Cascade 6V6 18Watt Tweaking Help Please

[ncusack]

Hello Everyone,

Been a little while since I've been around the forum but my latest build has me hunting for guidance. I recently put togeather a Liteiib with a cascaded preamp using 6V6's. Schematic with idle voltages has been attached to the post.

What I'm experiencing so far is too much gain and a decent amount of high end hash in the tone. The tone control is also not the most effective but I found a 5e3 tone mod that mentions lowing the 47n cap to ground to something like 7n to improve the high end roll off so I may give that a shot.

I'm wondering what suggestions you could make to reduce the preamp gain of the circuit. I know it's sort of a rule of thumb to try and bias the plates at half the B+ but I'm not sure of the best method to achieve that. Would it be better to adjust the lower for more current thus larger drop over the plate, Increase the plate for larger drop or try to reduce or increase the B+ itself in the early preamp stages?

Any and all suggestions would be more than welcome.

Cheers,

Neill

[eleventeen]

Offhand, and without conducting any sort of precision analysis, my comments would be:

The 2.2 Mohm resistors seem awfully large going into the 6V6 grids. Electronically, those resistors are hardly even there.

And the 47 pf coupling caps from the PI plates seem inordinately low. Those are .1's in most amps!

[ncusack]

Those are 2.2K going into the Grids and the 2.2M to ground are required for the LARMAR mastervolume to work. The 47p you mentioned are 47nF which I don't think is too far off the mark for most amps.

[sluckey]

I would first change the two preamp plate resistors to 50K. Then remove the 1µF cathode bypass caps if the gain is still too much.

You can also put a 470K voltage divider resistor between the 22nF cap and the top of the volume control, like what you see in the JCM-800 2204 amps.

[ncusack]

Ok so I did a quick swap out of the plate resistors and dropped them down to 56k and my voltages are now 277B+ preamp node. 228VDC at the first preamp stage plate and 1.876VDC across the cathode resistor and cap. 235VDC at eh second stage plate and 1.99VDC across the second cathode and cap. I wont have a chance to fire it up and test things until the weekend but do these voltages sound alright?

[sluckey]

Voltages are fine, but I'd concentrate more on how it sounds than the voltages on the preamp tubes.

[stratele52]

Remove the 22nf at the plate of V1 and 10nf at the output of V2 and try a 0.022 MFD.

Bypass cap at V1 and V2 are very low , try higner value ; 25MFD  ?

Plate's resitors had the good value, not less than 100K

[sluckey]

Remove the 22nf at the plate of V1 and 10nf at the output of V2 and try a 0.022 MFD.

22nF === .022µF

Bypass cap at V1 and V2 are very low , try higner value ; 25MFD  ?

That may make his marshall amp sound 'muddy'. And will also increase gain.

Plate's resitors had the good value, not less than 100K

He has too much gain. Smaller plate resistor equals less gain.

[stratele52]

Remove the 22nf at the plate of V1 and 10nf at the output of V2 and try a 0.022 MFD.

22nF === .022µF

OUPS !!!! SORRY

Bypass cap at V1 and V2 are very low , try higner value ; 25MFD  ?

That may make his marshall amp sound 'muddy'. And will also increase gain.

MUDDY ?  NOT IMO

Plate's resitors had the good value, not less than 100K

He has too much gain. Smaller plate resistor equals less gain.

YOU WON'T HEAR LESS GAIN

[dnator]

The cascaded 18 watt with 6V6s is actually really close to a Fender 6G3.
It uses a split load plate resistor setup on the second triode to reduce the gain.

http://www.el34world.com/charts/Schematics/files/fender/DELUXE_6G3.pdf

The 15k and 100k

[stratele52]

For less gain I'll put a resitor just before the 1 M volume , someting like 200K

[sluckey]

YOU WON'T HEAR LESS GAIN

I disagree. But since he's already changed them, let's let him decide.

For less gain I'll put a resitor just before the 1 M volume , someting like 200K

That's what I said.

[terminalgs]

PI to power tubes:

LARMAR or not,  the R2 (lower half) of those 250K pots are parallel with the 2.2M's, so the 2.2M's don't influence grid-leak that much. At full-tilt, you are a hair under 250K.  At 50%, you are two hairs under 125K.  (so I don't think the 2.2M's hurt anything).

however, I thought LARMAR was designed for fixed bias amps.  I don't think it does anything for you here.  and,, I think this general area of the circuit is where your bass response is disappearing.

the low frequency response of the PI to output stage is determined by the 47nf coupling cap, the 250K grid leak, and the output impedance of the PI (50K?).  

edit --- I was wrong on the following (now strike-thru), the 47nf always sees 250K resistance to ground to the right.  bass response shouldn't be effected by wiper position.

as you turn down the MV 250K, your bass-roll off frequency climbs.  So if you tried the amp with the MV set low,  you might have lost some/lots of your bass response.  turned up to "11",  you more/less have the low freq. roll off factors of a Marshall,  but as you turn down the MV,,  that  250K goes south,, and bass response disappears.


preamp:

I'm with sluckey on adding a 470K or a 1M above your 1M volume.  However, you can approximate the effect of such a mod by testing the volume at 50% without soldering.

usually, in a design like this, the 2nd gain stage is there to  make up gain for a complex tone stack that drops the Vrms of the signal so low that it needs recovery.  Where there are simple low-load volume/tone setups (like yours), the 2nd stage isn't used, signal is big enough to go to PI.  additionally, those 2 stage preamps usually have a 2nd channel that gets mixed with another channel (and thus attenuate everything).  PLUS the gain-increasing cathode bypass caps are usually on one gain stage or the other, not both.  PLUS no neg. feedback.  You don't have any of these components that reduce gain in other circuits.

So, yea "too much gain"..  you built a gain monster!!


Is there any combination of the volume controls that yield a decent sound?

[DummyLoad]

2.2M resistors under MV - they're for protection in case MV pot wiper fails - the power tube still has a grid-leak path in that event. personally, i like 220K-470K there with value dependent on tube type. LARMAR works with auto-bias power stage - use fewer parts; 2 caps and 2 resistors not needed. see decent example here  - weber is using 270K safety Rs.

too much gain? try removing second stage bypass cap. as also suggested; divider network(s) for first, second, or both stages.

use 500pF/5nF w/ 1M log taper pot. it's a decent 1 knob tone setup.

not enough bass? use 5-25uF for bypass of 1st & 2nd stages - IOW, use something larger than 1uF. what is OT freq. response?, what speaker/cabinet? add NFB?; all have some influence on bass response.

schematic of compilation of suggestions so far attached... :-)

--pete

[tubenit]

I'd take a different approach.   

I'd use a 12AY7 in V1 .............. & a 12AT7 or 5751 in the LTPI.  That will certainly lower your gain.  The 12AY7 that Doug carries is a really nice sounding tube in my opinion.  I seldom use a 12AX7 except in the LTPI anymore.

Then I'd use an "enhance cap" of 120p to 220p across the entrance plate resistor of the LTPI to lower the high end hash. 

With respect, Tubenit

[loogie]

Buy Merlin's books.

http://www.valvewizard.co.uk/

[printer2]

Rather than cascade the triodes how about do a parallel triode in the first stage as in the Spitfire or simple 18 Watts.

[jerrydyer]

why do you want a larmar on such a low watt amp.If you feel you need it just use the simple version like a cross line mv. better for that circuit
also a traditional MV works perfectly in these 18 watters. 1 meg pot before the phaze invertor. its going to sound way better and bigger without any of the 3

[DummyLoad]

for jason villa.


--pete

[davidwpack]

I agree with tubenit. I replaced my 12AX7s with 12 AY7s and it calmed my overly-gainy amp down significantly.

[ac427v]

+1 on using a traditional pre PI master volume. There is way too much gain hitting the phase inverter. or If you really want the LARMAR You can try a voltage divider before the PI. Use the Hoffman Stout schematic in the "Tube Amp Library of Information" on the bottom of this page as a model.
Craig