I can't guarantee it, but historically, I think Prof Marshall Leach is the culprit here. In his early text on his first amplifier, in the mid-70ies "Audio" magazine, he used an 8 Ohms load in parallel with 1 uF, with 10 kHz test waves, simply to illustrate that his amp was stable even under very adverse conditions.
After that event, everybody and their dog just had to prove worthy of the same load. My point is, I believe Prof Leach inadvertently used a value to prove just one, his own point, never stating that this was anything but random choice, a fact which was somehow later accepted as some sort of the norm.
Just one of those things which happen anyway, perhaps even without intent.
Indeed. After all, this IS diy
Or you get fancy and call it Model Capture.
jan
Yes, it is, but so what? With the kind of theory being flung about here, we might as well be a remote chapter of AES.
What strikes me as exceptional is that over the last month or so, I have learnt more than over the last 5 years. True, this is not my profession, strictly speaking, but again - who cares? Knowledge has no second location (beside one's head), no colour and no price.
Yes, we have turned a bit towards modelling in simulators, but I wish to God somebody started the subject like 5 years ago. In my view, there is NOTHING to compare with than a running process of modelling, commentary and changes/improvements - that's like learning on steroids. A few of John's comments are priceless to me, and Thorsten's project is icing on the cake.
It's good to run into a power packed crowd like this one, it helps us (well, myself at least) to keep our feet on the ground and avoid flying out into airless space. Theory of (social) systems teaches us that without a negative feedback, we do not have a functional system. Here, we get all that, plus spiced up with different ideas, perhaps not world shattering, but most useful anyway.
Hi,
Sure. The problem is, while it is described in the RDH (and many other books that share it's focus in other languages) , in this 21st century most EE's (and DIY'ers) do not even understand how transformers work or how to make measurements as to their behaviour.
With a bit of ingenuity even very primitive means can provide a lot of knowledge how a transformer behaves. FOr example, to see what it does with DC I rectify mains (beware, this is VERY dangerous) and use a big power resistor from the resulting DC, which incidentally also forms my known element in the AC test using a battery powered and isolated generator.
I cannot recommend doing this to anyone SO DON'T TRY THIS AT HOME, because, if you do not know what you are doing the voltages and currents available will kill you if given a chance (then again so does crossing a busy high speed road at red light)...
So yes, this is DIYAudio. I recommend the old RDH (it does float around as free download, I am unsure of the legal side) for information on many basic techniques for DIY Electronics and serious information on many parts we nowadays tend to take for granted...
Like, is there any basis for claims that transformers of identical wattage and voltage specification can produce sonic differences, just by being Torroidal/EI/M/LL or Double-C-Core*?
Ciao T
* Easy hint, leakage inductance and resonance frequencies as well as general leakage with frequency are widely variable with construction...
Indeed. After all, this IS diy
Or you get fancy and call it Model Capture.
Sure. The problem is, while it is described in the RDH (and many other books that share it's focus in other languages) , in this 21st century most EE's (and DIY'ers) do not even understand how transformers work or how to make measurements as to their behaviour.
With a bit of ingenuity even very primitive means can provide a lot of knowledge how a transformer behaves. FOr example, to see what it does with DC I rectify mains (beware, this is VERY dangerous) and use a big power resistor from the resulting DC, which incidentally also forms my known element in the AC test using a battery powered and isolated generator.
I cannot recommend doing this to anyone SO DON'T TRY THIS AT HOME, because, if you do not know what you are doing the voltages and currents available will kill you if given a chance (then again so does crossing a busy high speed road at red light)...
So yes, this is DIYAudio. I recommend the old RDH (it does float around as free download, I am unsure of the legal side) for information on many basic techniques for DIY Electronics and serious information on many parts we nowadays tend to take for granted...
Like, is there any basis for claims that transformers of identical wattage and voltage specification can produce sonic differences, just by being Torroidal/EI/M/LL or Double-C-Core*?
Ciao T
* Easy hint, leakage inductance and resonance frequencies as well as general leakage with frequency are widely variable with construction...
Aye, aye, Captain! Attached below.
Well, here it is – the LAS Mega power amplifier.
I read a review about it in a German magazine (forgot which one) sometime in 1978, I think. Aside from being acclaimed as an excellent sounding amplifier, I was really enachanted with what I saw as its sublime simplicty, yet not missing a trick.
Note that the voltage amp uses electronically stabilized power line, set at +/- 55V. I no longer have the stabililizer’s schematic, but it was a fairly simple 5 transistor circuit for each line. It was a single regulator for both channels. The power amp was fed +/- 50V off two 15,000 uF capacitors – even at thta time, I thought this was a little lean for such an amp.
You will no doubt note the output stage, which is a little unconventional, while still being a well known arrangement. For a start, note that the power devices are BD 249/250 C. This is TI’s European version of TIP 35/36 C, the 100V version. Nominally, there are also „D“ (120V) and „E“ (140V) versions, but they exist only in the catalog, and the same is true of the BD models. However, I have never actually seen them being offered anywhere.
While nominally the same old same, there are in fact differences between these two families of devices. The really important one, according to TI’s Power Semiconductor Datasheet book, is that the BD series have significantly shorter Ton, Tstore and Toff times – for example, 35C has a Ton time declared as 600 mS, while the BD is declared as 450 mS.
Both share TI’s prodigious current capability, as both can deliver 25 A continuously and 40 A in short bursts (SGS-Thomson’s version of 35C can do all of 50 A!). However, because of their low operating voltage, they are not really very suitable for rails over +/-40V, so the choice was seemingly based on availability, relatively low price and extreme current capability. Their Ft of 3 MHz minimum is about the standard of the day, although at that time, the first crop of Sanken’s RET transistors appeared in Pioneer and Sansui models.
The protection circuit is ALL about this. Note the extraordinarily large 1.5 Ohm/15W emitter resistors, you don’t see values like that every day. As explained in the magazine text, this is no idle chance, but reason – there is a time sequence of 200 mS when the output stage is limited by those resistors only. This is ensured by R7, R8 and C6, and R30, R31 and C7. Nominally, they have been worked out to stay silent (passive) for about 200 mS, and then to start stealing drive current. On the other hand, the manufacturer claims that the ciruit will almost instantaneously cease operation if the reason for the large current limiting ceases to exist.
Also interesting is the use of germanium diodes AA132 (D5, D10). The manufacturer claims that they serve as fast recovery diodes from large overloads, cutting the recovery time to 1/10 of what it would be without them.
For the rest, I’m quite sure you can find your own way. To me, this is a piece of art, not electronics, I find its extreme simplicty completely enchanting. Perhaps I do so because over the years, I have made 3 samples, and I can attest to the fact that it really sounds wonderful, although by current standards, it has no right to do that. Its open loop bandwidth is just 4 kHz, hardly high tech, but you need to remember that at the time, many competitors (e.g. Yamaha) rarely went over 1 kHz OL.
The only technical complaint I have is that despite using a current mirror, I would still have preferred to have a method of reducing its DC offset via a trimmer somewhere.
Its specifications were:
Frequency response at 1W/8 Ohms: 5 ... 1,000,000 Hz -3dB
Slew Rate: 100 V/uS
THD and IM: < 0.01% into 8 or 4 Ohms (unspecified otherwise, but assumed to be at 1 kHz)
Nominal power: 150 W/4 Ohms, 100 W/8 Ohms
Damping factor: > 300:1 into 8 Ohms
Input sensitivity: 1.4 V (+6 dBV) for nominal output.
A big problem here was the BFT28A transistor. Eventually, it turned out it was made only by a Californian company and was as rare as wisdom teeth coming out. I never managed to get hold of it, and The Good Lord knows I tried hard. The RCA 40264 was also hard, but attainable. Instead, I substituted both with the much easier and very similar 2N5415, generally quoted as a replacement for the BFT28A, or even better, 2N5416 (Ft not 15, but 50 MHz).
I didn’t much like this, because I believe in respecting the designer’s requirements in full, but I really had no choice.
The values of R17 and L1 were not given, so I used 2.2 Ohms for R17 and 2 uH for L1.
The only problem is that I cannot remember what was the designation of Q15. I drew this from memory, but knowing that its complement Q5 was 2N3551, it shouldn’t be too hard, and for me, it’s been like 15 or 16 years since I made the last version, so in plain language – I forgot.
What I got from this amp, beside its really very, VERY good sound, were two things.
It taught me that the everything else notwithstanding, the BD 249/250 C were almost ideal transistors for power amplifier output stages. If a transistor can have its sound, then this one would have a warm, relaxed sound I have to date not found in other transistor, only the MJ 21195/21196 approaching that, but not quite making it.
It also gave me its outstanding protection circuit. Ever since I saw it, I have been using it and I never once had a hitch. It’s easily adapatable to whatever you like using standard calculations. Its key trick is that it really is NOT invasive if set up well, thus kicking out the false fame of protection circuits being bad for the sound. Anything not done well will be bad for the sound.
I did something similar around 1990 with the Harman/Kardon-branded aftermarket-automotive TC600, which were required by the internal customer to have no protection action for a specified full-power 60-degree reactive-load tone burst at 20Hz, but of course also not to blow up into sustained short circuits. These amps, the design of which I sort-of inherited and was admonished not to change (if possible) had rows and rows of the big Toshiba TO3P devices.
The amusing thing was the "kids" who were buying big amps for their show cars disdained the TC series because, with its signal and temperature-gated-fan "transverse tunnel cooling" didn't look impressive enough --- most wanted massive heat sinks. This also totally ruined the Infinity-branded switchmode amplifier sales, a product that was bravely announced every CES for years running
When it finally worked, it was expensive and being so compact, nobody wanted it.You could be right. Hard as I tried, I couldn't find out anything about the original company Linear Audio Systems (LAS), other than it was from somewhere in Germany.
God alone (if even He) knows who these people were, but all the same, as I said, I regard this one as a piece of art of the times.
It is of particular interest to me since I have been very much in touch with the German audio industry while it existed in the 70-ies. This model is a true paradx, so very unlike other German products of the day (and even today), yet so very German in its think - cool, clear, focused.
Look at what people like Telefunken, Grundig, Nordmende, Elac, Studer/Revox, Loewe Opta, Dual, Perpetuum Ebner, Schneider, Koerting, etc were making at the time - many products look like they threw kilos and kilos of components in with a shovel (I refer to quantity, not the way it's arranged on the boards). One might think that the components alone cost more in retail than the whole unit - and one wouldn't be far off the mark.
This one is in complete opposition - as simple as it was possible to make it without any serious compromises, yet totally German in the logic behind it. I feel as if many years of experience were built into it, with much common sense and clear thinking. In terms of current delivery, they were at least a decade in front of the times.
Who in those days had regulated power lines for the voltage amp? Not many, let me tell you, in fact, VERY few. Sure, it's a compromise, it could all have been regulated, but I don't recall any Japanese products of the day, except for Kensonic Accuphase (the correct name of the company in those days), which in turn did cost at least twice the price of this one.
I'm ranting, I better stop now. Anyway, the point is, I admire this design for its technical simplicity, yet full completeness. My hat off to the designers, even after all these years.
Last edited:
I did something similar around 1990 with the Harman/Kardon-branded aftermarket-automotive TC600, which were required by the internal customer to have no protection action for a specified full-power 60-degree reactive-load tone burst at 20Hz, but of course also not to blow up into sustained short circuits. These amps, the design of which I sort-of inherited and was admonished not to change (if possible) had rows and rows of the big Toshiba TO3P devices.
The amusing thing was the "kids" who were buying big amps for their show cars disdained the TC series because, with its signal and temperature-gated-fan "transverse tunnel cooling" didn't look impressive enough --- most wanted massive heat sinks. This also totally ruined the Infinity-branded switchmode amplifier sales, a product that was bravely announced every CES for years running
Brad, these are the days of bling, not quality.
Few people care how it actually works, but many care how it LOOKS like it might work, and of course, the price. These days, the ideal consumer price is when you as the manufacturer pay them $10 to take it out of the store. Total chaos and madness.
I wouldn't be surprised. Siemens has done some outstanding work over the years.
Perhaps you may recall an audio power amp project, based on their BUZ MOSFETs, published in Elektor in the early 80ies. It was done by a man signed as head of their production of something, I forgot what. It is memorable for two reasons:
1. It was scalable, taking you in very clear steps from around 30 to around 120 Watts of output power, in if I recall correctly 4 steps, with VERY clear explanation of how and why, and
2. In terms of design logic, not unlike the LAS amp - simple, functional, focused and very sober, no fiddling about, straight to the point.
I never made one, but a lot of my friends did, so I did have several chances to audition it. In simple terms, it was one of the very few MOSFET amps I did like and would not mind owning.
I also remember it to this day, almost 30 years later, because Siemens is not a name I associate with audio in general, except for their semiconductors; with Motorola/ON Semi and Toshiba, they are on my preferred list. When ordering parts, I specify Siemens as a requirement sine qua non.
Hi,
This one is quite interesting. It has many features that are thoroughly modern, I suspect what held it back a little was the availability of semiconductors in it's day.
I find it interesting that the compensation is precisely the same that Wahab complains can never be stable... ;-) And it has a heavily degenerated VAS...
I likes it!
The Class A + AB driver stage is also interesting. I'd probably replace input stage BJT's with cascoded J-Fets and turn the cascode into a hawesford one. And maybe replace the drivers with Mosfets...
Oooops, this starts to look a lot like what I posted previously....
Ciao T
This one is quite interesting. It has many features that are thoroughly modern, I suspect what held it back a little was the availability of semiconductors in it's day.
I find it interesting that the compensation is precisely the same that Wahab complains can never be stable... ;-) And it has a heavily degenerated VAS...
I likes it!
The Class A + AB driver stage is also interesting. I'd probably replace input stage BJT's with cascoded J-Fets and turn the cascode into a hawesford one. And maybe replace the drivers with Mosfets...
Oooops, this starts to look a lot like what I posted previously....
Ciao T
Hi,
This one is quite interesting. It has many features that are thoroughly modern, I suspect what held it back a little was the availability of semiconductors in it's day.
I find it interesting that the compensation is precisely the same that Wahab complains can never be stable... ;-) And it has a heavily degenerated VAS...
I likes it!
The Class A + AB driver stage is also interesting. I'd probably replace input stage BJT's with cascoded J-Fets and turn the cascode into a hawesford one. And maybe replace the drivers with Mosfets...
Oooops, this starts to look a lot like what I posted previously....
Ciao T
I thought you might like it.
I completely agree, their biggest problem was that they had a rather poor choice of quality semicondustors in their day, which was probably something like 1976 or 1977 Their choice of BFT28A was certainly very, er, exotic, until you look at its specs - its gain "curve" is really much more alike to a straight line, a bit like 2SC3281. Anyway, 2N5415 is a good replacement.
At the time, I don't think you could stick in a FET pair, then cascode it, etc, because quality FETs for audio were just beginning to appear in any quantity and choice, and initiall, they were marketed in Japana and USA mostly. If you remember, it took the Japanese a hell of a long time to become available in Europe, a fact I cannot explain even today.
Nevertheless, I am happy to have learnt a great deal from it, in fact, surprisingly much. For one thing, I discovered BD 249/250 C trannies, with which I experiemented really a lot through the 80ies, and I repeat - if there ever was a good sounding transistor for audio, that's the pair, their only limitation being the relatively low voltage. They are great for lower powered amps, with rails up to +/- 45V, or up to about 100W/8 Ohms, but that's it. But it is well neigh indestructible, much like Motorola's MJ15XXX and MJ211XX series, it seems that you need a sledge hammer to hurt those.
It therefore souldn't surprise you that Grundig built some of their top of the line integrated amps in the early 80ies using them at the output.
Here's a suggestion - in your spare time (which you don't have), try building a live model. I don't think the sum total of all parts for two channels would cost more than say €200, but remember - no Elnas, those guys used Roe and such like. I do believe you would be as fascinated with it as I am. Technology, concepts, etc is nice, but the final proof is in the sound, and it's there that this one excells. Like I always say - it plays much better music than it has any right to, just looking at the schematic.
Or wait for me to get around to it, I've been promising myself that for years and somehow never managed it. But I will.
Now, for my next act, I'll show you guys how my beloved 1978-1980 Marantz brothers were made - US design work, I don't know who did it, but whoever it was, my hat off to him with a deep bow, that was intelligence and experience at work. Resulting in truly outstanding sound. I should know, I have a small collection of them, and I'm not in the habit of wasting money. All but one fully refreshed and readjusted, the last guy is just waiting for his turn.
Last edited:
Hola T,
Of course you would like it , it's a scrawny puss of an amp.....
A comment worthy of Megawatt Wayne.
I wonder how long it will take to make you realize that quantity is meaningless without quality, that quantity is not necessarily equal to quality, and that its much better for your ears to have a 50 W quality amp than a 250W piece of highly strung up junk.
I wish I could have you over for coffee, beer, lunch and auditioning, I'd look up a Naim integrated amp from the early 80ies, rated at just 30W/8 Ohms. I'll wager that 30 minutes with him would make you rethink your position, because that's the model which probably did the most for Naim as a company, and with good reason. It ran off fully electronically stabilized power supplies, the actual model is audio electronics schematic from the first year of a colledge, child's play it would seem, yet it blew away the vast majority of Japanese commercial amps rated at 100 W in terms of sound QUALITY. Of course, those were the days when the late Julian Vereker, Naim's (and partly Linn's) founding father ran the show, and you can earn more about his views of music from Naim recordings he made - quite out of the ordinary, way above it.
I am not saying this is in any way typical, just that it is possible, power alone accounts for very little. When I got my current speakers, years ago, I took the time and trouble to twist some arms and have some high class pro gear transported to my room, because I was honestly very curious to know just how much power was I really using. And those sexy looking blue back lit analog needle power meters on the Marantz 170 DC were really for show only, anything but very reliable. Anyway, I learnt that about 90% of the time, I was using less than 1 Watt, and in peaks that amounted to 10...15 Watts. As expected, those peaks were not really much louder than the rest, but did require a lot more than average energy to be reproduced as they should be; mostly things like tympani, and the taiko drummers session, the most drawn by the Blue Man Group, first audio CD, song No.8, when a guy runs up and strikes a drum with a diameter of 8 feet. That piece quickly swamps like 99% of tube amps, they just choke on it. But then, so do many solid state amps as well.
Hi,
You know what?
On any funloving righteous speaker a REAL MAN (not a housebroken, pussywhipped "new man") would care to use this is a double sword fighting, evil megawatt amp busting "Puss in Boots" that is, with Kitty Softpaws thrown in for extra sensual pleasure.
Big Megawatt Amp's are needed as compensation only by those who are sorely under-endowed, in the speaker department of course... ;-)
Ciao T
Of course you would like it , it's a scrawny puss of an amp.....
You know what?
On any funloving righteous speaker a REAL MAN (not a housebroken, pussywhipped "new man") would care to use this is a double sword fighting, evil megawatt amp busting "Puss in Boots" that is, with Kitty Softpaws thrown in for extra sensual pleasure.
Big Megawatt Amp's are needed as compensation only by those who are sorely under-endowed, in the speaker department of course... ;-)
Ciao T
DVV ,
Would love to , 50 watts per channel is fine by me , I had Dennis's 805B's, sounded fantastic on my mini's (speaker is similar in size to T's ) when quantity and quality match , Is when the experience takes on a different level ..
Big amps, big speakers , big room ..... SOTA is the goal ...
Would love to , 50 watts per channel is fine by me , I had Dennis's 805B's, sounded fantastic on my mini's (speaker is similar in size to T's
) when quantity and quality match , Is when the experience takes on a different level ..Big amps, big speakers , big room ..... SOTA is the goal ...
You have this wrong , it's those who say size doesn't count that's hiding something ...
- Status
- Not open for further replies.