Most certainly aggreed - recovery from clipping is a most important issue. That is the benefit of Neville Thiele's work, as I explained before (maybe not inthis particular thread).
In my opinion as a professional electronics engineer, with a lifetime's collection of electronics magazines, professional journals, and application reports - just about nothing.
However, there may be value in putting it together in one volume, or in a defined series of articles like Self did.
After, all, much of what Self did was already done and documented in a multitude of places. For example, one of his topics was extra distortion casued by operating the differential input pair with unequal emitter currents. I remember doing that one to death in 2nd year university - it was an examinable subject.
Am I'm sure that the professional engineers in major manufactuers of SS equipment, Quad, National Panasonic/Technics knew as much as Self - they just didn't publish in Electronic World, only peer reviewed journals.
Maybe not come up with complete new technology, but what I see from tube designs by for instance Bob Cordell (Linear Audio Vol 2) or Frank Blöhbaum (Vol 6 and 8) is that the best of these designers combine trusted tube technology with solid state to come up with a design that's better than the sum of the parts.
Combining ss with tubes allows you to improve the tube designs with near-perfect voltage regulators or constant current sources to mention just a few. Frank Blöhbaum's use of his BestPentode topology brings performance that simply isn't possible with only tubes.
So I am pretty sure that great things are still waiting to be done!
Jan
Fully agree Keith, I think the major contribution of Doug Self with his Distortion article series was that he managed to get it all together in one place, hanging together logically and consistently.
That is why I agreed to reprint the whole series, together with Peter Baxandall's views on it, in Baxandall & Self on Audio Power. And Svetlana let me do it, very kind of her! As well as the kind agreement by Peter Baxandall's daugter to include the 30+ page personal letter from Peter to Douglas.
Jan
Why not self publish? It costs you nothing and you get to keep all of the profits. Your works can be shipped worldwide and you don't have to lift a finger in the process.
Risky. I might make nothing at all. I am not a well known author in the audio field, though I have had non-audio articles published in Electronics World and elsewhere.
Publishing in a well established magazine guarantees a sizeable reader count, payment, and lends credibility to an unknown author. And editors are useful in weeding out the woffle and the odd bit of wonky prose.
@Keit,
You noticed myself and Jan are fishing for content?
Honestly the wages to be made from publishing an article, or a book for that matter, are not what they used to be. For articles ~$60 per page is about average, so unless you plan to write 1000 pages of worthwhile content you won't be making any real money.
For myself the audio hobby is just that, a hobby (albeit a serious one I take great pleasure in). I have generated a little extra income from being the editor in chief for the Elektor Audio Specials (it is a part time position) and publishing a number of articles, but considering the hours I've put in and the parts purchased I do not get to expense it is costing me money rather than making me any.
You noticed myself and Jan are fishing for content?
Honestly the wages to be made from publishing an article, or a book for that matter, are not what they used to be. For articles ~$60 per page is about average, so unless you plan to write 1000 pages of worthwhile content you won't be making any real money.
For myself the audio hobby is just that, a hobby (albeit a serious one I take great pleasure in). I have generated a little extra income from being the editor in chief for the Elektor Audio Specials (it is a part time position) and publishing a number of articles, but considering the hours I've put in and the parts purchased I do not get to expense it is costing me money rather than making me any.
@Keit,
You noticed myself and Jan are fishing for content?
Honestly the wages to be made from publishing an article, or a book for that matter, are not what they used to be. For articles ~$60 per page is about average, so unless you plan to write 1000 pages of worthwhile content you won't be making any real money.
For myself the audio hobby is just that, a hobby (albeit a serious one I take great pleasure in). I have generated a little extra income from being the editor in chief for the Elektor Audio Specials (it is a part time position) and publishing a number of articles, but considering the hours I've put in and the parts purchased I do not get to expense it is costing me money rather than making me any.
I was already well aware of that writing doesn't pay much. Especially when each page of an article on this topic written to the same quality as D Self would need 100's of hours of work - not the actual writing - the research and testing required in order to know what to write. But I would like to recover at least some of the cash outlay to buy parts I would not otherwise ever use.
The fact that writing doesn't pay much is a factor in why I want to assess how many out there would actually be seriously interested. So far I'm getting the feeling not a lot.
Unfortunately that's true. Linear Audio isn't making any money to speak of - I live off my 40 years of Dutch Airforce pension. As Keith mentioned, at EW they now no longer pay for articles; that's why you see all these company app engineers write articles - they charge it on advertisement.
I just heard from another author who wrote a paper for the AES that they want him to pay $ 50 a page for anything over a certain set number of pages. Can you imagine that? You want to write a good paper and you actually have to bring money to get it published??
Anyway, Linear Audio is a hobby-run-out-of-control, and as long as I can keep it going, I'll do it.
It's great to meet all these fantastic authors, and, I'll get to read all the articles first!
Jan
I just heard from another author who wrote a paper for the AES that they want him to pay $ 50 a page for anything over a certain set number of pages. Can you imagine that? You want to write a good paper and you actually have to bring money to get it published??
Anyway, Linear Audio is a hobby-run-out-of-control, and as long as I can keep it going, I'll do it.
It's great to meet all these fantastic authors, and, I'll get to read all the articles first!
Jan
I typically buy most DIY audio related books, Morgan Jones, Douglas Self, Bob Cordell, etc. simply because I always find something worthwhile in them, and typically use them for inspiration or to have the required formulas etc. handy if needed.
I am far from your typical DIY-er though, most DIY forums I visit revolve around folks that rather spend money on parts than books and would much rather buy a kit of some sort rather than designing something from the ground up or reading a book on how to design an amplifier to begin with.
I've seen this with the Elektor Audio Specials too. Back in '07 myself and Bruno Putzeys spent about four months on designing a SS class-A amplifier that offers astonishingly low THD and excellent performance overall.
I spent countless hours on the prototypes, PCBs, simulations etc. However I was quick to realize, after publication of the design, that most DIY-ers thought it had too many transistors and didn't want to bother building it. I've sold perhaps 50 sets of PCBs in the past 7 years.
In my humble opinion for a lot of people this hobby doesn't revolve around building something that outperforms store bought gear, but building something that works and for which they can claim they built it with their own two hands.
Add to that the whole quackery surrounding boutique parts and you'll quickly realize people rather substitute part A costing ten-fold of the original and get a subjective boost in performance than to tackle the real shortcomings of a design by educating themselves on the inner workings of their gear.
Reading books and understanding the principles that govern the operation of an amplifier is too much effort for most, they want a cheap and instant fix, that's what today's society is about mostly anyway, instant gratification. Attention to detail and true workmanship be damned, mediocre is the new high-end.
Alright, enough ranting, my apologies for getting a little carried away.
Last edited:
It's great to meet all these fantastic authors, and, I'll get to read all the articles first!
Jan
Do you get to edit out all the fluff and carry-on sentences too? That's the other gotcha with writing, some folks can easily put their thoughts to paper in a very comprehensive and easy to digest manner, others just ramble on in not very coherent sentences and jump from one topic to the next without any apparant structure
Where one can take a peek on that Thiele's works? other than the famous "Loudspeakers... "
Cheers
J.
Exactly! The majority of work written about vacuum tubes happened before the advent of modern solid state technology. Today we have far more options, far better modeling tech, and computing power on our desks that not even NASA could have imagined in the 1950's.
If we get over the notion that vacuum tubes must remain pure and void of all silicon based life forms, we can make a better amp. True, silicon devices have new challenges like huge gate capacitances that vary with applied voltage, but this can all be overcome with GOOD engineering practice.
About 15 years ago I started using CCS loads and mosfet buffers to ELIMINATE all overload recovery issues. I first published this here and on the Tubelab website.
Power Drive | Tubelab
At first the email was predominantly negative.....a few readers even suggested that I change my name to Transistorlab! After selling a few TSE amps incorporating PowerDrive, that changed, and the 10 year old TSE design still sells well. Why? Because there is no overload after clipping to recover from, because there is no coupling cap at the grid of the output tube.
If we really get down to it we can build a tube amp with a SMPS that is modulated by a DSP chip to feed the output tube precisely the voltage and current it needs to do it's job at any instant in the complex audio waveform, and thereby improve the efficiency of a class A tube amp beyond the theoretical maximum of 50%. Impossible? Well technically this would be a class H vacuum tube output stage, controlled by solid state electronics, but possible.
The amp I spoke of in the previous paragraph is not an engineering dream (or nightmare...depending on your point of view). I built this amp, and measured something like 75% efficiency on a SET amp that squeezed almost 20 WPC out of a 6AS7 with very low distortion. I did this in response to a design contest sponsored by Microchip that ran in Circuit Cellar magazine.
The design won a "category prize" in that contest, which got me a request from Circuit Cellar for a magazine article. I wrote an article, that ran in Circuit Cellar that described the amp, and its design.
So, what did it take to do this?
I was employed as a engineer at Motorola at the time and we were working long hours "advancing the state of the art in two way radio communications". The contest had a long deadline, nearly a year. It took me most of my free time over that year to design the amp, develop the algorithms, write the "C " code and get some of the Microchip tools to do something that they really weren't intended to do. After it was all working, it took about a month to document it all and prepare the contest submission. Then it took several months to write the article....I am a slow writer and change things a lot.
What did I get out of it?
The contest win netted me $750 USD.
At the time (2008?) Circuit Cellar paid $75 USD per page. The article ran 10 pages, so I got $750 for the article.
I got to know the Microchip guy and he gave me lots of free stuff, EVB's, a "C" compiler, and some chips.
I got to learn a lot about power supply modulation, DSP chips, and high efficiency low distortion tube amps. This technology was in its infancy when I did this, but it is now common in the solid state linear RF power amps used in cell phone base stations.
The article ran in Circuit Cellar, which is primarily all about imbedded computing. I doubt that to many of their readers were too interested in high tech vacuum tube amps, but the article did catch the interest of someone at Peavey, In fact it got me a few emails and phone calls from Hartley Peavey himself. We discussed this tech and its application to a guitar amp, and both agreed that as it existed, it wouldn't be a good fit because of the way it distorted when overdriven (brick wall clipping like a class D amp). Unfortunately I was too busy working my butt off trying to keep my day job to take it any further. The amp is in a box around here somewhere.
The magazine article used to be on Circuit Cellar's web site but it was removed a couple of years ago. I may put the details on my web site when I get caught up, but a picture appears a the bottom of this page:
Cathode Follower | Tubelab
Re this and also the Morgan Jones wincing comment:
If we are trying to make a Blameless tube amp, and we are allowed to use solid-state parts, then we are on a slippery slope. We go to a solid-state output stage to eliminate distortion from the OPT, then we find that the only thing limiting the amount of GNFB we can apply is the low gain of the tube driver, and before you know it all of the tubes have been optimized out of the design in the pursuit of Blamelessness.
To stop this from happening, some sort of arbitrary limit would have to be set on the amount of silicon allowed. I don't have a problem with silicon rectifiers or CCS. MOSFET followers... discuss...
Do you get to edit out all the fluff and carry-on sentences too? That's the other gotcha with writing, some folks can easily put their thoughts to paper in a very comprehensive and easy to digest manner, others just ramble on in not very coherent sentences and jump from one topic to the next without any apparent structure
I know the pain.
One time-eater is getting the illustrations up to pub quality. You'd be surprised how many pretty unreadable bitmaps I get. But over the years I've learned some tricks and collected some special software so that almost all illustrations in L|A are pretty good.
But I still shudder to see the bad quality in other often well-regarded journals.
Jan
I actually promised to post a scan on the other thread a few days ago.
Unfortunately I have a very large collection of books, jounals, magazines, and papers on electronics, and much of it is uncataloged, and not stored at my home, as I ran out of room. I've got about 3 rooms of stuff where I live, and another 3 rooms of stuff elsewhere.
So I have not yet had a chance to find and scan the thing.
In the meantime, I can tell you that if you compare Thiele's circuit with the Quad split load & push-pull driver design, you'll imediately see where he got his inspiration from. The main difference is that Thiele used more modern TV-type tubes and addressed a couple of subtleties.
If you look up Neville Thiele in Wikipedia, they give a reasonably complete listing of journal papers written by Neville Thiele, there's about 3 articles pertaining to audio design, but what I had in mind (complete amplifier design, consideration of recovery from clipping & cliiping behaviour etc) is not listed there.
Thiele was quite a noted engineer in Australia, prominent in television set design and aspects of broadcasting. Loudspeaker cabinets was actually a very small (pun not intended) diversion for him. Its quite stange that he became famous overseas for what was for him probably just a quick dabble.
Last edited:
IMHO the limit must be maintain the use of output transformers, and all the SS wonders as auxiliary (even if admittedly crucial) elements, not replacing any tube (the power drive thing isn't it in place of a tube, but the grid coupling caps, or not?)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.