Thursday, November 30, 2017

6SN7 NOS



An interesting article about the origins of the 6SN7 .
I dont agree with some of the comments , but some good info here, especially the historical content.
Double click to expand page.




Wednesday, November 29, 2017

FAQ -DHT PREAMP

What happened to the Directly Heated Triode preamps you were building ?

Yes I built DHT preamps for quite a long time , I built the first commercially available 300B preamp, and built quite a few different models, using different DHT tubes over the years. 45, PX4, 33, 101D, 4P1L and a bunch of obscure tubes.

I still very much like the lush, slightly warmer sound of DHT , and I still have a DHT preamp in one of my systems.

In the right system, or to cater to a certain subjective taste they can be very enjoyable, and I will still custom build them if required.

Its a bit of a shame that the best DHT preamp tubes are NOS from the 1930-50 period, I really like the 71A, 10/10Y/801A , but buying them now is a bit of a lottery as to what you will get.

And DHT tubes, because the filament is in the signal path,  are usually very microphonic. Some are worse than others, the 300B isn't too bad , the 4P1L can be quite "ringy".

It doesn't particulary worry me, I even think a little microphonics can be a good thing, but it drives most audiophiles crazy, as they think that something is "wrong" with the tube.

All the time I was building DHT tubes , around 10+ years, I still had 6SN7 preamps, and continued to work on them .

The octal 6SN7 tube was designed for applications such as preamps, and is maybe the ultimate preamp tube.

Everyone uses 12AX7,12AU7, 12AT7 etc, but these tubes were designed for TV purposes and to my ears aren't a patch on the 6SN7 and its Russian equivalent the 6H8C.

The 6SN7, in the right circuit, which lowers the output impedance to an optimum level , without any side effect degradation , is just about perfect in every way.

Its accurate, has a big soundstage and sounds natural and full of ease. Clean  solid bass and lots of air.

Most other tubes sound closed in, strained and brittle in comparism.

A couple of years ago I found a combination of componentry and circuit which gave the 6SN7 preamp a performance that I thought gave all the advantages of DHT without any disadvantages.

Its since been fine-tuned and is my number 1 product .

Tomorrow I'll talk about different 6SN7 tubes and the pitfalls and advantages of NOS.





Tuesday, November 28, 2017

That's not a tube, THIS is a tube!

Apologies to Crocodile Dundee.






211 on the left V1505/212E on the right.

I built quite a lot of SE amps 20 years ago, the V1505 is the most impressive tube , sounds pretty good too, like a giant 300B.

Monday, November 27, 2017

Sunday, November 26, 2017

6SN7 Preamp specs

Someone asked for specs of 6SN7 preamp,  specs don't really tell you how a preamp will sound, but here they are, in approx order of importance:
Output Impedance 1300 ohms.
Gain 0-26 dB adjustable.
Tube Compliment 6SN7  linestage 5881 5AR4 power supply
6688 6H23 phono preamp.
Power supply 300VA transformer, CLCRC shunt regulated
Copper chassis chrome plated
Remote volume
Lundahl, Magnequest amphorous input and output transformers

What Makes a good Tube Preamp- Pt3 Power Supply


A tube power amp can get away with a good solid power supply (PS) that consists of a big transformer , rectifier and a simple capacitor /inductor/capacitor bank.
If its big enough and of sufficient quality thats all that's really needed.

A preamp though is a different story- its a gain device and if the PS isn't clean, fast and powerful the good (the signal) will get amplified along with the bad (noise and distortion) .

A really good PS for a preamp will have 0.0mV of noise at the voltage output of the PS feeding into the loads of the triode tubes used for gain.

So for example, a Supratek will have 300V DC at the loads and will measure 0.0mV AC with no signal.

There's not many preamps that can achieve this and it takes an evolved and sophisticated design to do this competently and reliably.

Personally , I dont really take a PS for a preamp seriously unless it is regulated.

This means the voltage will always remain constant no matter what the mains voltage is or regardless of the demands of the circuit to produce distortionless amplification.

There's simple and more sophisticated ways of doing regulation, a good solid state regulator can be made with a handful of transistors and components, but being a died in the wool tube guy, I prefer a tube based regulator.

These come in either series or shunt versions, series is simpler.

Tube shunt regulators bleed off a portion of current to earth to maintain a constant current through the preamp circuitry, and from a sonic perspective are superior to series regs.

I've been using shunt regulators in my preamps forever, and they are a big determinant in the sonic signature of the Supratek preamps.

They provide a "blackness" around the amplified signal, with zero noise and distortion.

There's no strain with complex or dynamics passages as the voltage to the circuit is always constant and uninterrupted.

In contrast, an unregulated PS will "sag" with the demands of voltage peaks.

Component choices are also a big factor in performance. Once again, you will hear me say how detrimental electrolytic capacitors are to audio quality, and how so many audio devices these days are filled with them.  Its not uncommon for a modern preamplifier to have 50 or more electros in them, with a maximum lifetime of 10,000 hours.

There are 2 electro caps in a Supratek linestage, one for the tube heaters, and one for the remote volume.

Polyprop caps are much better, or even better again, minimal or no cap designs simplify design, performance and reliability.

There is no substitute for transformer size, you can definitely have too small a transformer, but its hard to go too big!

The Supratek preamps have a seperate power supply chassis simply because the PS is too big to fit in one chassis, and there are advantages in getting the transformer and inductors away from the circuitry.

I also use 1.2mm copper chassis for both preamp and PS, its expensive but if you live with a copper chassis for a while you will hear that a steel chassis does have a "sound" that is less neutral than copper.  It's almost intangible , but it's there.

There's a good reason the hyper expensive Japanese AudioNote gear is copper chassis based.


Power supply on the left. The middle tube is rectifier, the outside tubes are the shunt regulator tubes.


Saturday, November 25, 2017

60 Year old tube amp that kicks arse.

Ok lets get away from the technical stuff for a day and do some reminicising.
A very long time ago I found a pair of Altec-Lansing 604B coxial speakers in a big cabinets on the side of the road.
Turns out they were from the Australian Broadcasting Commission radio station, and in the back of each cabinet was an industrial looking tube amp.

This was back in the day when tube amps were being phased out, solid state was becoming dominant and tubes hadn't had the renaissance they enjoyed a decade later.
But you could still buy replacement tubes in the local electrical shop and I'd heard old guys tales that these old fashioned tube amps actually sounded pretty good.
A new 5Y4G rectifier got it going again , and goddam the 604B and amps sure sounded different to the English amps and speakers we were listening to back then.
The tube amp was an Australian made Trimax A52 , a push pull EL34 design with a 6AU6 input tube and 12AT7 driver.
These were built on commission to the ABC , Trimax was mostly a transformer winding company , who had a very clever and enterprising design dept.
They produced quite a varied range of professional gear for the ABC, including some interesting preamps  and mixers.
The A52 was loosely based on the Mullard 5-20 EL34 amp. It put about nearly 30 watts with only a minimal amount of feedback, achievable because of the very high quality of the output transformers.

Its a very elegant design, straight out of the Golden Era of tube design. A fantastic output transformer, that would take the very best winders to match today.
Only one electrolytic capacitor in the entire amp, and that was probably installed later in its life.
Todays electronic components have hundreds of electrolytic capacitors , mostly used because they are very economical- the paper in oil caps used back in the 50's and 60's were much costlier, and can be as good as new 60 years later!
The pair of EL34 tubes put out a solid 30 watts,  but its not stressed or working hard to achieve this, so it will keep working reliably year after year.


I dont have the original pair I found 40 years ago, but i do have another 2 sets of amps I bought about 5 years ago.
I have a lot of respect for these amplifiers, they are of a good solid, timeless design, have incredible reliability that comes from over engineering of components,  and sound excellent.
There are very few audio devices that can maintain this sort of integrity and quality over a 60 year period.
It may be as ugly as sin, but underneath that exterior is a real beauty.
The Trimax A52 inspires me in my own preamp building, my preamps are very over engineered, with massive power supplies, and a reluctance to use those damn electrolytic capacitors with their limited life.
Suprateks built 25 years ago are still working reliably  and I'm even more focused on the latest models being super reliable and long lifed.
This is what makes a classic in the vein of the old tube Marantz , Luxman products. They are simply built to last, something that is sadly not a priority these days.
I was using my Trimax's in an active/passive configuration with my Supraxox field coil speakers, but the Supravox are going back to France for a re-cone and i now have the Trimax's on my AR-LST speakers.

These are another high quality, over engineered product from long ago, a fantastic speaker that was very revolutionary when introduced. It has 9 drivers per speaker, a 10 " bass driver, 4 mid range and 4 tweeters per speaker, kind of like a big Bose .
Despite receiving much praise, the LST's never really took off as they came out at the end of the tube era and they needed high power solid state to work well.
They were just ahead of their time, the high power Crown solid state amps they were partnered with were quite horrible sounding amps.
Ive split the crossovers on mine so the three bass, midrange and tweeter sections can be each driven by a separate amplifier, with the Trimaxs driving the mid range and tweeters.
This relieves the need for a really high power amp and the 30 watt Trimaxs are adequate for mid and treble, with a high power SS amp on bass.
I've installed level controls on each amp, so tone can be adjusted to suit. Of course a powerful preamp that can drive 3 sets of amps is also required.
LST stands for laboratory standard transducer and that is pretty much what they are, monitor standard with the ability to colour  the tone if required.
There's a real satisfaction using products you know are well designed , engineered and constructed to perform at the highest level, and last a lifetime .
The Trimax was probably the highlight of the Australian audio manufacturing of its era, and I feel privelged to own them.






Friday, November 24, 2017

What makes a good tube preamp- Pt2 The Circuit

In todays marketing focused world there's a lot of hype - especially about hyper quality components- exotic capacitors , resistors etc .

But the determinant factor in quality sound reproduction is simply the requirement for a high quality electronic design (or circuit)

Super duper capacitors and resistors wont turn a pig of a circuit into something that sounds wonderful.

A really competent design with bog standard , nothing-fancy parts will rock the socks off a poor design that someone has stuffed with $1000 of hyper parts.

The trick is to use the most advanced, up to date and state of the art design and then carefully choose the componentry to achieve the desired sound signature.

Its a matter of synergy and the experience of knowing what works best in situation.

I've never been a big fan of electrolytic capacitors, but most audio devices have hundreds of them in them.

Eliminating them is difficult and expensive, but the result is very worthwhile, as it results in a superior phase performance, which means more 3D imaging and more expansive sound-staging.

My Supratek preamps have typically no more than 3 or 4 electro caps , and used only where absolutely necessary.

The painfully common direct coupled/cathode follower design is the stalwart of preamplifier design and is employed in the vast majority of low cost preamps on the market.

Cheap to make with 12AU7 or 12AT7 tubes , and with reasonable performance , there must be many thousands of these preamps built over the years.

Its a great way for designers who wish to produce a tube preamp without too much thought or development- its a proven design with just enough "tube" sound to classify as a tube preamp.

These designs are really hangovers from the 1950's , state of the art tube design has well and truly moved on from there , but surprisingly few modern designers deviate much from the tried and true.

More innovative and developed designs use a hybrid mix of tube and solid state components, and even iron chokes and transformers in the mix,  with sophisticated power supplies to give highly accurate frequency response , the ability to drive any amplifier and superior phase performance to give realistic imaging and sound-staging.

Tommorow- power supplies.






Thursday, November 23, 2017

What makes a good tube preamp? PT1 - Output Z

What makes a good tube preamp?

There's quite a few tube preamps available from the cheapest Chinese import to the exotic and expensive, so what makes a tube preamp good?

The most important feature is low output impedance ie the ability to drive loads. A power amplifiers load (input impedance) will vary from typically 5000 ohms (solid state) to 100,000 ohms(tube) and its generally recognised that a ratio of 1:10 is suitable, ie the preamp will have an output impedance 1/10 of the input impedance.

So if a tube preamp is going to sucessfully drive a solid state power amp with an input impedance of 5000 ohms it should have an output impedance of around 500 ohms.

This is just about worst case- 5000 ohms input is quite low, but not uncommon.

What happens if a preamp's output impedance (Z) is too high?

The preamp will not be able to transfer its frequency response in a linear fashion, it will be inaccurate and "sloppy' sounding.

This can happen when an unsuitable high resistance tube is used in a circuit that does not do any "shifting" from high to low output Z.


Tubes like the 12AX7, 6SL7 are often used in simple, minimal preamps but their high resistances require circuits such as the cathode follower to be used , which lower the output Z.

Unfortunately there's a price to pay , as 100% feedback is employed in this circuit, which does introduce a bit of haziness and veiling.

Without the cathode feedback, the high resistance tubes sound slow, warm and so "romantic" they will put you to sleep pronto.

Another way to lower Z is to use a "transformer" a step down device which can drop the ratio by a chosen amount,  for example a 4P1L triode tube has a resistance of 1000 ohms, if we use a 4:1 ratio step-down we will end up with a  Z of 250 ohms, which is low for a tube preamp, and lower than you will ever need!

Transformers vary tremendously in quality , and they must be carefully matched to each tube and application. Transformers don't work well with high resistance tubes and even with well matched tubes, its quite an art to build a really competent and high performance transformer.

I use the 6SN7 tube in my preamps , with a special circuit and Lundahl or Magnequest output transformers the preamps have an output Z of 1300 ohms , nice and low .

So output Z is important for a preamp in order to control a power amp with accurate frequency response and drive.

Tomorrow I'll talk about something else that's important.