John Curl's Blowtorch preamplifier part II

[wayne]

Originally Posted by PMA

The biggest issue is always to dampen transients that result from switching of rectifier diodes. Almost no one cares, though it is well measurable. If not damped, very short spikes get through the amplifier and occur at speaker terminals. On can find them by a fast digital scope in a single shot trigger setting.

These can show up very well on emission testing on the line also. I suspect many small companies don't measure this but I don't know for sure.
About $10K gets a decent EMI test setup for precompliance testing.

[Wavebourn]

What can be measured on a loop of shorted oscilloscope probe ends does not mean it went through the amp being amplified and exists on it's output. It is just an error of measurement that can be improved by fancy cables, ground lifts, other unrelated to sound quality means including super-duper diodes...

[scott wurcer]

John, I'll be out Wednesday and we will officially become neighbors. I'll be at Tilden Park at dawn for some golf you are welcome to join me. I met two 70 yr. old guys who carry their clubs and finish 18 holes in 2.5 hr. amazing. The geezer dawn patrol. Unlike Mayan ball we do not claim the head of the loser.

[Bonsai]

Short spikes caused by diode switching are coupled into surrounding circuits primary through radiation, not conduction. 500pf in series with a 470 to 1k resistor across each diode in te bridge rectifier fixes it. Also, use fast diodes with soft recovery characteristics - also helps here.

[Bonsai]

If you feed your power supply through a 50 ohm resistor and then sweep the frequency over say 1k to 5MHz, you may well be able to pick up the resonances. Another option is to feed in a square wave and look for ringing. 1-2 volts is enough and do it directly into the cap bank- ie after the rectifier. If you have a network analyzer, even better. Just for reference, 1cm of PCB trace is c. 10nH, so there's plenty of scope for problems.

[john curl]

Thanks for the invite, Scott. However, I have not played golf for over 50 years. Have a good time, and sometime we will meet together with something more compatible. It is pretty up there.

[VladimirK]

Originally Posted by Bonsai

Short spikes caused by diode switching are coupled into surrounding circuits primary through radiation, not conduction. 500pf in series with a 470 to 1k resistor across each diode in te bridge rectifier fixes it. Also, use fast diodes with soft recovery characteristics - also helps here.

One should also bear in mind the leakage inductance of trafo and how does it involved at diodes switching. Some RC is needed between the AC terminals of the diode bridge. Also, I prefer to build external PS for avoiding electro-magnetic coupling with sensitive parts of the main circuit.
If switching spikes are not suppressed, PS can induce some buzzing sound being approached close to speaker (EMI interaction with driver coil), not saying about EMI effects if PS is inside the amp. Some EMI is produced by the output sections of amp, therfore I prefer SE shunt-like output stage that has no variation of PS current consumption under signal.

[Bonsai]

Yes - it can cause ringing. This is a big issue in SMPS also where the leakage inductance is much lower, so the ringing is at higher frequencies.

[ThorstenL]

John,

Originally Posted by john curl

A typical Parasound Amp power supply input.

I would note that if one took care to design certain features directly into the mains transformer (including leakage inductance maximisation/compensation) the schematic would look the same (unless we show a detailed equivalent schematic of the transformer) and the power supply filter may be build on a PCB, with precise, designed in trace resistance and inductance filtering and damping resonant tanks.

Sadly I do not remember seeing anything like this in Audio Gear (well, except stuff I design)... This is one area where serious use of industrial methods can give serious improvements with what would seem a generic circuit.

Ciao T

[john curl]

T, I am not in charge of the transformer design of Parasound Amps, so I can't do much, one way or another. However, this design can be improved in other ways. This is next.