They did NOT measure well by 20 kHz THD. They showed a lot of crossover-like distortion when the dumpers kicked in, easily visible with THD-20.
Cheers,
Bob
I think QUAD were smarter than that. There's a class A amp that supplies output current untill the dumpers come on. The additional output current of the dumpers increases the feedback factor (beta) in such a way that the feedback factor is constant with dumpers off or on. So there is no dead band, no gm doubling or anything like that. VERY smart. And don't get me started about the choice of the L and C in the bridge to take care of stability.
jan
There's a lesson in there for all of us. People tend not to value what they don't pay for.
I couldn't agree more, and with the following comments about performance in the vicinity of 20kHz. I thought they sounded just terrible.
Hi John,
Agreed, that is, only at high output power levels. At 20 kHz the dumper is too slow to cope with the necessary slew rate and the class-A amp has to little power. I told Peter Walker (many years ago) that I have 'issues' with his concept, but he answered that you will never need full power at such high frequencies (and high slew rates), as such signals could never be produced by vinyl and the best cartridges you can get.
Needless to say that I wasn't really convinced by his arguments.
Cheers,
E.
Hi Gerhard,
Looks interesting, but could you show us some specs, please?
Cheers,
E.
I would like to point out a few things about the Quad 'current dumping' circuit. It has both good and bad points.
Its good point is that it can reduce distortion in an amp, even including a class B load line by perhaps 20 dB better than just plain negative feedback may do it, with the same stability.
The problem with Quad's approach with the circuit is that they used it to get away with making a VERY poor, unbalanced circuit design, with the cheapest components possible, especially because, in the middle 1970's SMPTE IM distortion ruled as a test measurement. Harmonic distortion was used less, in those days, because ST HD analyzers were still not generally available, in 1975. Sure, Bascom King and Mark Levinson had one, but I didn't, for example.
SMPTE IM, with its forgiveness of high frequency distortion, including slew rate, the measured distortion of even the original Quad 405 was not too bad, in fact, it was pretty good.
Later, with the advent of low distortion ST HD analyzers, the Quad 'problems' became more obvious.
Its good point is that it can reduce distortion in an amp, even including a class B load line by perhaps 20 dB better than just plain negative feedback may do it, with the same stability.
The problem with Quad's approach with the circuit is that they used it to get away with making a VERY poor, unbalanced circuit design, with the cheapest components possible, especially because, in the middle 1970's SMPTE IM distortion ruled as a test measurement. Harmonic distortion was used less, in those days, because ST HD analyzers were still not generally available, in 1975. Sure, Bascom King and Mark Levinson had one, but I didn't, for example.
SMPTE IM, with its forgiveness of high frequency distortion, including slew rate, the measured distortion of even the original Quad 405 was not too bad, in fact, it was pretty good.
Later, with the advent of low distortion ST HD analyzers, the Quad 'problems' became more obvious.
I can't find the original application note, but I am pretty sure that comes from 1969, and is part of an 'FET APPLICATIONS' article that I first saw the 'folded cascode' It was composed of a jfet for the input stage and a bipolar transistor for the folded cascode. I think that it had a video application, first. I did not start using it, until I realized that I could cancel its inherent 2'd harmonic by making a parallel complementary topology and then sum them together at the output.
The complementary differential folded cascode was made popular by Charles Hansen in his designs. I noted this approach for the second stage of the Vendetta Research second stage, in the late '80's, but I believed in negative feedback back then, more strongly, and I removed it from consideration, because it would not give me enough forward gain, without adding a complementary follower output stage, which I wanted to avoid. However, with Charles Hansen's success with it, over the years, I finally put it in the Blowtorch line stage, with some success, and have recently striven to make a quieter version for a future Vendetta, should I choose to make one. Tough problem.
The complementary differential folded cascode was made popular by Charles Hansen in his designs. I noted this approach for the second stage of the Vendetta Research second stage, in the late '80's, but I believed in negative feedback back then, more strongly, and I removed it from consideration, because it would not give me enough forward gain, without adding a complementary follower output stage, which I wanted to avoid. However, with Charles Hansen's success with it, over the years, I finally put it in the Blowtorch line stage, with some success, and have recently striven to make a quieter version for a future Vendetta, should I choose to make one. Tough problem.
Hi John,
Not quite true about the use of THD and availability of good distortion analyzers in the 70's. While it is true what you said about SMPTE IM being too easy on amplifiers at high frequencies, THD as a measurement tool was well-entrenched before the 70's. It is probably true that not all folks were diligent enough to specify THD-20, but we cannot do much about that.
The Sound Technology ST1700 came out in 1971, after ST was founded by some HP guys (who probably did the HP339A). The HP339A was good down to below 0.002% - more than enough to expose the flaws of the Quad design).
Cheers,
Bob
Mr Curl,
excuse me for coming late to the party, but which manufacturer/type were those ?
(shame you didn't get round to the BA, i was hoping to invite you for a hooters&booze arm wrestle)
- Status
- Not open for further replies.