to get this thread headed in a more fun direction and maybe provide some meaningless education by accident.......
Quote:
Melos used the tube filament as the current sink. Novel but......bass limitations.......limited voltage...yadda..yadda...(my adds the last) |
Have to agree with this one .I tried every conceivable load for what some call the Szekeres Amp but something i can no longer call by that name due to it resembling the original in Class-A mosfet stage only.
I had read an article in the tube CAD Journal about a hybrid amp where the heater was used in this manner and it struck me as a very cool solution at the time so being a simple thing i implemented this feature and then sat down for the listening,the actual evaluation by "ear' and not the theoretical what it SHOULD sound like.
I must say the end result was a very clean signal and some of the low end "confusion' was gone,but so was the GRUNT factor !
Anyone who has ever used Grado headphones knows one thing they do not lack for is eye blinking slam if driven correctly and played LOUD .
Probably the loud is not a good thing but this dog is too damn old to change so i like my rock beating my ass down and my blues or jazz smooth as silk.and both must have the note separation and inner detail to allow me to almost "see' the preformance unfold.
the heater as drain load was anemic at best and took from the music like taking away a full range loudspeaker system and replacing it with a mini monitor !
Yeah the music is there but where the hell is the rest of it ???????
Next up,light bulbs !
Yes ,i said light bulbs.This is not a new thing but the recent article by Nelson Pass ,the "ZEN Lite" modification of the original ZEN Amp was something that again punched my "cool" button.
this "mod" was a big time improvement over the heater as load resistor but was not exactly "correct" in this implementation .i used refridgerator bulbs (wife as she rooted around in the dark "what happened to the bulb in the refridgerator"-the rick with the long practiced innocent look married men get over time "jeez honey ! I have no clue ?!? It's not there ?")
Better but there was a loss of control at the bottom end.unlike the heater option which took from the bass this one left it there but the bass itself was flabby and uncontrolled.The apparent "speed" and the feeling of tightness was gone and in its place was what i would describe as overhang if it were loudspeakers being discussed .
Back to the 5W/50 ohm Carbon potentiomenter .Pretty damn good sound ,a real bi*ch to get the balance right between the two channels without continous current monitoring.the drift was a killer !
So here i was,not happy with the resistor ,tried the tube heater and light bulb filament option,again not pleased,the variable resitor was a pain but sounded good but only running at 300 ma per channel,the mosfet/transistor/LM317 option while having bass to die for (really tight,fast) butchered the top end.......
DAMMIT !
all the while i am using every conceivable front end from 6GM8 tubes to the 6CA7 (damn good actually),op-amps,transistors,various tubes,FETs..........
And still not happy !
I decide maybe an inductor might be the end zone but i will be damned if i am going to purches the parts and then if they suck have them sitting on my bench laughing at me with no actual job to preform !
So on a whim i took a roll of magnet wire ,yes the entire damn roll from radio shack , and plugged it in.
Not bad ! Not bad at all !
But not yet "there" so i procedded to remove wire and listening between removals until i "dialled in" the sound and it has stayed that way since !
So now this is out of the way but something is yelling at me to bump up the power supply voltage.Everything i know about the FET/Mosfet family,little as it be,is that they like two things
1-class-A operation.not"quasi-class-A" but full tilt idle current=max output current and nothing less .
2-these suckers like the juice as in GIMME SOME DAMN VOLTAGE !
running a MOSFET at +12V or +/-6 Volts is STARVING THE BEAST !!!!!!!!!
You want power ? Feed it !
Hit that sucker with some real he-man voltage and current and be prepared to listen to a new thing entirely !
Starve it and you are spitting in the wind.
so as Kevin stated ,the heater load+the seriosly low voltage supplied to the mosfet/darlington limits where this output stage can be but if you massage it then beat it over the head with some real juice !
Dr.Jekyl becomes the Hyde Monster !
And if i remeber the Melos biasing correctly from memory my first thought was "over done ! Way too complicated to get to the finish line intact" and many areas where things could go wrong.
My philosophy has always been to use what is required to get the end result then toss the rest in the trash as a good idea on paper gone bad.
And that includes building for specs unless you are willing to actually check the harmonic structure all the way into the MHZ range (including the power supply)
So long story short :
Doin the John Swensen/Kevin Gilmore/Melos hybrid gainstage/headphone amp as posted above is somerthing to take under serious consideration.
I HAVE the tube gain stage right here ,i HAVE been down every conceivable path with the Class-A MOSFET Buffer stage and if you get the power supply right the schematic posted will whip some serious ass PLUS give you the ability to vary the volume while changing NOTHING else !
that last is no small feat.The phase margins ,impedance relationships,bandwidth-all are "messed with" when you add a simple volume control to the mix.
A VCA suffers from none of the above if your design is tight
and the servo is not feedback in the classic sense.not even close
longwindedrickmonsterator