "outside".

I'm not a big SET fan. But maybe its time to revisit SET. The classic tubes for SET amps such as 2A3 & 300B were extremely scarce back in the 70's. My listening room is large, acoustically tuned, and VERY quiet. Noise of any kind is very easy to hear. And quite honestly, I like what I have, & have not heard anything better (excellent & different, yes, but not "better", overall ).

cheers!!!!!!

Link to comment
Share on other sites

If you like to tinker tube amps are where it's at, just by plugging in a different tube you can

change the sound of your amp. They are less completed to build as well, just be aware the

voltage present in a tube power supply will give you a jolt you will not so forget.

Bob

Link to comment
Share on other sites

Mark,

You've always had a knack for explaining things in such a way that enables me to comprehend more fully -- It's such a pleasure to have you back posting on the forum.

Is there a relationship between feedback and damping factor?

It is my understanding as of late that 12 to 18db of feedback can actually be a good thing -- or is this just a generalization that doesn't mean anything?

I've been studying texts explaining the differences between triode, pentode, and ultralinear. I didn't realize that Hafler was the big contributor to the ultralinear design. Because of my listening habits, and more specifically -- my musical tastes, I'm leaning very strongly to going back to class AB ultralinear with moderate feeback. My experience with triode and SET to this point, has me feeling that both suffer from constricted dynamics on peaks at the higher SPL's. I suppose if I had K-horns, I would enjoy either much more than I did.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...