Craig Sawyers

I've just sprung for the last pair of Justin's KGSS boards and a 50k 4-gang DACT. It'll be a while before I get it finished though, although I've already got most of the bits - but not transformer or casework.

Last series Chris Evans, a British radio DJ was on as the star in a reasonably priced car. He's stupidly wealthy having bought a company from Richard Branson and then sold it for a massive sum of money. Although he looks like a real plonker, he's a very savvy businessman. He went to a car memorabilia auction a few years ago intending to buy a signed photograph of some famous racing driver - but could not resist the temptation of buying James Coburn's 250GTO for a hammer price of £12m. Now that is serious folding money.

Jeez yes. That was a Jaguar XJ220 engine. 3.5 litre V6, twin Garrett turbos, 540 bhp and 475 lbft torque. Must have come out of a wrecked XJ220. Only 280 ever made, a fair number must have ended up wrapped round a tree in the last 17 years.

Not sure if this is the right list for this - but head case seems to be dropping way down the list of hits on Google - certainly here in the UK. Currently 6-7th hit, used to be second.

Yup - as Richard Feynman once said when a student knit-picked during a lecture "Don't listen to what I said - listen to what I meant"

That is exactly the way to go - and for sure is the way to remove interconnect cable dependence from sound quality. Kevin, my man - while waiting with bated breath for your KG-Blowtorch, I'm embarking on a Borbely pre with K216/J79 outputs that drive 50-ohm matched impedance. For precisely that reason.

I can see the confusion - both channels were identical, both oscillating with around 10V p-p at 650kHz wrt ground on + and - sides, even with the 5p capacitors across the 100k feedback resistors. The signal is in-phase, so the signal wrt bias is much lower. So at 650kHz, the gain is greater than unity when the closed loop phase shift gets to 180 degrees. And although the problem has its genesis in the frontmost EL34, it couples equally to the rearmost EL34 as a result of the way the circuit is both symmetrical and cross coupled. I think that it is some subtlety of layout. Stax clearly had the same sort of problem, which is why there is a 2p capacitor lurking in there. The physical size of any layout of this monster leaves it prey to any number of parasitics. I went for simple dominant pole compensation, although there are more advanced ways of achieving the same ends.

Now that is interesting - post 650kHz mod above, it now runs cooler. Previously, the temperature rise at the hottest part of the heatsinks was 28 - 29C above ambient. Now it is 24C above ambient. The only reason I measured it is that it felt noticeably cooler to the touch. Measured using a Fluke 87V DVM with thermocouple sensor.

Well, the K389 has a Ciss of about 30pF. Logically that is about right, since it is a low noise small signal JFET. So to roll off at a 3dB point of 25kHz would need a series resistance of 200k.

In my continuing quest to get rid of the 650kHz oscillation of the T2, I found that 5pF across each 100k feedback resistor did not completely sort the problem out. Currently, I have repositioned one of the 100k resistors (to the rearmost EL34). It is now connected diagonally directly between the anode and the 100 ohm, and cut the track at the 100 ohm. Removed the 2pF cap. Put 10pF across the remaining 100k resistor. Capacitance across the repositioned resistor had very little effect. I tried repositioning the other 100k (ie the one that now has 10p across it), but it had no significant effect. With that final combination there is no sign of instability, and the output noise measures <2mV rms in 20kHz bandwidth (ie 110dB below full output). Hum is about the same amplitude as the noise. I don't what to sign off on that mod until I've listened to it for a while and also measured the transient response.

Eh? You'll have to speak up, I'm not wearing my sculptedeers

That is the Brit in me - we apologise for everything, in many ways. Mine was the smart ass way of saying sorry

Maybe its an age thing - when you get to my almost walking frame age actual lectures are a dim and distant memory

And here's one for those of a mathematical persuasion Wherein it is related how that polygon of womanly virtue, young Polly Nomial (our heroine) is accosted by that notorious villain Curly Pi, and factored (oh, horrors!). Once upon a time (1/t) pretty Polly Nomial was strolling across a field of vectors when she came to the boundary of a singularly large matrix. Now Polly was convergent and her mother had made it an absolute condition that she never enter such an array without her brackets on. Polly, however, who had changed her variables that morning and was feeling particularly badly behaved, ignored this ignored this condition on the basis that it was insufficient, and made her way amongst the complex elements. Rows and columns closed in from all sides. Tangents approached her surface. She became tensor and tensor. Quite suddenly, two branches of a hyperbola touched her at a single point. She oscillated violently, lost all sense of directrix, and went completely divergent. As she reached a turning point, she tripped over a square root that was protruding from the erf and plunged headlong down a steep gradient. When she rounded off once more, she found herself inverted, apparently alone, in a non-euclidean space. She was being watched, however. That smooth operator, Curly Pi, was lurking innerproduct. As his eyes devoured her curvilinear coordinates, a singular expression crossed his face. He wondered, was she still convergent? He decided to integrate improperly at once. Hearing a common fraction behind her, Polly rotated and saw Curly Pi approaching with his power series extrapolated. She could see at once by his degenerate conic and dissipative terms that he was bent on no good. "ArcSinh!" she gasped. "Ho, Ho," he said. "What a symmetric little asymptote you have. I can see your angles have lots of Secs." "Oh, Sir," she protested, "keep away from me. I haven't got my brackets on." "Calm yourself, my dear," said our suave operator. "your fears are purely imaginary." "i, i," she thought. "Perhaps he's not normal, but homologous." "What order are you?" the brute demanded. "Seventeen," replied Polly. Curly leered, "I suppose you've never been operated on." "Of course not," Polly replied quite properly, "I'm absolutely convergent!" "Come, come," said Curly. "Let's off to a decimal place I know and I'll take you to the limit." "Never!!" gasped Polly. "Abscissa!!!" he swore, using the vilest oath he knew. His patience was gone. Coshing her over the coefficient with a natural log until she was powerless, Curly removed her discontinuities. He stared at her significant places, and began smoothing out her points of inflection. Poor Polly. The algorithmic method was now her only hope. She felt his hand tending toward her asymptotic limit. Her convergence would soon be gone forever. There was no mercy, for Curly was a heavyside operator. Curly's radius squared itself; Polly's loci quivered. He integrated her by parts. He integrated her by partial fractions. After he cofactored, he performed Runge-Cutta on her. The complex beast even went all the way around and did a coutour integration. Curly went on operating until he had satisfied her hypothesis. Then, he exponentiated and became completely orthogonal. When Polly got home that night, her mother noticed that she was no longer piecewise continuous, but had been truncated in several places. But, it was too late to differentiate now. As the months went by, Polly's denominator increased monotonically. Finally, she went to L'Hopital and generated a small but pathological function which left surds all over the place and drove Polly to deviation. The moral of our sad story is this: "If you want to keep your expressions convergent, never allow them a single degree of freedom."

One night when his charge was pretty high, Micro Farad decided to get a cute little coil to let him discharge. He picked up Millie Amp and took her for a ride on his megacycle. They rode across the Wheatstone Bridge, around the sine wave, and into a magnetic field next to a flowing current. Micro Farad, attracted by Millie Amp's characteristic curve, soon had her field fully excited and he couldn't resistor. He laid her on the ground potential, raised her frequency, lowered her resistance, and pulled out his high voltage probe. He inserted it in parallel and began to short circuit her shunt. Fully excited, Millie Amp cried, "Mho, Mho. Give me Mho!" "Ohmigod, this is good," shouted Micro. With his tube at maximum output and her coil vibrating from current flow, her shunt soon reached maximum heat. The excess heat had gotten her shunt pretty hot and Micro's capacitance was rapidly discharging, ... draining off every electron. They fluxed all night, trying various connections and sockets until Micro's bar magnet had lost all of it's field strength. Afterward, Millie tried self-inductance and damaged her solenoid. But it didn't phasor. With his battery fully discharged, Micro Farad was unable to excite his transformer. So they ended up by reversing polarity, and blowing each other's fuses.

Usually an anechoic chamber is used with conductive or ferrite loaded wedges which looks like an infinitely large environment - so trying to simulate free field, especially when the anechoic is big so you can get a good distance between antenna and DUT. I've tried finding stuff on alternatives using rf reflective walls (ie bare copper), and drawn a bit of a blank.

You stand a pretty good chance with Kimber - Ray Kimber was a recording engineer before he set up the company, and developed the woven speaker cable to stop thyristor light dimmer crud from getting back up the PA amps, getting into the feedback loop and then coming back down to the speakers as a buzz. He'd be pretty keen to get immunity measurements as a quantitative selling point for his cables. Cardas too, I would think, MIT, Nordost and others. Martin Colloms (HiFi Critic) has been trying to get a cables consortium together, part of the aim of which is to substantiate subjective claims with real-world measurements.

The other good idea is to route all the power cables closely together. That minimises the loop area that could inject airborne RF interference into the ground, or radiate common mode interference that is coming up the power or ground lines. Might be worth trying an air cored toroid in series with the overall ground. That way you keep the protective ground connection, but now have a high frequency block to stop utility ground shit from getting up the earth. Make a wooden donut maybe 6 inches in diameter, and wind as much heavy gauge green/yellow wire as you can around it. Wire in some very safe manner in series with the overall system ground.

Well, that sure is a pair of ebony erections

Original (low bias) SR-lambdas were the same - the baffles and electrostatic units were a single glued assembly.

Although Bob did his own attenuator using the Goldpoint roll-your-own version (and so deliberately chose 1dB per step), other commercial switched attenuators like the DACT, and attenuator transformers like Sowter http://www.sowter.co.uk/transformer-attenuators.php are available in fine step versions. The DACT fine is 0.5dB per step(!) But I'd certainly go for the more standard Goldpoint or DACT set up, with big steps for the first few, then 2dB per step thereafter. I find 2dB per step works well for me.

Boy am I feeling dumb! Been totally fooled by where the emoticons were. Then just happened to notice the smiley on the reply toolbar. Then click show all on the popup. Of course you all knew that. I can sometimes be a little slow on the uptake

Ah - there's the problem. For your preferred listening levels, and source output voltage you definitely need attenuation to centre the -12dB point of your Goldpoint. And yes, I guess that the gain of the amp could be changed. But going to say 40dB from 60dB has to done with a little care to make sure it is still stable at the lower gain, or at least has not picked up any overshoot.

FWIW the T2-clone also has a gain of 1000. My D-A has an output of 3Vrms and that puts the DACT at pretty much centre travel at the typical loud level at which I listen. Takes maybe three clicks each way to compensate for recording level differences. Half way on a DACT is -28dB, and the steps are 2dB over most of the range.

Classic! James May in a 500hp go kart that the New Stig threw around the track faster than anything else they have tested. Richard Hammond in a 1 mile race between a VW Beetle and a Turbocharged 911 - the beetle having been dropped vertically from 1 mile and falling under gravity. Oh yes!!