Tube Amps / Music Electronics
|For current discussions, please visit Music Electronics Forum.|
|5/4/2000 4:13 AM|
||Preamp Design Question...Confusion|
OK, so I'm playing around with a preamp design on a test rig, looking at load lines, and generally confusing the hell out of myself.
Here's the confusion: If I want to connect two gain stages together in the "standard" manner, what's the correct way to do it? I know plate of the first stage to grid of the second stange, but my confusion starts when I look at the load lines. Say I'm driving the first stage input with a 200mV peak-peak input (using a 12ax7 here) and I'm seeing a 40V peak-peak output (yeah, I can get more gain out of a 12ax7, I know, but I'm using an unbypassed cathode currently). Anyway, can this 40V peak-peak signal be applied directly to the next stage, or does it need to be attenuated to get it down to a level more in line with the available grid swith w/o driving the tube too far into saturation or cutoff?
Of course to get some distortion, some drive into saturation/cutoff is desired (isn't it???), but do I really apply that large a signal? What about after the second stage, which I'm planning on applying to the phase inverter?
I seem to be missing something fundamental here. Any help appreciated!!!
Incidentally, as most of you probably know, don't rest your hand on a tube amp transformer's HV secondary...I've never felt my teeth go numb before (except at the dentist)!! Not something I want to repeat!
|And now, a word from our sponsors:
|5/4/2000 5:46 AM|
Mention "load lines" again and I'm outa here! But I just wanted to suggest that you look at some schematics to see how many of the amps have been designed. For a high gain monster, look at the Soldano SLO100 (which was evidentally the inspiration behind the Boogie DR and the PV 5150EVH amps). In any case, between many of the stages most of the signal is thrown away, with the overdrive and distortion built up gradually over several stages. Less gainy amps will use fewer tube stages, with less of the signal discarded. Work your way down to a tweed Champ and with just a few tubes almost all of the signal ends up at the output. Each of these amp designs have their positive and negative aspects... probably best to have a few of each!
Just my very unscientific answer to your question...
|5/4/2000 2:04 PM|
chris, one thing i may suggest to you that proved invaluable to my own learning curve is this:
make up a test rig, and use linear taper pots (rigged as rheostats) for the cathode resistors and plate load resistors, and use them wired as potentiometers between stages. if you've got the pots, sprinkle a few here and there as rheostats providing shunt impedance to ground. these are best placed in the grid circuits.
in this way you can adjust cathode current, plate voltage and load, grid signal voltage, and grid circuit impedance "on the fly." of course, be careful when you use your fingers!
this manner of empircal learning taught me more than days, MONTHS even, of hardcore book learning--and i'm not afraid of books either. don't worry about "proper loads" and loadlines--we're talking about deliberately screwing with the operating points/impedances in order to GENERATE distortion, preferably of a pleasing nature.
what i've found is that you can pump an AMAZING amount of ACV into a grid, as long as the grid circuit impedance is low and the time constant is fast. that's how you really get the distorion/clipping.
|5/4/2000 2:14 PM|
Yeah, this is a good suggestion...it's what I finally started doing last night after not finding what I needed to know from books, simulations, etc.
I'm gonna go pick up some stuff today to build a better test rig (using an amp chassis now...not really very stable!). What happeded to that pic of RG's (I think) test rig? I want one of those!
|5/4/2000 2:19 PM|
I know this doesn't answer the main question, but isn't that a gain of 200. What am i missing here?
Some of the things i've noticed in higher gain amps are smaller coupling caps along with some way of dumping some of the highs - small cap to ground, or small amount of local feedback from the plate to the grid ala the Dumble 'smoothe switch', or maybe a small cap in parallel with the plate resistor. Marshall used 470k resistors going to the grid of the next stage to lose some signal...i believe this resistor will also cut some of the high end....Please correct me if i'm wrong on any of this.
|5/4/2000 2:40 PM|
Woops...meant to say 4V peak-peak output.
|5/4/2000 2:57 PM|
Common in marshalls is having a divider
between stages, 470k connecting
plate to grid, with 470k to ground,
dividing the signal by 2, BUT there is a
small cap (say 470 pF) across the
first resistor, so this arrangement
actually reduces the bass, not the treble.
|Page 1 of 1|